Digital PDFs

EB-26013-42

2000

306 pages

Original

9.2MB

Document:	Digital;s Networks: An Architecture With A Future
Order Number:	EB-26013-42
Revision:	0
Pages:	306
Original Filename:

OCR Text

Digital' s Networks:
An Architecture With A Future

mamaama

Digital Equiprnent Corporation makes no representation that the interconnection of its products in the manner described herein will not infringe on existing
or future patent rights, nor do the descriptions contained herein imply the
granting of liccnse to make, use, or sell cquipment constructed in accordance
with this description.
The information in this book is subject to change without notice and should not
be construed as a commitment by Digital Equipment Corporation. Digital
Equiprnent Corporation assumes no responsibility for any errors that may
appear in this document.
The following are trademarks of Digital Equipment Corporation: ALL-IN-ONE,
DEC, DEcnet, DEcsystem-20, DECUS, the Digital logo, Micro/Rsx, MicrovAx I,
Observer, PACKETNET, PDP, Professional, RSTS/E, RSX, Tops-20, UNIBUS, VAX, and
VT100.

IBM is a registered trademark of International Business Machines.

Table of Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. VII
CHAPTER 1 . Introducing and Defining DECnet
andDNA........................................ 1
What is DEC net ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1
Uses of a DEcnet Network. . . . . . . . . . . . . . . . . . . . . . . . . . .. 4
Digital Network Architecture ........................ 5
In terfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 6
Protocols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10
Importance of a Network Architecture .................. 13
DEC net Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 14
Topology Alternatives ............................... 16
DEC net Phases .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 19
CHAPTER 2 . DECnet Concepts. . . . . . . . . . . . . . . . . .. 21
Nodes ........................................... 21
Lines and Circuits ................. " ................ 24
Transmission Modes ..... . . . . . . . . . . . . . . . . . . . . . . . . . .. 26
Data Link Protocols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27
DDCMP Fuctional Description . . . . . . . . . . . . . . . . . . . . . . . .. 28
DDCMP Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 31
Data Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 31
Control Messages .................................. 32
Maintenance Messages ............................... 34
DDCMP Operation .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34
Typical Message Exchange ........................... 35
Maintenance Mode ................................. 37
Ethernet Functional Description ....................... 37
Ethernet Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 41
Ethernet Operation ................................. 44
Routing .......................................... 47
Routing Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 49
Routing Algorithms and Databases .................... 52
Routing Features ................................... 53
How Routing Works on an Ethernet LAN. . • . . . . . . . . . . . .. 55

Routing Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 56
Routing Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 59
Data Flow ....................................... '. 64
Data Flow at the Source Node .................. '....... 66
Data Flow Across the Network to the Destination Node .... 68
Data Flow at the Destination Node ... . . . . . . . . . . . . . . . .. 69

CHAPTER 3 . DECnet Configurations . ........... " 71
DEcnet Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 72
Wide Area Networks ............................... 72
Local Area Networks ............................... 74
N ode Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 81
Phase III and Phase IV Nodes. . . . . . . . . . . . . . . . . . . . . . . .. 82
Routing Capabilities ................................ 82
General Purpose and Dedicated Nodes ................ " 84
Line and Circuit Characteristics ....................... 88
The Total Picture ................................. " 89
CHAPTER 4 . Common MechanislDs in
DECnet Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92
Logical Links .................................... " 92
When Are Logical Links Required? ................... 93
How to Create a Logical Link ........................ 94
Behind the SceneS-DEcnet Software and Logical Links .. " 95
End-to-End Communication Layer and NSP . . . . . . • • . . . . " 95
Session Control Layer ........................... ; . .. 99
Requesting a Connection ............................ 101
Receiving a Connect Request ......................... 101
Session Control Messages ............................ 103
How the Program Identifies Logical Links .............. 103
Multiple Logical Links \Vithin a Program ............... 104
Data Types Within a Logical Link ..................... 105
Segmentation and Reassembly of Data .................. 105
Error Control ..................................... 106

III

Flow Control ..................................... 108
Access Control .................................... 109
User Transparency .................................. 111

CHAPTER 5 . Task-to-Task COlnmunications ........ 112
OEcnet Task-to-Task Calls ............................ 112
Requesting a Logical Link ........................... 113
Object Types and Names ............................. 116
Accepting/Rejecting a Logical Link Request ............ 117
Sending and Receiving Data .......................... 119
Terminating the Link ............................... 121
CHAPTER 6· Remote File and Record Access ....... 122
The Data Access Protocol (OAP) Interface ................ 123
Programming Remote File Access ..................... 126
File System Capabilities .............................. 128
Initiating Remote Access ............................ 130
The File Specification ............................... 131
Access-Control Information .......................... 131
File Characteristics ................................. 132
Accessing Remote Files From a Terminal ................ 133
Access Control .................................... 134
File Protection ..................................... 134
Remote File Specifications ........................... 135
Remote Command File Submission .................... 135
CHAPTER 7 . Network Terminal Facilities .......... 136
Interactive Terminal-to-Terminal Communcations ......... 136
Network Virtual Terminal Facilities .................... 138
Network Virtual Terminal ........................... 139
Network Virtual Terminal Operation .................. 146
Local Area Transport (LAT) Protocol .................... 148
Relationship Between LAT and CTERM ................... 149

CHAPTER 8 . Internetwork Conununications ....... 151
X.25 Communicatons .............................. 151
DTES and DCES ..................................... 153
X.25 Circuits ..................................... 153
X.25 Gateway Access Protocol ........................ 155
X.25 Gateway Access Modules ........................ 157
X.25 Access Methods ............................... 165
CCITT Recommendations for X.3, X.28, and X.29 ........ 168
DEcnet/ SNA Communications ......................... 169
DEcnet/ SNA Gateway Access Functions ................. 170
DEcnet/sNA Gateway Access Modules .................. 173
DEcnet/sNA Circuits ................................ 177
Loading the Gateway Node Software ................... 178
CHAPTER 9 . How DECnet Supports Applications .. 179
DEcnet's Task-to-Task Capability in an
Application Environment ............................ 181
Task-to-Task Communication: Order Entry and
Production Scheduling .............................. 182
Network and Application Interaction .................. 183
DEcnet's File Transfer Capability in an
Application Environment ............................ 185
U ser-Initiated File Transfer: Market Research and
Sales Analysis ...................................... 185
Program-Initiated File Transfer: Quality Control and
Vendor Analysis .................................... 188
DEcnet's Remote File Access Capability in an
Application Environment ............................ 189
Network Virtual Terminals. . . . . . . . . . . . . . . . . . . . . . . . ... 189
Short-Term Cash Management ........................ 190
DEcnet's Terminal Facilities in an Application Environment .191
Order Entry and Production Scheduling ................ 192
Market Research ............................. , ..... 193
Short-Term Cash Management ........................ 194
DEcnet as a Consideration in Relocation Planning ......... 196

CHAPTER 10· Network System Management ..... . 198
Network Management Utilities ....................... 199
Network Management Messages ...................... 201
Planning for Node Generation ........................ 205
Generating Network Software ........................ 208
Defining Configuration and Other Static Parameters ...... 209
N ode Addresses and Names .......................... 209
N ode Verification Passwords .......................... 209
Network Object Parameters .......................... 210
Routing Parameters ................................. 211
Line Identification .................................. 212
Circuit Parameters .................................. 212
Transmission Mode ................................. 216
Operating a Node .................................. 216
Controlling the State of a Node ....................... 216
Controlling Line or Circuit State ...................... 217
Monitoring Node Activity ........................... 218
Downline Loading ................................. 220
Downline-Loading Definitions ...................... 221
Downline-Loading Database Parameters ............... 221
Performing a Downline Load ......................... 222
Downline-Loading and Checkpointing RSX-IIS Tasks ..... 224
CHAPTER 11 . Monitoring and Testing
DECnet Performance . ............................ 228
Integrated Testing Tools ............................. 229
Loopback Testing .................................. 229
Upline Dump ..................................... 237
Monitoring Network Operation with Observer .......... 250
Glossary . ........................................ 256
Index . ........................................... 271

VII

PREFACE
Distributed data processing is a computing philosophy Digital
introduced and popularized. For over 20 years, Digital has been
developing products that bring computing power right into
people's work areas, whether these be offices, labs, or factory
floors. Digital has made it possible for people to use computing
resources from terminals on their desks, rather than depend on a
computer in an isolated room. The corporation's wide range of
communications products testifies to its commitment to bringing
computing power to user locations, where the resources are most
needed.
Introduction to DECnet and the Digital Network Architecture (DNA)
discusses DNA-the model of structure and function upon which
Digital's communications products are based-and the various
implementations of DEcnet-Digital's family of networking software. This handbook introduces readers to the capabilities of the
following new products:
• DEcnet-vAX
• DEcnet-RsX (Rsx-llM,RSX-llM-PLUS, and Rsx-lls)
• DEcnet-20
· DEcnet-10
Since 1975, Digital has been augmenting DNA. The architecture
supports a broad range of applications and a variety of network
topologies. (A network topology is a particular configuration of
nodes and lines.) This handbook summarizes DNA design, structure, and functional specifications. The current DNA version is
Phase IV.

VIII

The handbook introduces readers to the following new Phase IV
concepts and products:
• PRo/DEcnet
• Ethernet and related hardware
• DEcnet/sNA Gateway (for VAX/VMS and Rsx-ll systems)
• DEC net Terminal Server
• DEC net Router Server
• DEC net Router/x.25 Gateway
• Observer
Introduction to DECnet and the Digital Network Architecture (DNA)
describes the concepts and capabilities of DEcnet networks. A
DEC net network consists of two or more Digital computer
systems, enhanced with DEcnet software and linked by physical
channels or communication lines. Systems in a DEcnet network
can communicate with each other and share resources. Products
are also available that enable DEC net systems to communicate with
other network entities, such as IBM SNA mainframes and packetswitched data networks (PSDNS). DEcnet networks can be local area
or wide area configurations or configurations that include wide
area and local area networks.

All implementations of DEcnet (DEcnet VAX, PRo/DEcnet, and
DEcnet-Rsx, for example) embody the same network concepts.
The specific capabilities of a DEcnet network depend on the type of
system participating in the network and on the network's application. The objectives of this handbook are:
• To describe the common DNA network concepts behind all implementations of DEC net
• To identify the major network functions that DEcnet provides

One of the primary aspects of a DEC net network is the peer
relationship among its systems. Network operation is efficient,
since any system in the network can initiate and accept communications without the services of a controlling system. This handbook describes the relative ease and speed with which network
users can access the resources of any system in a peer-oriented
network.
This handbook is intended for readers with a knowledge of
communications technology who want to understand the DNA
structure and who want to learn about the concepts and capabilities of DEC net systems. It assumes that the reader is familiar with
Digital operating system concepts, but not with DEcnet. Typical
readers will include the personnel at the site of a newly installed
DEcnet system who can read this manual to learn about the kind of
work DEC net enables them to perform. Another group of readers
will include system managers and designers who are thinking
about using DEcnet to expand the capabilities of their existing
Digital computer systems. And, finally, the handbook is also
intended for system managers and designers who do not yet use
Digital systems but who are considering the implementation of a
computer network.
Introduction to DECnet and the Digital Network Architecture (DNA)
consists of three parts, encompassing eleven chapters:

• The first part, Chapters 1 to 4, introduces DEcnet concepts and
general configuration guidelines for DEcnet networks .
• The second part, Chapters 5 to 9, defines specific network functions
and explains the mechanisms that programs and users at terminals
can employ to implement those functions. This part of the
handbook also presents examples of how these capabilities can be
used in an application environment.

• The third part, Chapters 10 and 11, introduces DEC net functions
related to the management of the systems that make up a DEcnet
network.

Chapter 1· Introducing
and Defining DECnet and DNA

This handbook introduces readers to the Digital Network Architecture (DNA)-the framework of specifications within which Digital
designs its communications products-and to DECnet-the family
of networking products that Digital has implemented in accordance with DNA specifications. Readers will learn how and where
DEC net is used and what capabilities it offers. In addition, they will
understand the importance of a network architecture such as DNA.
There are several implementations of DEcnet, one for each operating system from Digital. This handbook discusses features and
capabilities that are common to all DEcnet implementations. For
details on specific implementations, refer to the documentation
accompanying the implementations .

. What is DECnet?
DEcnet is a family of software and hardware communications
products that enables Digital operating systems to participate in a
network environment. A network is an entity of two or more
computer systems that are connected by physical links (cable,
microwave, or satellite, for example) for the purpose of exchanging data and sharing resources. Every system in the network is
called a node. Figure 1.1 illustrates an example of a DEC net
network.
This handbook refers to two types of network users:
(a) individuals who employ the network to perform such highlevel functions as accessing a file on a remote node or updating a
database on a remote system. In this handbook, these individuals
are called "users." They usually have little knowledge of the
details of how data is transferred from one node to another and are
not concerned with the technicalities of network performance.

2· Introducing and Defining DECnet and DNA

Transmit Via Satellite
(Physical Link)

Figure 1.1
A DECnet Network

(b) individuals responsible for configuring and maintaining the
network, for monitoring network performance, and for ensuring
a smooth-running network environment. These individuals are
called' 'network managers." Occasionally, the manual refers to
"system managers," individuals who are responsible for the operation of the individual nodes making up a network. Although
network managers are primarily responsible for implementing
and maintaining networks, system managers can contribute
greatly to high-quality network performance by closely monitoring their specific nodes.

DEC net permits great flexibility in configuring a network. A
DEcnet network can start with two nodes and be expanded to
include over 64,000 nodes. Network managers can configure a local
area network, with nodes located in a building or spanning a cluster
of buildings, and expand the configuration to a wide area network,
with nodes distributed in different cities across the country and
even in different countries. Optionally, very large DEC net networks (with over 1,023 nodes) can be configured as interrelated
groups of nodes or" areas." DEC net can support up to 63 areas with
up to 1,023 nodes in each area. Area-based networks are particularly effective in large organizations; each department can have a
separate area in a corporate-wide network.
The nodes in a DEcnet network can run any of Digital's many
operating systems. A DEcnet network can comprise any combination of VAX systems running VAX/VMS; MicrovAx I systems
running MicrovMs or VAXELN; PDP-ll systems running Rsx-llM,
Rsx-lls, Rsx-llM-PLUS, RSTS/E, or RT-ll; DECSYSTEM-20s and
DEcsystem-10s running Tops-20 and Tops-10 respectively; and the
Professional 350 personal computers running the p/os operating
system.
A very important feature of a DEcnet network is that its nodes have
a peer relationship with each other. This means that every node in
the network can communicate with every other node without
having to consult a central controlling node. The advantage of
such a decentralized network is that each node can be equally
responsive to user requests. A network user at one node can easily
gain access to applications and facilities that exist on other nodes.
Since every data exchange and remote ·access operation does not
have to pass through a central node, communication overhead is
reduced and network performance increases in efficiency.

4· Introducing and Defining DECnet and DNA

Decentralization also provides a great deal of flexibility in the way
networks can be configured. Chapter 3 describes the type of
configurations possible in a decentralized network .
. Uses ofa DECnet Network
It is DEC net that enables various Digital systems in an organization, which are fulfilling differing application requirements, to
communicate among themselves. It lets them exchange data and
share resources, thereby enabling the implementation of all efficient data processing ~peration.
Networked systems can share expensive resources such as laser
printers and mass storage devices. Costly peripherals, therefore,
don't have to be duplicated.
Systems in a network can exchange information so that fewer
copies of a given file can be maintained. That way, users can be
assured that they are accessing the most current version of a file. In
addition, a network can speed up information flow within an
organization, since file transfers between systems are rapid.
In a network, data that has been entered interactively at a personal
computer can be sent to update a database on a large system. The
resources of the large system are not tied up in interactive data
entry and are available for complex tasks requiring extensive cpu
power.
Networks make it possible for small systems to take advantage of
the greater computing power of larger systems. Tasks that need
extensive computation can be sent to large systems for quick
execution. Conversely, large systems can take advantage of smaller
systems by offloading certain applications onto machines designed
for those specific jobs.

From their terminals, network users have access to a wide expanse
of resources. They can access the computing power or data
resources of any node on their DEcnet network. They can also
access the resources of non-Digital systems on other networks.
DEC net supports gateways that enable Digital systems to communicate with systems from other vendors. The DEcnetl SNA Gateway enables nodes in a DEcnet environment to communicate with
systems in an IBM System Network Architecture (SNA) environment. The DEcnet Router/x.25 Gateway enables DEcnet nodes to
communicate with Digital and non-Digital systems implementing the x.25 protocol and connected to packet-switched data
networks (PSDNS). See Chapter 8 for further explanation of
communication with foreign-vendor networks.
The gateways enable organizations to maintain vendor independence and take advantage of the capabilities of systems from
different manufacturers. At the same time, networks make it
possible for the different systems in an organization to exchange
information and share resources.
Gateways offload communication functions like routing and
protocol translation from network nodes, freeing up the latter to
process user applications with greater efficiency. See Chapter 3 for
further details .

. Digital Network Architecture
The Digital Network Architecture (DNA) is the framework for all
Digital communications products. The International Standards
Organization (ISO), a group responsible for providing industry
standards, has created an architectural model for open systems
interconnection (OSI). Figure 1.2A shows that DNA and the ISO
model are similar. DNA and the ISO model have evolved from
common roots.

6· Introducing and Defining DECnet and DNA

Architecture Layers

ISO

DNA
User

Application
Network
Management
Presentation

Network
Application

Session

Session Control

Transport

End-to-End
Communication

Network

Routing

Data Link

Physical Link

Figure 1.2A
ISO and DNA Models

DNA includes the specifications that govern the interrelationship of
the various software components making up DEcnet. There are
two kinds of intercomponent relationship that DNA defines:
interfaces and protocols.
Interfaces
Interfaces are definitions of specific functional boundaries between
DEC net software components residing within a single node. The

boundaries are structured as a hierarchical set of layers (shown in
Figure 1.2A). DEC net software is arranged, according to function,
as modules within these layers (see Figure 1.2B). DNA layers, from
highest to lowest, are:

User Layer-The user layer encompasses user-written programs and
user-level services that access the network, network services that
directly support user and application tasks, and overall system
management. The most common services accessed by users at this
level include resource sharing, file transfers, remote file access,
database management, and network management. All are
described in greater detail in subsequent chapters.
Network Management Layer-The Network Management layer
defines the functions used by operators and network managers to
plan, control, and maintain DEC net networks. This layer interfaces
with all the DNA layers below it to gather data on network
performance and change network operational parameters. Functions include downline loading of remote nodes and loopback
network line testing. Chapters 10 and 11 discuss network management in greater detail.

Network Application Layer-The Network Application layer, which
contains user- and Digital-supplied modules, defines network
functions used by the User and Network Management layers. The
most important DEC net functions currently operating in this layer
are remote file access, file transfer, remote terminal capability
(including the network virtual terminal function), access to x.2S
connections using a Packetnet System Interface (PSI) or x.2S/x.29
extension package, and access to SNA Gateways. This layer also
defines a universal II a language within DNA so that resource
sharing among heterogeneous systems is simplified. Chapters 6
through 8 discuss the functions of this layer.

8 • Introducing and Defining DECnet and DNA

The Network
Control Program,
which allows
manager/operator
to monitor/control
network activity

CD A user program
communicating
with a
remote program

® A user program
accessing
remote files

User
Layer
Network
Management
Layer

Network
Management
Routines

Network
Application
Layer
Session
Control
Layer
End-to-End
Communication
Layer
Routing
Layer
Data
Link
Layer
Physical
Link
Layer
Communications
Facilities

Line A (e.g., a
telephone line)

Line B (e.g., a
local cable)

Figure 1.2B
DNA Modules Resident in a Typical
DECnetNode

Line C (e.g., a
Lines D & E
telephone line) (Ethernet cables)

Session Control Layer-The Session Control layer defines the
system-dependent aspects of process-to-process communications.
The functions of this layer include name-to-address translation,
process addressing, and access control. See Chapter 4 on common
mechanisms in DEcnet functions for a further explanation of the
role of the Session Control layer.
End-to-End Communication Layer-The End-to-End Communication layer software handles the system-independent aspects of
communications. These include connection management, data
flow control, end-to-end error control, and segmentation/
reassembly of user messages. See the discussion on common
mechanisms in DEC net functions in Chapter 4 for a discussion of
these concepts.
Routing Layer-The Routing layer software defines the mechanism
for transporting or routing user data from the sending node to the
receiving node. Modules in this layer also provide network
congestion control and packet lifetime control. See Chapter 2 for
an explanation of these concepts.
Data Link Layer-Modules in the Data Link layer define a mechanism for creating an error-free communication path between
adjacent nodes, whether they are connected by an x.2S link, an
Ethernet link, or a Digital Data Communications Message Protocol (DDCMP) link (point-to-point or multipoint, as described in
Chapter 3). See Chapter 2 for a discussion of DDCMP and Ethernet
and Chapter 8 for a discussion of x.2S.

10· Introducing and Defining DECnet and DNA

Physical Link Layer-The Physical Link layer defines the manner iIi
which device drivers and communications hardware (modems and
lines, for example) should be implemented to move data over a
transmission line. The functions of modules in this layer include
monitoring channel signals, clocking on the channel, handling
interrupts from the hardware, and informing the Data Link layer
when a transmission has been completed.
A module can use the services provided by modules in a lower layer
and can provide services to modules in a higher layer. A module
cannot access services offered by a module at a higher layer. The
Network Management layer modules are the only ones that can
interface directly with modules in each of the lower layers for access
and control purposes.
In building-block fashion, DNA causes each layer of software to
build on the services offered by the modules in a lower layer. DEC net
has been designed in accordance with these rules so that high-level
network functions, like program-to-program communication,
occur in a predictable and regulated fashion on each DEcnet node.
Protocols

Modules with equivalent functions in the same layer, but residing
in different nodes, communicate using protocols. A protocol is
both a set of messages and the rules for exchanging the messages.
DNA defines the message formats. and rules very specifically.
Protocols provide a common code of behavior that two computer
systems in a network can use to understand the role each is playing
in corumunications and information exchanges. Figure 1.3 illustrates the relationship between protocols and interfaces. Figure 1.4
illustrates protocol communication between two nodes.
DNA does not define protocols for the User Layer. In this layer,
programs must include the means to interpret received data (this

Mo~ules

,
1

I~ Protocols

---I

Interfaces

Modules

,
1

Layer n

Interfaces
Layer 3

~Protocols

Layer 2

M_o_d_u_l_es_---'~ Protocols

Layer 1

'---__

Node A

NodeB

Figure 1.3
Relationship between Protocols and
Intetfoces

process is similar to interpreting the contents of a record on disk).
Some DNA layers support more than one function and therefore
include multiple protocols. An example is the Data Link layer,
which supports the DDCMP, x.2S, and Ethernet protocols. The
protocols that DNA defines are not exclusive; customers can
substitute their own, if they choose, as long as these protocols are
implemented consistently by equivalent modules throughout the
network.
The DNA framework of interfaces and protocols is transparent to
most user activities. Users need not concern themselves with how
data moves through the network.

12· Introducing and Defining DECnet and DNA
Node 1
Modules

User Program

Node 2
Modules

Protocols

-----------

t--

User Layer

User-Defined Protocol

---1

User Program

----------Network Management Layer
Network Information and
Control Exchange
Protocol (NICE)

Network
Management
Module

----------Network Application Layer

Network File
Access Routines
(NFARs)

File Access
Listener
(FAL)

Data Access
Protocol (D AP)

----------Session Control Layer

Session Control
Module

Session Control Protocol

End-to_E~~t~::~::V~~::ion Layer1

NSP Module

r
-

Protocol - - - - - . (NSP)

NSP Module

=============== ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ '===============
Routing Layer

I Routing Module -1IIrt----- Routing Protocol -----11~ Routing Module
I ~DCMP .Module

t...

Data Link Layer
Digital Data
Communications
Message ~
.
DDCMP Module I
Protocol (DDCMP)

----------Physical Link Layer

Device
Controller

Electrical Signals

Device
1 Controller

r- - - - - - - - - - - l
Communication lines
forming physical
connection

Figure 1.4
Protocol Communications between Two
Nodes

. Intportance ofa Network Architecture
DNA is an open-ended architecture designed to absorb new and
increasingly efficient communications technologies. It is highly
adaptable to different -communications situations and supports a
broad range of user applications and functions. Digital's adherence
to a network architecture in the design of its communications
products results in several important advantages for network
customers.

Communications technology is changing continually. Future
phases of DNA may model additional layers, additional modules
within layers, or alternative modules for certain layers. With the
wide range of available DNA facilities, organizations will be able to
implement networks that are closely tailored to meet specific
application needs without incurring the expense of custom-designed
networks. Currently, DNA'S Data Link layer supports three
protocoIS-DDCMP, x.2S, and Ethernet. Network managers can
implement the protocol best suited to meet their organization's
application needs. For example, one organization's network of
systems located within a limited geographical area may have a very
high traffic volume among nodes. In this case, the Ethernet
protocol over an Ethernet cable would be a good solution. For a
network with medium traffic volume among geographically
dispersed systems, the x.2S protocol might be a more costeffective solution.
A second advantage of a network architecture is that it specifies
common communication mechanisms and user interfaces to
which computer systems of different types must adhere when
communicating. Without such a networking standard, data
exchange and resource sharing among the variety of available
operating systems, communication devices, and .computer hardware would be highly difficult to effect.

14· Introducing and Defining DECnet and DNA

Managing the different systems in a network would also be a very
complicated task in the absence of a network architecture. Network management, error recording, and maintenance are easier
when there are standard procedures for error detection, recording,
isolation, recovery, and repair. An architecture provides this
standardization so that network management can take place
consistently across different systems. Standardization enables a
network manager to manage an unattended remote system from a
local node, even if the remote node is running a different operating
system than that of the local node.
Yet another benefit of a network architecture is that it enables
system-'independent functions to be designed specifically enough
to be implemented in hardware for improved performance. For
example, the DEUNA communications device implements the
Ethernet data link protocol; the DMRll , DMPll, and DMvll
implement the DDCMP data link protocol; and the KMSll-BD
implements the x.25 frame-level protocol.

. DEC net Capabilities
DEC net enables users and applications to perform a number of

high-level network functions. These include:
Task-to-task communication-Cooperating programs running under
different operating systems and written in different languages can
exchange data. For example, PDP-ll medical analysis systems can
transfer data to a central VAX system without an operator's
commands. See Chapter 5 for further details on task-to-task
communication.
Remote file and record access-Programs and users can access files that
reside on remote nodes. Remote file access facilities enable the

transfer of files between two nodes, the manipulation of files
(opening a file, deleting data from or appending data to a file) on
remote nodes, and the submission of command files for execution
to a remote node in order to gain access to that node's resources
Consult Chapter 6 for additional information on remote file and
record access.

Terminal-to-terminal communication-On some DEcnet implementations, a user at one terminal can use a DEcnet utility to send
messages to users at other terminals in the network. Refer to
Chapter 7 for further explanation.
Network terminal communication-From their terminals, users can
log on to a remote node and execute commands as if they were
typing on a local terminal at that node. DEC net Phase III supports
this capability if the remote node is running the same operating
system as the user's local node. DEC net Phase IV supports this
capability even if the operating system on the remote node is
different from that on the local node.
The Phase IV capability is known as network virtual terminal. This
capability is particularly useful for program development. From
their terminals, programmers can draw on the resources of any
system in the network. Depending on the application they are
developing, they can log on to the system best suited to their task.
For example, for rapid compilation, they can submit source-code
files to a high-performance VAX/VMS system. See Chapter 7 for
more details on network terminal communication.

Problem isolation and Network management-Network managers can
use DEcnet commands and procedures to generate and define nodes

16· Introducing and Defining DECnet and DNA

and to isolate network problems. By performing loopback tests,
trace facilities, and dump analyzers (see Chapters 10 and 11 for
further explanation of network management functions), network
managers can monitor network nodes and keep abreast of network
performance in order to identify and resolve potential difficulties
before they become crises. Users can thus be assured of an errorfree network.
Downline loa ding-With this DEenet capability, an entire software
system or task developed on a node with adequate peripheral and
memory support can be sent to target systems for execution. For
example, programmers in a manufacturing setting can develop
new process monitoring and control parameters on a central
system and then transfer the parameters to smaller target systems
located on the factory floor.
Upline dumping-When a target system begins to fail, DEenet's
upline dumping capability automatically sends a system-image
dump upline to an adjacent, larger system. For example, an RSX11s DEe net node that has crashed can upline dump a copy of its
system image to an adjacent Rsx-llM DEe net node for interpretation. When the failing node is restored, it can be downline loaded
to resume operation .
. Topology Alternatives
DEenet provides considerable flexibility in configuring a network.
A wide area network can include nodes distributed across the
country and even in different parts of the world. Depending on an
organization's application requirements, a network manager can
decide where the nodes will be located, which nodes will be
directly connected to each other, and what physical channels
(leased lines, coaxial cable, and PSDN links, for example) these
connections will use.

Greatly facilitating the configuration of wide area networks is a
DEcnet feature known as routing. Routing eliminates the need for
every node in a D£cnet network to be physically connected to every
other node in the network. Communication hardware and line
costs are greatly reduced. With the reduction in the number of
physical lines, the risk of downtime due to channel failure also
decreases.
Routing ensures that information from a source node is delivered
to a specified destination node even if the information has to travel
through several intervening nodes. In the event a line fails,
DEcnet's adaptive routing feature finds an available alternative path.
DEC net Phase IV offers area routing, the method by whi~h DEC net
can support very large networks (up to 63 areas with up to 1,023
nodes in each area). In an area-based network, two types of
routing mechanisms operate: Levell routing is the mechanism
responsible for communication within an area; Level 2 routing is
responsible for communication between areas. Levell and Level 2
routing enable a node in one area to exchange data and share
resources with any node in any other area of the area-based
network. Chapter 2 includes a comprehensive discussion of
routing in networks with and without area routing.
The most recent version of DEcnet, Phase IV, incorporates the

Ethernet baseband local area network specification. What this means
is that systems running DEC net Phase IV can be attached to a
coaxial cable in order to communicate at high speeds (a maximum
cable data rate of 10 Mbits per second). Using DEC net facilities,
individuals on an Ethernet network can perform any of the highlevel network functions described above.

18· Introducing and Defining DECnet and DNA

In addition to enabling high-speed communications, Ethernet
networks support a variety of communication servers (see Chapter
3 for details) that greatly reduce the cost of implementing a
baseband local area network and significantly increase network
performance. Communication servers are special purpose or dedicated nodes that perform communications functions for the other
nodes in the network, thereby freeing the latter of communications overhead.
DEcnet Phase IV also incorporates the X. 25 protocol for communication over packet-switched data networks (PSDNS). A DEC net wide area
network can, therefore, include systems linked through a PSDN.
x.2S is a protocol recommended by the International Telephone
and Telegraph Consultative Committee for data communications
over common-carrier lines. By linking geographically dispersed
systems through a PSDN, the cost of data communications can be
reduced significantly. The tariffs for a packet-switched data network are based on the volume of data sent, rather than the distance
or connect time between communicating systems. Cost savings
with PSDN links can be substantial, especially when there is
medium traffic volume among systems.
Through ~me physical link to a PSDN, logical links among many
systems can be created. See Chapter 2 for an explanation oflogical
links. Users need not worry about how the information is routed
through the PSDN; the carrier that supplies the PSDN is responsible
for ensuring that data is routed to the correct destination.
The incorporation of the x.2S protocol in DEcnet Phase IV gives
users of packet-switched data networks easy access to all high-level
DEC net functions. PSDN users need not write any additional code
for their application programs in order to perform network
functions such as file transfer, remote resource access, network
virtual terminal, and program-to-program communication
between DEC net Phase IV systems linked across a PSDN.

The Digital Data Communications Message Protocol (DDCMP) is the
third protocol that DEcnet Phase IV incorporates at the Data Link
level of DNA. Depending on what is best for their applications,
network managers thus have the option of configuring networks
with links that support DDCMP, x.2S, or the Ethernet protocol.
Chapter 2 discusses these protocols in further detail.
In applications wherein peer relationship among nodes is not
critical, organizations can save dramatically on expenses for lines
and communication hardware by implementing a multipoint configuration. In such a configuration, several nodes are dropped from a
single line with a central node controlling access to that line. (See
chapter 3 for additional information on multipoint lines.)

• DECnet Phases
Digital has been developing networking products based on DNA
since 1975. This development effort has occurred in a series of
phases. So far, there have been four DEC net phases. The products in
each new phase are fully compatible with those produced in the
previous phase. The sections below list the capabilities of Phase III
and Phase IV DEC net nodes. (Appendix B in the Overview ofDig ita I
Networking Products contains a brief history of DEcnet and discussions of earlier DEC net phases.)
Phase III
Functions that DEcnet Phase III systems can perform include taskto-task communication, remote file access, terminal-to-terminal
communication, network terminal communication (between systems running the same operating system-homogeneous systems),
network management, downline loading, upline dumping, and
loopback testing.
In addition, Phase III systems can perform message routing,
multipoint communication, and communications with packetswitched data networks using x.2S protocols.

20· Introducing and Defining DECnet and DNA

The following DEC net products provide Phase III capabilities:
DEcnet-20, DEcnet-l0, DEcnet/E, and DEcnet-RT.

Phase IV
Phase IV includes additional capabilities that establish Digital as a
leading supplier of products for multivendor network environments. In addition to being backward-compatible with Phase III,
Phase IV features include:

Data link independence, to increase data link options for customers
in implemehting networks. DEC net Phase IV-supported data link
protocols are x.25, DDcMP,and Ethernet.

Dedicated communications server products that offload communications overhead from general purpose nodes (see Chapter 3).

Increased network routing support allowing over 64,000 nodes to be
configured in a network.

Enhanced network management capabilities (see Chapters 10 and 11).
Gateway support for communications with nodes using x.25
protocol over a packet-switched data network (see Chapter 8).
Gateway support for communications with IBM systems in an SNA
network (see Chapter 8).
DEcnet-vAX, DEcnet-Rsx (for 11M; 11M-PLUS,l1S, and Micro/Rsx
systems), DEcnet-20, and PRo/DEcnet have already implemented
Phase IV capabilities. PRo/DEcnet enables a Professional 350
personal computer to be connected to an Ethernet local area
network to exchange information and share resources with other
Phase IV nodes. Through a DEcnet Router Server (see Chapter 3),
the Professional 350 on an Ethernet local area network can
communicate with Phase III and Phase IV systems that are not on
the Ethernet.

Chapter 2' DECnet Concepts

Network activity involves the flow of information between nodes
across lines and circuits. Data originating at a node is routed through
the network until it reaches a specified destination node.
Nodes, lines, circuits, and routing are some major DEC net concepts. Knowledge of them is helpful in understanding how data
flows through the network. A knowledge of routing, in particular, is essential in appreciating the flexibility DEC net permits users
and network managers in performing their tasks .
. Nodes
A node is a network entity-an identifiable unit capable of
processing, sending, and receiving network information. A node
consists of an autonomous software entity, such as an operating
system, with processors and other associated hardware that
implement DNA specifications. Every node in a DEC net network
has a unique numeric address. This address designates the location
of a node in the network, just as a street address designates the
location of a building in a town.

In area-based networks, a node's address includes an area number
and a node number. The area number specifies the area in which
the node is located, and the node number identifies the node
within the specified area. A node address in an area-based network
is a 16-bit number. The first 6 bits are reserved for the area number,
and the next 10 bits for the node number. This division of the 16bit node address limits the maximum number of areas in an areabased network to 63 (26 - 1 = 63), and the maximum number of
nodes in an area to 1,023 (2 10 - 1 = 1,023).
Data traveling from one node to another is enveloped by control
information, which includes the address of the sending node,
called the source address, and the address of the receiving node,
called the destination address. (Data enveloped by control information is a packet.) These addresses inform nodes intercepting the

22· DECnet Concept

packet en route to its destination where the packet is going. Node
addresses ensure that a data packet can always be identified in terms
of the node that is sending it and the node that is receiving it,
regardless of the path the packet travels. (See below, under Routing,
for a discussion of the routing mechanism that determines data
packet paths.)
Most users find it difficult to remember numeric addresses,
particularly when there are many nodes in a network. Software in
the Session Control layer of DEcnet eliminates this difficulty. The
system manager for each node in the network specifies names for
all the nodes with which that node expects to communicate.
These names are defined and stored in a database on that node and
are meaningful only to the Session Control software running on
that node.
Against the names in a node's database, Session Control software
stores the unique numeric addresses associated with the names. A
single node can be known by several names, since system managers
for the different nodes in a network can assign a different node
name for the same remote node. Consider a node in San Francisco.
This node communicates with nodes in Boston, New York, Los
Angeles, and New Mexico. Users at these other nodes access the
San Francisco-based node by different names. Although a remote
node can have several names, it has only one address. This address is
consistent throughout the network, regardless of the name by
which the node is accessed.
A Session Control module in each node maintains a node-name
mapping table that stores the unique node addresses associated
with the names of all the nodes with which that node communicates. Although the node-name mapping tables in the Boston,
New York, Los Angeles, and New Mexico nodes will contain

different names to refer to the node in San Francisco, the tables in
each of these four nodes will have the same address against the
name by which they refer to the San Francisco node.
When a user requests connection to a node and refers to it by
name, the Session Control module consults the node-name mapping table to determine the address of the destination node. It
passes on this information to the End-to-End Communication
Layer, which creates logical links between systems (see Chapter 4).
For incoming connect requests from the End-to-End Communication Layer, the Session Control module uses the node-name
n1apping table to identify the node from which the request
originated.
Session Control software enables users to refer to nodes by names
that are meaningful to them (and which they can therefore
remember easily). By means of the node-name mapping table,
Session Control software finds the correct address of a node so that
data arrives at its proper destination.
When system managers initially generate or configure their node
into a network, they usually assign the name and address by which
the node is to be known in the network. For small networks, these
assignments can be hand-typed into each system. For large
networks, an up-to-date master list of node addresses and names
can be maintained on a particular node so that other nodes in the
network can simply copy the information from the master file.
(Existing nodes in the network exchange their node naInes and
addresses with each newly configured node.)
Other node characteristics, such as routing capabilities and dedicated functions, are discussed in Chapter 3.

24 • DECnet Concept

. Lines and Circuits
A line is a physical data path from a DEe net node to another DEenet
node or from a gateway to an IBM SNA network or packet-switched
data network (PSDN). A line is connected to a node by a hardware
controller that drives or controls the line. Lines that use the public
telephone network are often connected to a modem. Lines can be
either dedicated (hard-wired) or dialup. Lines can be attached to a
local area network cable or configured in point-to-point or
multipoint configurations (see Chapter 3 for details on the latter
two configurations).

Lines form the physical connections between adjacent nodes. There
is a higher level of internodal connection called circuits. Circuits
are logical connections that carry information between two nodes
and operate over the physical medium oflines. Each point-to-point
line has one circuit associated with it, and each tributary on a
multi poin t line has a circuit (see Figure 2.1). A single physical
connection to an x.2S packet-switched data network (PSDN) can
support many virtual circuits.
An Ethernet baseband network consists of one physical line with
multiple circuits. Each circuit corresponds to a particular Ethernet
protocol type. An Ethernet protocol type identifies the type of
operation performed over a circuit. Figure 2.1 shows four Ethernet
protocol types:
DNA routing protocol enables the operation of the DEenet routing
mechanism and is required for all DEe net communications.
DNA remote console protocol provides simple internodal terminal
communications used for diagnostic and maintenance procedures.
DNA dump/load protocol enables operators/managers to load system software remotely in nodes that do not have mass storage (for
example, communication servers).

Point-to-Point Connection

Multipoint Connections
I'

Node A
J------/-i

-Ci~c~it: - -Ci~c~it A-C - -: - - Cir~~it AD- - :
1
1
AB :
1
1

1
1

NodeB
1/

NodeC
/

NodeD
I'

Ethernet Connections
Remote Console Circuit _____________ I
Loop C·lrcmt
. __________ 1_ - - - - - - - - -:- - -1- - - - - - - - - - - - I
Routin~ Circuit --- _ -Dump/Load Circuit _ _
~

_1:_ -------- -:--1-: ----------I :
1 1
1 1

Node A

Figure 2.1
Circuits

1
1

1 1
1 1

NodeB

1
1

NodeC

26 • DECnet Concept

Loopback protocol enables the performance of loop tests and is
required for any type of testing procedure (see Chapter 11).
There are other protocol types not shown in Figure 2.1. Refer to
the bibliography in the Overview ofDigita INetworking Products for a
list of Ethernet documentation that further explains protocol
types.
All communication activity in the network is considered in terms
of circuits. DEC net 's routing mechanism establishes a succession of
circuits between a number of intervening nodes to get data from
one node to another. These circuit pathways enable nodes to send
data to other nodes that are not physically adjacent to them.

Transmission Modes
Information is transmitted over DEcnet lines in half-duplex or fullduplex mode. In half-duplex mode, the line can transmit data in
either direction, but only in one direction at any given time. Data
cannot be sent and received over the line simultaneously. In fullduplex mode, the line can transmit data in both directions
simultaneously, thereby making it possible for a node to send and
receive data at the same time. The concepts of half duplex and full
duplex relate to the physical line only.
Information flow at the higher level of the circuit always appears to
be full-duplex. User data passes through the various DNA layers
until it reaches the routing mechanism in the Routing layer. On
the basis of predefined information, the routing mechanism
determines the circuit over which the data will be transmitted to
an adjacent node. The data then passes down to the Data Link
layer.

Once the data leaves the Routing layer, the user services and the
routing mechanism assume that it has been transmitted over the
specified circuit. However, if the data cannot be transmitted over
the designated physical line because it is a half-duplex line already
carrying data, or, in the case of an Ethernet transmission, because
the channel is not free, software in the Routing layer will buffer
the data until it can be transmitted. The user is not aware of when
the data is actually buffered or when it is sent out directly on a free
line .
. Data Link Protocols
Currently, there are three protocols residing in the DNA Data Link
layer:

Digital Data Communications Message Protocol (DDCMP)-a byteoriented protocol
Ethernet protocol-Carrier Sense Multiple Access with Collision
Detect (CSMAI CD) with physical channel encoding and operating
over a coaxial cable
X.25 Levels 2 and 3-operating over Level 1 of the CCITT x.2S
recommendation, which defines a standard interface between data
terminal equipment (DTE), such as a DEcnet node, and the data
circuit terminating equipment (DCE) of a packet-switched data
network.
This chapter will describe the DDCMP and Ethernet protocols. For
additional details on the x.2S protocol, see Chapter 8.
The primary advantage of DEcnet's support of three data link
protocols is that network managers can configure networks based

28 • DECnet Concept

on the data link protocol that best suits their organization's
application requirements. Regardless of the data link protocol
they implement, network managers can be assured of error-free
transmission of data, since a primary function of the Data Link
layer is to ensure the integrity of transmitted data.
DDCMP Functional Description
There are three general types of data link protocols: byte-oriented,
character-oriented, and bit-oriented. DDCMP is a byte-oriented protocol. Such a protocol provides a count of the number of bytes that are
sent in the data portion of each message. A character-oriented
protocol uses special ASCII characters to indicate the beginning of a
message and the end of a block of text, and a bit-oriented protocol
uses flags to frame data sent in undefined lengths. Neither the
character-oriented nor the bit-oriented protocol contains provisions for checking whether all the transmitted data has arrived at
its destination. The advantage of a byte count in a byte-oriented
protocol is that it facilitates checking on the part of the receiving
node to see whether all transmitted data has been received.
DDCMP was designed in 1974 specifically for the Digital Network
Architecture. DDCMP is functionally similar to HDLc-High-level
Data Link Control-which was adopted in 1975 by the International Standards Organization. (HDLC is a bit-oriented protocol,
however.) Another type of data link protocol that is commonly
implemented is BISYNC (binary synchronous), which is characteroriented.
DDCMP is a general purpose protocol. It operates on a variety of
communications devices, from the very small to the very large.
DDCMP makes maximum use of channel bandwidth and handles
data transparency efficiently (data transparency is the capability of

receiving, without misinterpretation, data containing bit patterns
that resemble protocol control characters). Character-oriented
protocols cannot handle transparent data as efficiently as byte- or
bit-oriented protocols. Other major features of DDCMP include
error recording to warn of impending channel failure on degraded
lines and a simplified mode for bootstrapping and testing functions.
DDCMP transmits data grouped into physical blocks known as data

messages and provides a mechanism for exchanging error-free
messages. This mechanism works in the following manner:
DDCMP assigns a number to each data message, beginning with the
number one (after each initialization) and incremented by one for
each subsequent data message. In addition, DDCMP places a 16-bit
cyclic redundancy check (cRc-16) error-detection polynomial at
the end of each data message transmitted.
The receiving DDCMP module checks for errors and, if there are
none, returns an acknowledgement that it has received the
message. Acknowledgement is efficient since the receiving DDCMP
module does not have to acknowledge each message sent.
Acknowledgement of data message n implies acknowledgement
of all data messages up to and including data message n. If the
receiving DDCMP module detects an error, it uses time-outs and
control messages to resynchronize and trigger retransmission;
DDCMP FeatureS-DDcMP is a powerful and versatile data link

protocol. It has been designed to provide an error-free communications path between adjacent nodes. Features include the ability to
• Obtain data from the Physical Link layer in blocks of 8-bit bytes .
• Sequence data by message numbers.

30 • DECnet Concept

• Send up to 255 messages without waiting for individual acknowledgement of each successive message. This capability is known as

pipelining.
• Operate independently of channel bit width (serial or parallel) and
transmission characteristics (asynchronous or synchronous).
• Operate with a wide variety of communications hardware and
modems.
• Detect errors by means of CRc-16 message trailers.
• Retransmit to correct errors.
• Achieve optimum performance with techniques such as pipelining, piggybacking (sending an acknowledgement within a returned
data message), and implying a postive acknowledgement of
previous messages by negative acknowledgement of current message (see below, under DDCMP Control Messages).
• Operate in half-duplex and full-duplex modes.
• Support point-to-point and multipoint communications.
· Synchronize transmission and reception on byte and message level.
• Frame (construct envelopes) data messages.
• Provide a maintenance mode for diagnostic testing and bootstrapping functions.
• Provide data transparency.
• Notify the other end of the link when restarting or initializing.
• Maintain error counters.
· Record the occurrence of events for automatic error reporting to
the user.

DDCMP Messages
There are three types of DDCMP messages: data, control, and
maintenance. Data messages consist of user data. Control messages return acknowledgements and other control information to
ensure data integrity and error-free transmission. Maintenance
messages consist of information for downline loading, upline
dumping, link testing, or controlling a remotely located, adjacent
system .
. Data Messages
DDCMP formats all messages received from the Routing layer to be
sent across the physical link into a data message format (Figure
2.2). The data message format ensures proper handling and error
checking of both the header information and the data being sent.
(In the message format figures in this chapter, the numbers below
each message field indicate the length of the field in bits.)

Data Message Format

ISOH ICOUNT IFLAGS IRESP INUM IADDR IBLKCK11 DATA IBLKCK21
8
14
SOH
COUNT
FLAGS
RESP
NUM
ADDR
BLKCKl
DATA
BLKCK2

the numbered data message identifier
the byte count field
the link flags
the response number
the transmit number
the station address field
the block check on the numbered message header
the n-byte data field, where O<n = COUNT<2 14
the block check on the data field

Figure 2.2
DDCMP Data Messa,Re Format

32· DECnet Concept

. Control Messages
DDCMP has five control messages that carry link control information, transmission status, and initialization notification between
DDCMP modules. All the control messages operate to ensure an
error-free link between sending and receiving DDCMP modules:

Acknowledge Message (ACK) acknowledges the receipt of correctly
numbered data messages that have passed the CRc-16 check. The
ACK message is used when no numbered data messages are to be
sent in the reverse direction. The ACK message conveys the same
information as the RESP field (see Figure 2.3) in numbered data
messages.
Negative Acknowledgement Message (NAK) passes error information from the DDCMP data receiving module to the DDCMP data
sending module. The NAKTYPE field (see Figure 2.3) indicates the
cause of the error. The NAK message servers two purposes: it
acknowledges receipt of all previously transmitted messages with
a number less than the current message number received and it
notifies the sender of error conditions relating to the current
message.
Reply to Message Number(REP) is a message from the data sender to
the data receiver requesting a received-message status. The data
sender issues a REP message when it has sent a data message, has not
yet received acknowledgement of that message, and the time
allocated of an acknowledgement has expired.
Start Message (STRT) establishes initial contact and synchronization on a DDCMP link. A DDCMP module sends this message during
link startup or reinitialization.
Start Acknowledge Message (STACK) is the response to a STRT
message. It tells a receiving DDCMP module that the transmitting
module has completed reinitialization.

Figure 2.3 shows the formats of the DDCMP control messages.

Acknowledge Message (ACK) Format

I ENQ I ACKTYPE , ACKSUB IFLAGS IRESP , FILL IAD DR IBLKCK31
8

Negative Acknowledge Message (NAK) Format

I ENQ , NAKTYPE IREASON' FLAGS IRESP IFILL IADDR IBLKCK31
Reply to Message Number (REP) Format

I ENQ I REPTYPE IREPSUB IFLAGS IFILL INUM IADDR IBLKCK31
Start Message (STRT) Format

IENQ I STRTTYPE ISTRTSUB , FLAGS IFILL' FILL' ADDR , BLKCK31
Start Acknowledge Message (STACK) Format

I ENQ , STCKTYPE I STCKSUB , FLAGS' FILL' FILL' ADDR , BLKCK31
ENQ
ACKTYPE
NAKTYPE
REPTYPE
STRTTYPE
STCKTYPE
ACKSUB
REASON
REPSUB
STRTSUB
STCKSUB
FLAGS
RESP
FILL
ADDR
BLKCK3

the control message identifier
the ACK message type with a value of1
the NAK message type with a value of2
the REP message type with a value of 3
the STRT message type with a value of 6
the STACK message type with a value of 7
the ACK subtype with a value of 0
the N AK error reason
the REP subtype with a value of 0
the STRT subtype with a value of 0
the STACK subtype with a value of 0
the link flags
the response number used to acknowledge
received messages that checked out to be correct
a fill byte with a value of 0
the tributary address field
the control message block check

Figure 2.3
DDCMP Control Message Formats

34· DECnet Concept

. Maintenance Messages
Figure 2.4 illustrates the fonnat of the maintenance message,
which is used in DDCMP maintenance mode. The maintenance
message is a DDCMP envelope for data controlling downline
loading, upline dumping, link testing, and controlling an unattended computer system. The DNA protocol used in performing
maintenance functions is the Maintenance Operation Protocol
(MOP). MOP messages (see Chapter 10 for a description) are sent
within the DnCMP maintenace message.
DDCMP Operation
The DDCMP module has three functional components: framing,
link management, and message exchange.
Framing- The framing component locates the beginning and end
of a message received from a transmitting nDCMP module. Framing involves locating and locking onto, or synchronizing with, a

Maintenance Message Format
IDLEI COUNTI FLAGS I FILL I FILL I ADDRIBLKCK11DATAIBLCKCK21
8
14
2
8
8
8
16
8n
16
DLE
COUNT
FLAGS
FILL
ADDR
BLCKCKl
DATA
BLKCK2

the maintenance message identifier
the byte count field
the link flags
a fill byte wi th a value of 0
the tributary address field
the header block check on fields D LE
throughADDR
the n-byte data field, whereO<n = COUNT<2 14
the block check on the DATA field

Figure 2.4
DDCMP Maintenance Message Format

certain bit, byte, or message and then receiving subsequent bits,
bytes, or messages at th~ same rate as that at which they come in.
At the Physical Link level, modems and communications interfaces synchronize bits. The DDCMP framing component synchronizes bytes by locating a certain 8-bit window in the bit stream.
On asynchronous links, DDCMP uses start/ stop transmission techniques to synchronize bytes. On synchronous links, DDCMP
searches for a SYN character. Byte synchronization is inherent in 8bit multiple parallel links. DDCMP synchronizes messages by
searching for one of the three special starting bytes (SOH, for data
messages, ENQ for control messages, and DLE for maintenance
messages) after achieving byte synchronization. To maintain
message synchronization, DDCMP counts out fixed-length headers
and, when required, counts out variable-length data based on the
count field of the header.
Link Management-This component controls transmISSIOn and
reception on links connected to two or more transmitting systems
and/ or receiving systems in a given direction. Link management
controls the direction of data flow on half-duplex links and, with
the use oflink flags, controls the selection of tributary stations on
multipoint links. In addition, link management uses selective
addressing to control the receipt of data on multipoint links.
Message Exchange-This component transfers data correctly and in
sequence over a link. Message exchange operates at the message
level (after framing has been accomplished) to exchange both data
and control messages.

Typical Message Exchange
is a positive-acknowledgement-with-retransmission protocol.
This means that, for each data message correctly received and
passed to the Routing layer, DDCMP returns a positive acknowl-

DDCMP

36· DECnet Concept

edgement. Such an acknowledgement is either an Acknowledge
Message (ACK) or a piggybacked acknowledgement in the
response (RESP) field of a data message.
If a DDCMP module receives a message out of sequence or with an
error detected by the CRC, DDCMP does not pass the data on to the
Routing layer. Typically, DDCMP does not acknowledge this
message. Eventually, a time-out occurs, and one of two things
could happen: the data sending DDCMP module retransmits the
message or the data receiving DDCMP module sends a NAK to the
data sending module.
The transmission ofDDcMP data messages includes the following
steps:
1. The transmitter increments the message number and puts the
number, n, in the data message. This message is transmitted
within the required framing envelope. A timer is started.

2. The receiver frames and receives the message, checks the
received CRC value against a computer CRC value, and compares the
message number with the number it expects to receive. If the
message checks out correctly, the receiver returns a positive
acknowledgement (ACK) with that number, passes the message to
the Routing layer, and increments the next expected number to
n + 1. If the message fails to check out, the receiver ignores the
message or sends a NAK.
3. The transmitter then follows one of three procedures:
If the transmitter receives a positive acknowledgement, it checks
the number received to see ifit is an acknowledgement of receipt of
an outstanding message (a message that has been awaiting transmission). If this is the case, the transmitting DDCMP module
notifies the Routing layer of successful transmission of that

message as well as of any previous lower-numbered outstanding
messages. If all outstanding messages have been acknowledged,
the timer is stopped. If one or more message remains outstanding,
the timer is restarted.
If the transmitter receives nothing, the timer expires. The transmitter sends a REP control message to initiate error recovery.
If the transmitter receives a negative acknowledgement, it retransmits the message and all higher-numbered messages.
The transmitter can send several data messages before requiring
that the first one be acknowledged. Acknowledgement of the
highest-numbered message implies acknowledgement of all
lower-numbered messages. A negative acknowledgement implies
postive acknowledgement of any previously transmitted, lowernumbered messages.
Figure 2.5 shows a message exchange involving positive acknowledgement and pipelining. Figure 2.6 shows error recovery from a
NAK.

Maintenance Mode
Maintenance mode uses the DDCMP framing and link management
components but not the message exchange component. Sequencing or acknowledgement, if required, must be handled within the
data fields of the maintenance messages and is part of a higher-level
protocol.
Ethernet Functional Description
Computer systems and other digital devices on an Ethernet local
area network can exchange data at a speed of up to 10 Mbits per
second. An Ethernet local area network can be located in one
building or can span a cluster of buildings. The incorporation of
the Ethernet specification in Phase IV DNA means that systems
running DEC net Phase IV can be connected directly to an Ethernet

Node A

Messages

Client A requests
transmit of first data
message

ACK received

DA TA (message number n)

Client A requests
transmit of second
data message

Message received
and client A requests
transmit. Acknowledgment of client
message B is piggybacked in RESP field
of data message

NodeB

2 Message received

ACK (oj message number n)

and given to client B;
client B now expects
next data message to
be n+l

DATA (messaJ?e number n + 1) 5 Message received

DATA (message number m)

and client B requests
transmit of its message. Acknowledgment of client A
message is piggybacked in RESP field
of data message
These 3 data
messages are
pipclined

DA TA (message number n + 2) 7 Message received

Client A requests
transmit

~ Message received

Client A requests
transmit

~ Message received;

Message received

DATA (message number n+3) 8

DATA (messagenumbern+4) 9

10 DATA (message number m +1)

Figure 2.5
Typical Message Exchange, with Positive Acknowledgement, Piggybacking,
and Pipe lining

client B requests
transmit. Messages
n+2, n+3 and n+4
are all acknowledged
with the acknowledgment of n+4 in
RESP field of client B
data message

Node A
Client A requests
transmit

~Message receIve
'd'
Data (messaJ:e number n)

N AK received
Retransmit

NodeB

Messages

error

NAK (of messaJ:e number n)

~ 4 Message received
Data (messaJ:e number n)

Figure 2.6
DDCMP Message Exchange, showing
Error Recovery

coaxial cable to perform high-level network functions (file transfer
and remote resource access, for example) at the high speed
supported by the Ethernet cable.
The Ethernet local area network specification is the result of an
extensive collaborative effort by Digital Equipment Corporation,
Intel Corporation, and Xerox Corporation. The specification
includes a Physical layer and a Data Link layer.
The primary characteristics of the Physical layer are: a data rate of
10 million bits per second, a maximum station (node) separation of
2.8 kilometers, a maximum of 1,024 nodes or stations on a single
Etherent LAN, a shielded coaxial-cable medium using Manchesterencoded digital baseband signaling, and support of a bus structure
in the shape of a branching tree.
The characteristics of the Data Link layer include multiaccess
network control, in which access to the channel is fairly distributed to all nodes, and regulation of channel access through the
Carrier Sense Multiple Access with Collision Detect (CSMAICD)
technique.

40 • DECnet Concept

The functions of the Ethernet Data Link include data
encapsulation/ decapsulation and link management:
Data Encapsulation/Decapsulation comprises framing, addressing,
and error detection. Framing defines the format of message
packets (the different fields of information within packets) that are
broadcast over the local area network. In addition, framing
constructs packets from data supplied by the nodes through the
higher layers, disassembles network messages, and supplies data to
the higher layer protocols of the node. Addressing handles source
and destination addresses, and error detection detects physical
channel transmission errors.
Link Management comprises channel allocation and channel access.
Channel allocation is responsible for amount of channel use,
which is determined by the definition of packet size. Channel
access is controlled by the CSMA/ CD technique, part of which is
carried out in each of the two layers. The Data Link layer responds
to the channel or carrier sensing of the Physical layer by deferring
transmission in case of traffic, transmitting in the absence of
traffic, and backing off and resending in the case of a collision.

Figure 2.7 shows the two functional components of the Data Link
layer, the Data Encapsulation sublayer, and the Link Management
sublayer.
The Ethernet Data Link layer provides a best-effort delivery
service. It does not provide an error-control facility to recover from
transmission errors. In DNA, the End-to-End Communication
layer is responsible for the error-control functions that are essential
for reliable communications. Ethernet local area networks do not
require additional error control at the data link level since the
Ethernet physical channel has an inherently low error rate.

Client Layer
Client-to-Data Link Interface

Data Link-to-Physical Interface
Physical Layer

Figure 2.7
Ethernet Data Link Layer Functions

Ethernet Messages
The Ethernet Data Link layer has one type of message, the frame.
The data encapsulation function of the Data Link layer comprises
the construction and processing of frames. The frame format (see
Figure 2.8) reflects the data encapsulation functions of framing,
addressing, and error detection in the following manner:
Framing: no explicit framing information is contained in the
format since the necessary framing cues are present in the interface
to the Physical layer.
Addressing: there are two address fields to identify the source and
destination nodes of the frame. Both address fields contain 48 bits.
In DNA Phase IV, each network node has a 16-bit node address. The
48-bit Ethernet data link address of a DNA Phase IV node is derived

Ethernet Data Link Frame Format

IDESTINATION ISOURCE ITYPE IDATA IFCS I
48
DESTINATION =

SOURCE

TYPE

DATA

FCS

the destination data link address. A data link address
is one of three types:
• Physical Address: The unique address associated with
a particular station on the Ethernet. The 48 bit field
permits a station to have a unique address over all
Ethernets. In DNA Phase IV, each network node has
a 16 bit node address. The 48 bit Ethernet data link
address of a DNA Phase IV node is derived by
prefixing the 16 bit address with a 32 bit prefix
assigned to DNA Phase IV nodes. Thus, DNA
Phase IV addresses are unique over a single DNA
network, which may include multiple Ethernets,
DDCMP links, and X.25 links.
• Multicast Address: A multi-destination address associated by higher level convention with a group of
logically related stations on an Ethernet. DNA assigns multicast addresses to the group of all Ethernet
End nodes and to the group of all Ethernet Routers.
• Broadcast Address: A distinguished, predefined
address which denotes the set of all stations on an
Ethernet.
the source data link address. This field always contains the physical address of the station transmitting
a frame on the Ethernet.
the type field. The type field is reserved for use by
higher level protocols to identify the higher level
protocol associated with the frame, permitting multiple higher-level protocols to coexist in the same
Ethernet. Ethernet type field values are assigned to
the DNA Phase IV Routing protocol and to the
DNA Maintenance Operation protocols.
the data field. The Ethernet data field contains higher
level protocol data and is 8n bits long, where
46:5 n :51500. Full transparency is provided, in the
sense that any arbitrary sequence of8 bit bytes may
appear in the data field. The minimum length of the
data field ensures that all frames occupy the channel
long enough for reliable collision detection.
the frame check sequence. This field contains the
CRC-32 polynomial check on the rest of the frame.

Figure 2.8
Ethernet Data Link Frame Format

by attaching to the 16-bit address a 32-bit prefix assigned to DNA
Phase IV nodes. DNA Phase IV addresses are unique over a single
DNA network that can include multiple Ethernet networks,
DDCMP links, and x.2S links.
Error Detection: a Frame Check Sequence field is used to check
the accuracy of the information contained in each transmitted
packet.

In Figure 2.8, the numbers below each field indicate the length of
the field in bits.
DNA defines a standard convention for padding higher-level proto-

cols to the minimum Ethernet data field size. Higher-level protocol data is preceded by a 16-bit byte count and followed by any
required pad bytes. When this convention is in effect, Ethernet
frames have the format shown in Figure 2.9

Ethernet Frame Format with Padding

I DESTINATION I SOURCE I TYPE I COUNT I DATA I PAD I FCS I
48
COUNT
DATA

°
PAD

the 16 bit count of DATA bytes, where COUNT = n
the data field. The Ethernet data field contains
higher level protocol data and is 8n bits long, where
::S n ::S 1498. Full transparency is provided, in the
sense that any arbitrary sequence of8 bit bytes may
appear in the data field.
zero or more bytes of zeros, where m = max(O,44 - n).

Figure 2.9
Ethernet Frame Format with Padding

44 • DECnet Concept

Ethernet Operation
The higher layers of a node (layers above the Data Link layer)
initiate transmission requests. These layers are referred to as the
client layers, since the Data Link layer provides them with transmission and reception services.
Transmission Without Contention The client layer requesting transmission passes data, source and destination addresses, and data
format type to the Transmit Data Encapsulation component of
the Data Link layer. This component constructs a frame from
the client-supplied information, with or without padding
(depending on what the client has specified). The Data Link layer
then appends a frame-check sequence for error detection and hands
the frame to the Transmit Link Management component for
transmission.

Before proceeding to transmit, Transmit Link Management
attempts to avoid contention with other traffic on the channel by
monitoring the carrier-sense signal of the Physical layer and by
deferring to any passing traffic. The Physical layer is able to detect
carrier-in this case, the electrical baseband signal that is present on
the coaxial cable-whenever there is a frame being transmitted.
When the channel is clear, the Data Link layer passes the frame to
be transmitted to the Physical layer as a serial stream of bits. Before
transmitting any of the bits in the frame, the Physical layer sends an
encoded preamble that allows the receivers of all the nodes on the
channel to synchronize their clocks. The Physical layer then begins
translating the bits of the frame into Manchester-phase-encoded
form and generates electrical signals onto the coaxial cable that
represent the bits in the frame.

The Physical layer of a transmitting node monitors the channel
during the entire time it is transmitting (it is the energy level or the
electrical-signal level of the channel that the Physical layer monitors). If there is a collision, the electrical-signal level increases. The
Physical layer detects this increase. In a contention-free transmission, the collision-detect signal remains off. When the transmission
has been completed without contention, the Data Link layer
informs the client layer and awaits the next frame for transmission.
Reception Without Contention When a node begins to transmit in
the absence of traffic on the channel, the Physical layers of all other
nodes on the network sense its carrier signal and alert their
respective Data Link layers so that there are no conflicting
attempts to transmit. At that point, all other nodes on the
network become receiving nodes and listen to the transmitted
signal. The Physical layer of each receiving node synchronizes to
the incoming preamble, receives the encoded bits from the cable,
and translates the phase-encoded signal back into binary data,
discarding the preamble.

Meanwhile, the Receive Link Management component of the
Data Link layer has been waiting for the incoming bits to be
delivered, since it was notified of a carrier signal. Receive Link
Management collects bits from the Physical layer as long as the
carrier-sense signal remains on. When the carrier-sense signal goes
off, the frame is passed to the Receive Data Decapsulation
component for processing.
The receiving node's Receive Data Decapsulation component
checks the frame's destination address field to determine whether
the frame is intended for that node. Ifit is, its type field is checked
to decide which client module (Routing or Network Manage-

46 • DECnet Concept

ment, for example) should process the incoming frame. If the
transmitting client module has specified that padding is in use, the
frame is decapsulated appropriately. The contents of the frame are
then passed to the receiving client layer with an appropriate status
code. The status code is generated by inspecting the frame check
sequence to detect any damage to the frame and by checking for
proper byte-boundary alignment at the end of the frame. If the
destination address portion of the frame indicates that it is not
intended for that node, the bit stream to the Data Link layer is
halted after the preamble and destination address, and the frame is
not accepted.

Collisions: Handling of Contention- It is possible that two or more
nodes could determine that the channel is idle and attempt to
transmit at or about the same instant. When this happens, their
transmissions overlap and interfere with one another. The resulting contention is called a collision. A node can experience a collision
only during the initial part of its transmission (the collision
window), before its transmitted signal has had time to propagate to
all parts of the Ethernet channel. Once the window has passed, the
node or station is said to have acquired the channel, since all properly
functioning stations would have detected the carrier signal by that
time and would have deferred transmission.

In the event of a collision, the Physical layer first notices the
interference on the channel and turns on the collision-detect
signal. The Transmit Link Management of the Data Link layer
notices this signal and initiates collision-handling procedures.
First, the frame being transmitted is permitted to continue for a
brief interval (called ajam), to ensure that the collision is noticed by
other transmitting nodes involved in the collision. Then Transmit
Link Management terminates the transmission, or backs off, and

schedules a retransmission attempt for a randomly selected time in
the near future. When repeated collisions occur, retransmission
attempts are also repeated, accordingly. Since repeated collisions
indicate a busy channel, Transmit Link Management attempts to
adjust to the channel load by voluntarily delaying its own retransmissions to reduce its load on the channel. Eventually, either the
transmission is successful, or the attempt is abandoned with the
assumption that the channel has failed or has become overloaded.
At all receiving nodes, the bits from a collision are received and
decoded by the Physical layer as if they were bits from a valid frame
(this happens because the collision detect signal is not acknowledged in receiving nodes). The Data Link layer's Receive
Link Management component can distinguish between a collision
fragment and a valid frame, since collision fragments are always
shorter than the shortest valid frame. Receive Link Management
does not accept collision fragments .

. Routing
Routing is the DNA layer whose function it is to determine the path
or route along which a packet of data travels to its destination. A
path consists of the sequence of circuits and nodes between a
source node and a destination node. Users at a node need only
specify the names of remote nodes with which they wish to
communicate. Routing takes care of the details, ensuring that data
travels to its appropriate destination. The path the data takes
through the network is entirely transparent to users. In many
cases, there are multiple paths between network nodes (see Figure
2.10), some more efficient than others.
In area-based networks, data travels within and between areas.
Regardless of the number of areas between source and destination
nodes, routing nodes forward a packet of data to its intended

48 • DECnet Concept

NodeD

NodeE

NodeF

Node A

Figure 2.10
Routing Paths

location. Router is a general term for any node in a network that
can receive and forward messages or packets addressed to other
nodes in the network.
There are two types of routers in area-based networks: Level 1
routers, which are responsible for communication within an area,
and Level 2 routers, which are responsible for communication
between areas.
A Levell router operates in the same way as traditional Phase III
and Phase IV DEcnet routing nodes. It handles routing within an
area and keeps itself informed of the state of all nodes in that area,

but it does not concern itself about nodes outside its area. When a
node in one area (area 5, for example) wishes to communicate with
a node in another area (area 7, for example), a Levell router in area
5 forwards the request to the nearest Level 2 router in the same
area.
All inter-area communication takes place through Level 2 routers.
Each Level 2 router keeps track of the least-cost path to each area in
the network, as well as the state of the nodes within its area.
The Routing layer delivers packets on a best-effort basis. What
this means is that the layer provides no guarantees against packets
being lost, duplicated, or delivered out of order. A higher layer of
DNA, the End-to-End Communication layer, provides these assurances. The Routing layer selects routes on the basis of network
topology and operator-assigned circuit costs (see below, under
Routing Terms) and adapts to changes in network topology (like
finding an alternate path if a circuit or node fails).

Routing Terms
The total distance between a source node and a destination node is
called the path length. The path length is measured in hops. A hop is
the logical distance between two adjacent nodes. A path never
exceeds a maximum number of hops; this number is a value that a
network manager sets or one that is determined by the DEC net
implementation (Phase III or Phase IV).
Routing allows the network manager to assign a cost to each
circuit connected to a node. Cost is an arbitrary integer (within the
size limits of the cost field) that is used to determine the best path
for a packet. The cost of each path between a source node and a
destination node is the sum of positive integer values assigned to
the circuits that compose the path.

50· DECnet Concept

The Routing layer routes packets on the path of least cost, even if
this is not the path with the fewest hops. The DNA Routing
Functional Specification does not dictate how to assign circuit
costs, but it does suggest a cost-assignment algorithm based on
circuit bandwidth (since packets take the path with the lowest
cost, a network manager could assign low-cost figures for circuits
with high bandwidth and high-cost figures for circuits with low
bandwidth).
The design of the Routing layer gives network managers great
flexibility in configuring networks. Node locations and the
assignment of circuit costs and maximum-hop values are left
entirely to a network manager's discretion. Thus, a DEC net
network can be tailored to meet the varying needs of differing
application requirements.
Figure 2.11 illustrates some Routing terms. The glossary contains
definitions of these and other Routing terms (including network
diameter, maximum path length, maximum visits, maximum
path cost, and maximum cost).
Routing parameters like circuit costs and maximum path lengths
are defined by network managers when installing a node into a
DEC net network. These values can also be modified after a DEC net
system has been installed to improve network performance or to
reflect changes in network configuration. The values of routing
parameters should be determined only after careful consideration
of their effects on the local node and on the network as a whole.
When a node is up and running, an operator can dynamically
increase or decrease the cost for each circuit defined for that node.
Altering circuit costs can result in alteration of packet-routing
paths. The ability to change circuit costs dynamically can be an

NodeD

NodeE

\::

3
I'

NodeC
v

NodeB

NodeF
/

1 hop { 2
/

Node A
/

Node A wants to send a packet to Node D. There are three possible paths.
Path
A to B

B to C

A toB

B taD

A to B

B to F

C to D

F to E

E to D

Path Cost

Path Length

2 + 2 + 3 = 7*

3 hops

2+7=9

2 hops

2+3+4+2=11

4 hops

* 7 is the lowest path cost; Node A therefore routes the packet to Node D
via this path.

Figure 2.11
Routing Terms

52· DECnet Concept

advantage when network performance suffers due to excessively
high levels of traffic on low-cost circuits. (The Routing module
itself does not automatically adjust to traffic flow, since packets
always travel on the path with the lowest cost.)

Routing Algorithms and Databases
Routing algorithms operate on the routing parameters that
system managers assign for every node in the network. One
Routing algorithm calculates the path length and path cost to
every possible destination via every circuit defined for that node.
Another algorithm determines which circuit represents the least
costly path to each destination in the network. The algorithms
operate on locally available values (circuit costs and path lengths,
for example) as well as on values supplied by all adjacent nodes.
Each node maintains routing databases in which are stored the
results of the algorithms. In an area-based network, the Level 1
routers maintain routing databases that are similar to the routing
databases in traditional Phase III and Phase IV DEC net nodes. The
routing database of a Level 1 router includes information on path
length and path cost to every node and Level 2 router in its area. A
Level 2 router makes use of a second database-an area routing
table-to determine the least-cost path to the other areas in the
network.
Upon receiving a packet to be transmitted or forwarded, the
Routing layer first consults the databases to find the least costly
path to the packet's destination and then sends the packet via the
selected circui t. A packet is discarded or returned to the sender if i ts
destination is not accessible via any circuit known to the Routing
module.
When the packet arrives at the next adjacent node in the path, the
Routing module there also chooses the least costly path for the

packet, according to the local node's routing databases. The
Routing module performs the services for packets being routed
through the local node, as it does for packets that are originating at
the local node.
Each Routing module executes the algorithms over and over again
in response to events of different kinds on the network. If a
physical line somewhere in the network goes down, the Routing
algorithms recalculate the path length and cost to destinations
affected by the line failure.
Whenever the algorithms reexecute, each node reveals the contents of its databases to all adjacent nodes, which use that
information to update their own databases. In this way, changes
affecting path lengths and costs ripple back and forth through the
network so that all the routing databases are automatically
updated to reflect the current state of the network.
In an area-based network, when a Levell router receives a packet
addressed to a node in another area, it uses Levell routing to send
the packet to the nearest Level 2 router. That router forwards the
packet through Level 2 routers to a Level 2 router in the destination
area. The destination Level 2 router then sends the packet through
Levell routers in the destination area to the destination node.
Routing Features
In addition to determining packet paths, path length, and path
cost, the routing mechanism performs other functions that ensure
efficient data transfer over the network. These include:

Fonvarding packets-If a packet is addressed to the local node,
Routing delivers it to the End-to-End Communication layer. If a
packet is addressed to a remote node, Routing forwards it to the

54 • DECnet Concept

adjacent node in the path. In large area-based networks, Levell
routing forwards packets between nodes within an area; Level 2
routing forwards packets between areas.
Adapting to topology changes-If a circuit or node in a path fails,
Routing finds an available alternative path.
Adapting to difforent kinds oj circuits-Routing operates over paths
consisting of multiple types of Data Link circuits, including
DDCMP point-to-point and multipoint lines, Ethernet links, and
x.25 virtual circuits.
Periodically updating Routing modules on other nodes-Routing modules in other nodes are periodically updated to reflect any topology
changes (circuits going down or up and operators modifying
routing parameters, for example. See above).
Returning packets addressed to unreachable nodes-Routing returns
packets addressed to unreachable nodes to the End-to-End Communication layer, thereby keeping it informed of the status of
packet transmissions.
Limiting the number ofpackets queuing up for transmission on individual
circuits-This prevents a circuit from becoming overloaded.
Managing buffirs-Routing manages the buffers at nodes that are
capable of routing packets from a remote source to a remote
destination.
Regulating the ratio ofpackets to be forwarded through a node with those
that are generated on that node-This achieves a balance between
incoming and outgoing traffic.

Tracking the number of nodes a route-through packet has visited and
discarding packets that have exceeded a predefined lim it- This ensures
that packets can never loop endlessly through the network tying
up network resources.
Performing node veriflcation-Routing will exchange verification
passwords with an adjacent node if so requested by the Network
Managemen t layer.
Maintaining counters and gathering event data Jar network management
purposes- This enables a network manager to identify and locate
network problems such as traffic congestion.

See below, under Routing Operation, for details on these features.
How Routing Works on an Ethernet LAN
When a message is sent to an Ethernet node from a Phase III or
Phase IV node that is not on the Ethernet or from a node on a
different Ethernet network, a special routing node on the
Ethernet-the DEe net Router Server (see Chapter 3)-or a host
system routing node receives the message. Either the DEenet
Router Server or the host system routing node submits the
message to the Ethernet channel. (A DEenet Routerlx.25 Gateway can perform this function, as well). The message travels along
the Ethernet channel until it reaches the destination node (the
destination node can recognize and retrieve messages addressed to
it). If a message originates on an Ethernet and is addressed to
another node on the same Ethernet, a routing node is not
necessary. The source node transmits the message directly to the
destination node as if there were a point-to-point connection
between the nodes.

56 • DECnet Concept

When the x.25 protocol is implemented through a gateway node
(a VAX node or a DEC net Router/x.25 Gateway) rather than
through data link mapping (discussed in Chapter 8), DEcnet
routing is not used over the PSDN. The PSDN handles all routing
functions that occur within its boundaries. (See Chapter 8 for an
explanation of x.25 implementations.)
DEcnet supports area routing on an Ethernet channel. Groups of
Ethernet systems may thus be segregated into separate areas. The
routing overhead resulting from such a configuration does not
make it a cost-effective solution, however. Nodes that would
normally communicate directly with one another over the
Ethernet cable now have to communicate through intervening
Levell and Level 2 routers.
Routing Messages

There are two types of Routing messages: data packets and control
messages. Data packets carry data to and from the End-to-End
Communication layer. Routing adds a packet route header to
messages sent by the End-to-End Communication layer. Routing
modules in adjacent nodes exchange control messages to initialize
the Routing layer, maintain routing data, and monitor the status
of adjacent nodes.
Packet Route Header-The Routing layer ofa node uses one of two
formats for data packet routing headers, depending on whether
the adjacent node is on an Ethernet or on some other type of
circuit. When the adjacent node is an Ethernet end node (an
Ethernet node that does not have the ability to receive and forward
messages intended for other nodes), the Routing layer uses the
Ethernet Endnode Data Packet format. For any other type ot

Phase IV Data Packet Routing Header Format

I RTFLG I DSTNODE I SRCNODE I FORWARD I
8

Ethernet Endnode Data Packet Routing Header Format

I RTFLG I DSTNODE I SRCNODE I RES I FORWARD
8
RTFLG

DSTNODE
SRCNODE
FORWARD
RES

RES
16

the set of flags used by the routing nodes, including:
• Return to sender flag (indicates whether or not the
packet is being returned)
• Return to sender request flag (indicates whether to
discard or try to return packet)
• Intra-Ethernet Packet (indicates to the Ethernet
Endnode receiving this packet that the source of the
packet is on the same Ethernet, and can be
communicated with directly).
the destination node address
the source node address
the number of nodes this packet can visit
a reserved field

Figure 2.12
Packet Route Header Formats

adjacent node, the Routing layer uses the Phase IV Data Packet
format. The Ethernet Endnode Data Packet format has expanded
addressing and reserved fields, currently not used in DEC net Phase
IV, for later expansion.
Figure 2.12 shows the data packet routing header formats. The
numbers below the fields in this and in the next figure indicate the
length of the fields in bits.

58 • DECnet Concept

Routing Control Messages-A node's Routing layer uses six types of
Routing control messages. It uses one type, the Routing message,
on all types of circuits. It uses two types of Routing control
messages-the Ethernet Endnode Hello message and the Ethernet
Router Hello message-for Ethernet circuits only. (See above, under
Routing, for definition of Router). And it uses the three other
types of Routing control messages-the Hello and Test message, the
Initialization message, and the Verification message-only on nonEthernet circuits.

The Routing message provides the information necessary for updating the routing database of an adjacent node. This message carries
path cost and path-length values for a set of destinations.
The Ethernet Router Hello message is used for the initialization and
periodic monitoring of Routers on an Ethernet circuit. Through a
multicast operation, each Etherent router periodically broadcasts
an Ethernet Hello message to all other nodes (routers and end
nodes) on the same Ethernet circuit. The message contains a list of
all routers on the Ethernet circuit from which the sending router
has recently received Ethernet Router Hello messages. By
exchanging Ethernet Router Hello messages, all routers remain
informed of the status of the other routers on the circuit. The
message also provides input to an algorithm that selects a single
router on the Ethernet circuit to be the designated router for that
Ethernet. The designated router assists two end nodes in discovering that they are both on the same Ethernet. It does this by setting
a special intra-Ethernet bit in a packet that it forwards from one
end node to the other. Subsequent communications between the
end nodes can be direct, without router intervention.
The Ethernet Endnode Hello message is used for the initialization and
periodic monitoring of end nodes on an Ethernet circuit. By

means of a multicast operation, each Ethernet end node periodically broadcasts an Ethernet Endnode Hello message to all routers
on the Ethernet circuit. The routers use this message to maintain
the status (up or down) of end nodes on the Ethernet.
The Hello and Test message tests an adjacent node to determine if an
adjacency is operational (an adjacency is a circuit and node pair).
The Routing layer of a node sends this message periodically on
non-Ethernet circuits in the absence of other traffic. When a
node's Routing layer receives this message or any other valid
message from an adjacent node, the Routing layer starts or restarts
a timer. If the timer expires before the Routing layer receives
another message from the adjacent node, Routing considers the
adjacent node down.
The Routing layer of a node sends an Initialization message when
initializing a non-Ethernet circuit. The message contains information about the node type, required verification, maximum Data
Link layer receive block size, and Routing version.
A node's Routing layer sends a Verification message if, on a nonEthernet circuit, the initialization message requires that node
verification accompany the initializing process.
Figure 2.13 illustrates the Routing control message formats.
Routing Operation
The Routing layer of a node consists of two sublayers with
associated functional components. The Routing Control Sublayer
performs routing, congestion control, and packet-lifetime-control
functions. The Routing Initialization Sublayer performs the initialization function.

Routing Message Format

ICTLFLG ISCRNODE IRTGINFO ICHECKSUM I
16

16n

Ethernet Router Hello Message Format

ICTLFLG IVERS IID IINFO ILIST I
8

56m

Ethernet Endnode Hello Message Format

ICTLFLG IVERS IID IINFO I
8

168

Hello and lest Message Format

ICTLFLG ISRCNODE ITEST DATA I
16

8-128

Initialization Message Format

ICTLFLG I SRCNODE I INITFO I
16

Verification Message Format

I CTLFLG ISRCNODE IFCNVAL I
16

CTLFLG

SRCNODE
RTGINFO
CHECKSUM
VERS
ID
INFO
LIST

TEST DATA
INITFO
FCNVAL

8-64
Routing control flag, with the following types:
• Initialization message
• Verification message
• Hello and Test message
• Routing message
• Ethernet Router Hello message
• Ethernet Endnode Hello message
Identification of source node's Routing Module
Path length and path cost to a set of n destinations
One's complement add check on routing information
Routing module version number
Identification of node sending message
Node type (Router or Endnode), maximum Data Link
layer receive block size, hello timer
list of known Routers on the Ethernet circuit (Each of
the m entries consists of a 7 bit priority field for selecting the Designated Router, 1 bit to indicate two-way
connectivity, and the 48 bit identification of a Router)
sequence of up to 128 bytes of data to test the circuit
node type, required Verification message, maximum
Data Link layer' receive block size, Routing version.
type-dependent verification information; function
value.

Figure 2.13
Routing Control Message Formats

Routing- The routing function is made up of five processes:
decision, update, forwarding, and receive.
The decision process selects routes to each destination node in the
network. This process uses a connectivity algorithm that maintains path lengths and a traffic-"assignment algorithm that maintains path costs. When a routing node receives a Routing message,
the routing node executes the connectivity and traffic-assignment
algorithms. This execution results in updates to the databases that
the routing node uses to determine packet routes.
The update process constructs and propagates Routing messages.
It sends Routing messages to adjacent nodes whenever the
decision process requires it to do so, and it also periodically sends
Routing messages to adjacent nodes to ensure the integrity of the
routing databases.
The forwarding process supplies and manages the buffers required
to support packet route-through to all destinations. It looks up a
table to select the output circuit for a packet. If a destination is
unreachable, the forwarding process either returns the packet to
the sender or discards it, depending on the option flagged in the
packet route header.
The receive process inspects a packet's route header. If the packet is
intended for the local node, the receive process dispatches it to the
local End-to-End Communication layer. If the packet is intended
for another destination, the receive process dispatches it to the
Routing layer of the next node in the packet's path.

Congestion Control-This function consists of a single process:
transmit management. This process manages buffers by limiting the
maximum number of packets on a queue for a circuit. If a queue for

62· DECnet Concept

a particular circuit reaches a predetermined threshold, additional
packets for that queue are discarded to prevent congestion.
Transmit management also regulates the ratio of packets received
directly from the End-to-End Communication layer to routethrough packets. This regulation prevents locally generated packets from degrading a node's route-through service.

Packet Lifetime Control-This function prevents excessive looping
by discarding packets that have visited too many nodes.
Initialization and Circuit Monitor-By means of the Routing initialization function, which is both a startup and periodic procedure,
the Routing layer of a node ascertains the identities of neighboring
nodes and adapts to the characteristics of the circuits to these
neighboring nodes.
On an Ethernet circuit, the Ethernet Endnode Hello message and
the Ethernet Router Hello message perform initialization and
monitoring functions (see above, under specific Routing message
types for a description).
On DDCMP circuits (point-to-point and multipoint), the Initialization message, the Verification message, and the Hello and Test
message perform initialization and monitoring. In addition to the
functions these messages perform, described above, the Initialization message adapts the Routing layer's operation to the block size
of the physical line that the Data Link layer manages.
On x.25 virtual circuits (permanent and switched), the Initialization and Verification messages work in the same manner that they
do on DDCMP circuits. The difference is that on x.25 circuits, the
Routing layer uses a Data Link block size that is taken from the
x. 25 data packet size selected when the virtual circuit is initialized.

x.2S circuits are initialized when the Network Management layer
commands the Routing layer to do so. The Routing Initialization
Sub layer contains a database, maintained by Network Management, that includes a mapping between Routing layer circuits and
x.2S packet-level permanent virtual circuit identifiers (in the case
of permanent virtual circuits), and between Routing layer circuits
and virtual call parameters (in the case of switched virtual circuits).
The database contains items such as DTE subaddress range, maximum packet size, and the number of times failed calls should be
reattempted.
When placing a virtual call, the Routing Initialization Sub layer
provides the necessary parameters to the x.2S packet-level component of the Data Link layer, which places the call to the remote
Routing layer at the supplied foreign DTE address. If the remote
DTE accepts the call, the Routing Initialization function proceeds
with circuit initialization.
The Routing Initialization module listens for incoming virtual
calls managed by the x.2S packet-level component of the Data
Link layer. When informed of an incoming call by the x.2S packetlevel component, Routing Initialization executes an algorithm to
determine whether to accept the call, reject it, or leave it pending
in case some other DNA module (such as the x.2S Gateway Server
module-see Chapter 8) wishes to process the call. The algorithm
uses the mapping database described above and makes its decision
by comparing the foreign DTE subaddress in the incoming call
packet to the DTE subaddress range in its database. If the decision is
to accept the call, initialization proceeds as described above.
Once an x.2S circuit is established, the Routing Initialization
Sublayer performs additional functions not performed for

64 • DECnet Concept

Ethernet or DDCMP circuits. These functions, which maximize the
performance and integrity of x.2S circuits, include:
Blocking and Deblocking Routing Layer Messages-Since communication over x.2S circuits is in the form of packets, Routing Initialization will block and deblock DNA datagrams into x.2S packets in
order to minimize the number of packets transmitted over the
virtual circuit.
Fragmenting and Reassembling Routing Layer Messages-If the x.2S
virtual circuit packet size is too small to contain an entire Routing
layer message, Routing Initialization will fragment and reassemble Routing layer messages so that they fit in the available x.2S
packet size.
Checksums of X. 25 Packets-To insure the integrity of transmitted
data, the Routing Initialization Sublayer appends a checksum to
every data packet transmitted over the virtual circuit.

As it does for other circuits, the Routing Initialization Sublayer
monitors x.2S circuits for non-recoverable errors, including x.2S
resets, restarts, and clears. If any of these events occurs, Routing
Initialization declares the circuit down, informs the Routing
Control Sublayer of the condition, and attempts to reinitialize the
circuit.
Figure 2.14 summarizes the operation of the Routing layer.
. Data Flow
The primary purpose of a network is to pass data from a source
node to a destination node. Data traveling from one node in a
network to another node passes from a source process in the User

Transmit Interface
to End-to-End
Communication

Receive Interface
to End-to-End
Communication
End-to-End Communication Layer
Routing Layer

Forwardin<~

---

Database

~--

Contains:
• reachability
vector
• output
circuit
to be used

Forwarding
Process

Decision Process

Routil1g
Database

· selects paths
• maintains Routing
Database
· maintains Forwarding Database

Contains:
• Routin,R
Information
(hop, cost,
flags, etc.)

t
Receive Process

Update Process

• receives packets
. passes them to appropriate process
. discards packets
that have visited
too many nodes
(Loop Detection)

• generates and
propagates
routing
messages

I-----------------~I----------------

· supplies and manages buffers for
route-through
· rejects packets from
End-to-End Communication if queue
is too full
· reads Forwarding
Database and decides on which circuit to send packet
· takes care of packets
for unreachable
nodes
drops incoming
packets if the queue
is already full

• •

Control sublayer
-- -- -- ; - -- -- -- -- -- -- --.-- -- -- -- -- -- -- --. -- -- Initi;l~t~n~hlay~

DDCMP
Initialization

Ethernet
Initialization

• sends and receives
Initialization,
Verification, and
Hello and Text
messages

. sends and receives
Ethernet Router
and Endnode
Hello messages
. selects Designated
Router

DDCMP Module

Ethernet Module

Figure 2.14
Routing Layer Components and their
Functions

X.25
Initialization

X.25 Circuit
Database

• initializes PVCs
· issues outgoing
and accepts
incoming calls
· blocks, deblocks,
fragments, and
reassembles packets

Contains:
• Circuit
mapping
data

Data Link Layer

I-----------------I~----------

X.2S Level 3
Module

66 • DECnet Concept

layer down through each layer of the DNA hierarchy of the source
node, before being transmitted across a line. If the destination
node is not adjacent to the source node, the data travels up to the
Routing Layer of the adjacent node, where it is routed (or switched),
sent back down through the two layers, and transmitted across the
next line in the path. The data travels in this manner until it reaches
its destination node. At this node, the data passes up the hierarchy
of layers to the destination process in the User layer.
As the data passes down through the several layers in the source
node, it acquires control information. Each layer adds control
information to the packet of data it receives from the layer above it.
The control information by which data is enveloped as it travels
over a line ensures that it is transmitted with no errors and that it
arrives at the proper destination. Figure 2.15 illustrates what
happens to data as it passes through the DNA layers at a node. In this
example, the Network Management layer is not involved.

In the following scenario a user attempts to form a logical link
with another user. The requesting user passes initial connection
data to the destination user. The numbered steps correspond with
the numbers in Figure 2.16 following this explanation.
Data Flow at the Source Node
1. The source user requests a connection to the destination user
and passes connect data.
2. The Session Control module receives the data, maps the
destination node name to a numerical address, and places the data
in the next transmit buffer, adding control information to the
message. The message then passes to the End-to-End Communication layer.

..
D

User
Layer

C\
'-l

[§J

Considered as data
by layer

--------------------------------------~~~~~='----------Network
Network
Application
User
Application
Control
Data
Layer
Information

Session
Control
Layer

Session
Control
Control
Information

Network
Application
Control
Information

User
Data

End-to-End
Communication
Layer

End-to-End
Communication
Control
Information

Session
Control
Control
Information

Network
Application
Control
Information

User
Data

Routing
Layer

Routing
Control
Information

End-to-End
Communication
Control
Information

Session
Control
Control
Information

'--Network
Application
Control
Information

User
Data

Routing
Control
Information

End-to-End
Communication
Control
Information

Session
Control
Control
Information

Network
Application
Control
Information

User
Data

Data
Link
Layer
'I

Data
Link
Control
Information

,
,

Complete Envelope Transmitted
ftom One Node to an Adjacent Node

Figure 2.15
Information Building as Data Traverses
DNA Layers

,
,

Data
Link
Control
Information

68 • DECnet Concept

3. The End-to-End Communication layer adds its control infor-

mation and passes the message (now called a datagram) to the
Routing layer.
4. The Routing module adds a header conSIstIng of the

destination- and source-node addresses and selects an outgoing
circuit for the message based on routing information. Routing
then passes the message (now called a packet) to the Data Link
layer, specifying the outgoing circuit address and, if necessary, the
station address of the receiving node on the circuit.
5. The Data Link module adds its protocol header, consisting of
framing, synchronizing, addressing, and control information, and
its protocol trailer consisting of a cyclic redundancy check (CRC).
The frame is now enveloped for transmission.
6. The Physical layer module transmits the enveloped message over

the physical line.
Data Flow Across the Network to the Destination Node
7. The enveloped data message arrives at the next adjacent node.
The Physicallayer module receives the message and passes it to the
Data Link layer.
8. The Data Link module checks the packet for bit errors in
transmission. On links with a significant probability of transmission errors, the Data Link protocol performs error correction.
DDCMP and x.2S provide error protection by retransmitting. On
links with a low probability of transmission error-Ethernet links,
for example-packets received with errors are discarded, with
higher layers providing error recovery. The Data Link header and
trailer are removed from a correctly received packet, which is then
passed up to the Routing layer.

9. The Routing layer checks the destination address in the header.

If the address is not that of the local node, Routing selects the next
outgoing circuit from its routing table and passes the message back
to the Data Link layer. The Routing layer has routed the message
on to the next line in its path. The message proceeds as in steps 5
and 6 above.
10. The message proceeds through the network, switching at
routing nodes, until it reaches the Routing layer of the node with
the same address as the destination address in the message.

Data Flow at the Destination Node
11. The packet passes to the Routing layer of the destination node
as described in steps 7 and 8, above. The destination Routing
module removes the Routing header and passes the datagram to
the End-to-End Communication layer.
12. The End-to-End Communication layer module examines the
End-to-End Communication layer header in the datagram. If the
module has the resources to form a new logical link, it passes the
connect data, without the End-to-End Communication layer
control information, to the Session Control layer.

13. The Session Control module performs any necessary access
control functions (to ensure if the incoming connect request is
valid and has the privilege to be given access) and passes the
message to the appropriate process in the User layer after removing
Session Control header information.
14. The destination process interprets the data according t,)
whatever higher-level protocol is being used.

~~
S- ~

Node 1

Node 2

/o=l

2! N

~;....
C\

L ser
ayer

Data

Node 3

Data from user is enveloped in control
information as it passes through layers
to the routing layer.

0
~

Data;::

n
Cl

;::

Session
Control
Layer

Header information removed as data
rr----1
passes through layers to the user layer. --------+- I I r.._L_ I

r "

I Data I

rf-

End-to-End
•.
C
L ommUDlcatlon
ayer

3
1--.

R outing
. Iayer receIVes
.
packet from data link layer
and sends it to node 3.

iii

Routing
Layer

, ~MnA.' , ,

Data
Link

8 I neaaer! I I
Headers and data sent
from node 1 to node 2.

•I

Layers ~~~~__~"____- L____~
Data link control removed
in data link layer, node 2.

Data link control added
in data link layer, node 2.

....

Chapter 3 . DECnet Configurations

This chapter discusses how nodes and lines (presented in Chapter
2) can be arranged in various configurations.
The previous chapter explained how the routing mechanism
enables nodes that are not linked by a direct physical line between
them to communicate. Routing permits great flexibility in the
way that nodes and lines can be combined in networks, since every
node in the network does not have to be connected by a direct
physical line to every other node.
The major factors affecting network configuration are requirements of the applications that will use the network and capabilities
of the network components (environment, node characteristics,
and line and circuit characteristics). These capabilities will dictate
the limits within which networks may be configured.
Network environment is one of the capabilities affecting configuration. There are two basic environments: wide area networking
environment, with nodes distributed over a vast geographical area
(across a country or even across different countries), and local area
networking environment, with nodes distributed in a limited
geographical area (a building or cluster of buildings). Wide area
networks can include local area networks (see Figure 3.9). Within
wide area and local area networks, nodes can be configured in
several ways, depending on application requirements.
Node characteristics, such as the ability of the node to receive and
forward messages intended for other nodes and the type of
network services it can perform, greatly affect how a network
manager can position it in a network configuration. Whether the
node implements DNA Phase III or Phase IV software is also an
important consideration in determining how it will be configured
in the network.

72 • DECnet Configurations

Lines and their associated circuits can be arranged in bus structures,
point-to-point arrangements, or multipoint configurations.
This chapter on configurations does not describe configuration
details for specific DEC net systems or give definite specifications for
configuring a network. Rather, it is a general discussion on the
types of configurations that are possible using DEcnet. For details
on how to configure a specific system into a network, refer to the
system manager's guide for that particular DEC net system .
. DECnet Environments
There are two basic environments for DEC net configurations: wide
area networks and local area networks (LANS). Figure 3.1 shows a
wide area network and Figure 3.2 shows a local area network. Each
type can be configured with a variety of nodes and links. However,
only Phase IV DEC net nodes can be directly connected to a local area
network cable.
Performance and response time are better over a LAN than over a
wide area network. From an end user's perspective, this is the only
noticeable difference between the two types of network. All other
aspects of communication between nodes appear to the user to be
identical in either network.

Wide Area Networks
A wide area network is composed of nodes connected by individual communications links configured in various patterns, as
shown in Figure 3.1. In general, wide area networks are used for
long-distance communications. Nodes in wide area networks can
be located many thousands of miles apart.
In a wide area network, messages can be transported over dialup,
leased, or switched lines. Wide area networks typically use

Start-stop
modeDTE
To
Other
DEcnet
Nodes

NodeK
DEcnetVAX/PSI

DEcnetVAX/PSI

Other x.2S
User or
Network

NodeE
DEcnet
-l1s

Node A
DEcnet-11M
-PLUS/PSI

NodeD
DEcnet11M-PLUS

Multipoint

F
NodeG
NodeB
DEcnet
-VAX

NodeC
DEcnet
-11M

J
Other DEcnet Nodes
PSDN = Packet Switched
(Various Operating Systems)
Data Network
DeE = Data circuit-terminating equipment,
a PSDN switching node
DTE = Data terminal equipment, a user computer (packet mode)
or a terminal using the PSDN (start-stop mode)

Figure 3.1
Wide Area Network Configuration

74· DECnet Configurations

common carriers, like the telephone network, to transport messages over most or part of the distance. Because wide area
networks use the telephone system for most communications,
modems or modem equivalents are generally needed on each end
of a communications link to convert digital data to analog form,
and vice versa, and to permit transmission over long distances.
DEC net supports transmissions ranging from 1,200 bits per second

to 56 Kbytes per second for remote connections over leased or
dialup lines and over packet-switched data networks. A network
manager can specify different transmission speeds for the various
communications lines in a wide area network. Transmission
speeds depend on the supported hardware devices and the type of
lines used. Specifying transmission speeds gives a network manager some leeway in distributing workloads and adjusting performance levels for various network components.
See below under Node Characteristics and Line and Circuit Characteristics for guidelines on how to combine nodes in wide area
configurations. Any node restrictions in wide area configurations,
such as nodes that can be connected only to a local area network,
are also identified in these sections.
Examples of wide area networks include private packet-switched
networks such as the data networks developed by many companies, packet-switched data networks such as public data networks
using the x.25 interface (Telenet and Transpac, for example), and
leased -line networks.
Local Area Networks
Local area networks (LANS) are privately owned networks that
offer reliable high-speed communication channels. They are optimized for connecting information-processing equipment in a
limited geographical area, such as an office, a building, or a

DEe net
-RSX

Ethernet Coaxial Cable

DEenet

-RSX

-VAX

DEenet

-VAX

-RSX

Figure 3.2
An Ethernet Local Area Network

76· DECnet Configurations

complex of buildings. LANS can be designed with a variety of
technologies and arranged in different configurations. Consequently, they vary significantly with respect to their transmission
speeds, the distances they span, and the capabilities and services
they offer.
An Ethernet baseband LAN consists of a segment of coaxial cable or
several segments of coaxial cable joined together, with each
segment ranging in length from 20.4 to 500 meters. The Ethernet
cable is terminated at both ends (see Figure 3.3).
A Phase IV DEC net Ethernet network supports up to 1,023 nodes.
Ethernet supports a bus topology, with each node attached to the
cable by a single line. The channel bandwidth of an Ethernet cable
is 10 Mbytes. The controller on the host system and host-system
bottlenecks (caused by disk I/O, for example) limit CPU
throughput to considerably less than 10 Mbytes. Typically, each
node can achieve a throughput between 400 and 1,400 Kbytes.
Nodes that support a UNIBUS structure, such as PDP-11 and VAX
nodes, connect to the Ethernet cable by means of a network
adapter called the DEUNA. Nodes with a Q-BUS structure, such as
MicrovAx and MICRO-11 systems, connect using a network
adapter called the DEQNA. The DECNA controller connects the
Professional 350 personal computer to the Ethernet cable. The
DEUNA, DEQNA, and DECNA adapters, which are installed as

•
Figure 3.3
Ethernet Cable

•

communications devices in a DEenet node, incorporate hardware
module boards, a distribution panel, and cabling. All three
adapters connect to Digital's H4000 transceiver physically and
electrically by a transceiver cable (see Figure 3.4 for DEUNA and
DEQNA connections). A maximum of 100 transceivers can be used
on a single SOO-meter segment of Ethernet cable.
A communications controller called the DELNI enables the connection of up to eight nodes or devices to each H4000 transceiver. The
DELNI can also be used in a stand-alone configuration, unconnected
to the Ethernet coaxial cable. In this configuration, the DELNI
supports up to eight devices (systems, not terminals) using
standard Ethernet transceiver cables up to 50 meters away. The
DELNI and connected devices form a local area network of their
own. The DELNI can also be arranged in a hierarchical stand-alone
configuration that is not connected to a main Ethernet segment.
Here, the DELNI supports up to 64 devices using Ethernet transceiver cables up to 50 meters away. Configurations using the DELNI
are cost-effective for LANS that span a limited area. When an
Communications
Device

ITransceiver Cable

~Maximum

Ethernet Coaxial Cable

Figure 3.4
Ethernet Hardware

;W()'()"Transceiver and
Coaxial Cable Tap

78· DECnet Configurations

Ethernet Coaxial Cable

Figure 3.5
DELNI

organization's networking needs grow, a larger Ethernet LAN can
be easily implemeted by connecting the DELNI configuration to a
coaxial cable via an H4000 transceiver. Figure 3.5 shows PRof
DEC net nodes on a DELNI that is attached to an Ethernet cable.
Segments of coaxial cable can be connected to extend an Ethernet
LAN beyond the SOO-meter limit of a single segment. Two cable

segments are joined using a device called a repeater (see Figure 3.6).
A repeater enables the two cable segments to function as if they
were one cable. It amplifies transmission signals and passes data
packets between two coaxial cable segments without filtering
signals. Digital local area networks can use two types of repeaters:
Local repeater-Connects two coaxial cable segments within a
limited geographical distance (within a building, for example).

There can be a maximum of 100 meters between the two cable
segments, and each cable segment can be a maximum of 500
meters long.

Remote repeater-Connects two coaxial cable segments spanning a
large distance (between two buildings, for example). The remote
repeater consists of two local repeaters connected by a fiber-optic
link up to 1,000 meters long.
Both local and remote repeaters connect to an H4000 transceiver
on the coaxial cable using transceiver cables (see Figure 3.6). No
more than two repeaters can be placed between any two nodes on a
single Ethernet.
Other characteristics of Ethernet configurations are:
· Transceiver cables connecting nodes to the Ethernet cannot exceed 50
meters.
• Cable segments must be terminated in 50 ohms.
• The maximum distance between any two nodes is 2,800 meters (this
includes the length of the Ethernet cable, the transceiver cables, a
local repeater, and a remote repeater).
All DEcnet nodes directly connected to an Ethernet cable must be
Phase IV nodes. However, Phase III nodes or Phase IV nodes that
are not on the Ethernet can gain access to Ethernet resources
through a DEC net Router Server, a DEC net Router/x.2s Gateway,
or a full-function Ethernet node acting as a Phase IV Router.
All the server nodes described below are supported on the
Ethernet. (There is aDEcnet/sNA Gateway implementation that
supports DEcnetl SNA communications in a wide area network
configuration as well.) Server nodes are special purpose nodes on

80 • DECnet Configurations

Node
B

Node
C

Node
D

Fiber Optic Cable
::s 1000m

Ethernet Coaxial Cable

Node
E

Node
F

Figure 3.6
Repeaters

the Ethernet that perform communication functions for the other
nodes~ thereby freeing the latter of communication processing.
Server nodes are identical in hardware design. The software that
distinguishes them and prepares them for their special communications functions is downline-loaded from a designated host node
on the same Ethernet LAN. A host node is a full-capability DEC net
node whose services are required in the startup procedures or

operations of another node. Documentation for each server node
includes installation information describing the requirements of a
host node.
One type of DEC net system, PRO/DEcnet, can partICIpate in
network activities only as an Ethernet node. (PRo/DEcnet software enables a Professional 350 series personal computer to
participate as a node in a DEcnet Ethernet network. PRo/DEcnet is
distinguished by a menu-driven user interface. Refer to the PRof
DEC net manual set for more specific information.) Once connected
to an Ethernet cable, however, PRO/DEC net can access all the
resources available to any other Phase IV DEC net node. Routers and
servers enable the Professional 350 running PRo/DEcnet to communicate with systems on another Ethernet and with remote
Phase IV and Phase III systems not on an Ethernet network (see
Figure 3.9).
Nodes connected to an Ethernet cable can connect with wide area
networks, other local area networks, packet-switched data networks that implement the x.25 protocol, and systems in IBM SNA
networks. Figure 3.9 shows that x.25 and DEcnet/sNA Gateways
can connect directly to the Ethernet cable.
For more information on Ethernet local area networks, refer to the
following Digital manuals:

• Introduction to Local Area Networks
• Networks and Communications Buyer's Guide
· Node Characteristics
Characteristics of a node that affect network configurations are whether
it supports Phase III or Phase IV DNA implementation, what its
routing capabilities are, and whether it is a dedicated or general
purpose node.

82· DECnet Configurations

Phase III And Phase IV Nodes
DEC net products have been developed in a series of phases. Not all
Digital operating systems implement the same phase of DEcnet,
nor do they all implement the full capabilities of each phase.
Chapter 1 discusses the DEC net phases and capabilities currently
supported by each Digital operating system.
In planning a network configuration, the phase of DEcnet that is
installed on a node is important, since Phase IV DEcnet extends the
networking capabilities of a node beyond what is offered in Phase
III. Only Phase IV nodes can connect directly to an Ethernet cable.
Phase III nodes can communicate with Ethernet nodes through a
DEcnet Router Server or a Phase IV node. The Phase IV area
routing mechanism supports networks of over 64,000 nodes (refer
to Routing in Chapter 2 for details), while Phase III supports up to
255 nodes.
Communications servers, described below, all implement Phase
IV software. Servers are important to configuration plans because
they offload functions, such as routing decisions (Chapter 2) and
virtual terminal capabilities (Chapter 7), that are normally performed by a general function node. An Ethernet LAN configured
with a DEC net Router Server could support a number of nodes
without routing software (such nodes are called end nodes). The
router would provide routing functions for these end nodes.
Throughout this handbook, differences between Phase III and
Phase IV will be identified and discussed as they affect various
aspects of network operations.

Routing Capabilities
From the point of view of routing capabilities, there are two types
of DEcnet nodes: end nodes and full-function nodes.

End node-End nodes do not have route-through capability. This
means that they cannot receive and forward messages intended for
other nodes. An end node can send packets to an adjacent node. If
an end node has multiple circuits to one or to several adjacent
nodes, only one of those circuits can be active at a time.
If a node adjacent to an end node is a full-function node, its routethrough facilities can be used by the end node to establish
connections with other nodes in the network. Nodes A, D, and H
in Figure 3.7 are end nodes.

Fulljunction node-A full-function node can send packets from itself
to any other Phase IV node, regardless of its location in the
network. It can receive packets from any other Phase IV node and
it can route packets intended for other nodes. Full-function nodes
can have multiple circuits actively communicating with one or
several nodes at the same time. Full-function nodes also maintain
an up-to-date map of the network in their database. The map
provides routing information, such as the node address of each
accessible node in the network and circuit costs. This information
is used by the routing mechanism to determine the most costeffective paths for sending data to a destination node. Data
destined for a node that is not adjacent to the sending node can be
routed though one or rnore full-function node. Nodes B, C, E, F,
and G in Figure 3.7 are full-function nodes.
Levell and Level 2 routers in area-based networks are full-function
nodes. A Levell router routes packets between nodes within its
area. When it receives a packet that is intended for a node in
another area, the Levell router forwards the packet to the nearest
Level 2 router in its area. The Level 2 router then routes the packet
through other Level 2 routers to a tevel2 router in the destination
area, which forwards it to a Level 1 router in that area. Level 1
routers in the destination area forward the packet to the destination node.

84 • DECnet Configurations

NodeD
DEcnet

-20

NodcC
DEcnet
-VAX

Ethernet Coaxial Cable

Figure 3.7
Node Routing Capabilities

General Purpose And Dedicated Nodes
Also affecting network configuration is whether or not particular
nodes in the network are general function or special purpose.
Dedicated or special purpose nodes have specific functions that can
service the needs of many other nodes. Figure 3.8 shows dedicated
DEC net nodes providing routing services to a group of nodes on an
Ethernet cable. Also in the figure are gateways that connect
Ethernet LANS to x.2S and IBM/ SNA networks.

Most of the dedicated nodes discussed in this handbook must be
configured on an Ethernet cable. The one exception is an implementation of the DEcnet/ SNA Gateway, which can also be config-

DEenet
-RSX

Ethernet Coaxial Cable

DEenet
-VAX

-VAX

DEenet
-RSX wi

Observer

Figure 3.8
General Purpose and Dedicated Nodes

ured in a wide area network. General purpose nodes can be
configured in a wide area or local area network environment.

86· DECnet Corifigurations

DEC net nodes performing dedicated communications functions
are called communications servers. With the exception of the
Terminal Server, the communications servers implement Phase IV
DEC net software. (Because these servers are dedicated to a specific
function, they cannot be accessed as regular Phase IV nodes for
general user functions.) The Terminal Server implements the
Local Area Transport (LAT) protocol (see Chapter 7 for details).
DEcnet communications servers must be downline-Ioaded by a
DEC net host node on the Ethernet cable. (The exception to this
rule is the wide area version of the DEcnet/ SNA Gateway. A
DEcnet/sNA Gateway intended for a wide area network can be
located in a wide area network and loaded directly from a floppy
disk.) DEC net communications servers perform the following
functions:

DECnet Router Server-This is Phase IV DEC net node that performs
routing functions only. The DEcnet Router Server handles the
routing of traffic between nodes on two Ethernets and between
nodes on an Ethernet network and remote Phase IV or Phase III
nodes that are not on the Ethernet. The DEC net router offloads
routing functions from general function host nodes so that their
resources can be used more efficiently to support user applications.

Terminal Server-This server enables the connection of several
terminals to a host system on an Ethernet network (see Chapter 7
for details on the LAT protocol operating over the Ethernet
terminal-host link). Through the use of a single command, users
on terminals connected to this server can establish a nonblocked
logical connection to any Phase IV node on the Ethernet network.
Since the terminal server provides access to multiple hosts, it
potentially reduces the number of direct terminal connections to
individual hosts within a LAN. Thus, power and packaging

requirements on the hosts are reduced. The terminal server allows
terminals to be distributed close to where they are used, thereby
reducing cabling costs and complexity. By locating the terminal
server with the terminal users, the task of managing the multiple
cables running from the terminals to the server can be simplified.
Network reliability increases with the tern"linal server. When a
host with many terminal connections fails, all the terminals
connected to it become inoperative. However, if a host to which
terminals are logically connected via a terminal server fails, the
terminals are not inoperative and can access other hosts on the
network.

DECnet Routerlx.25 Gateway-This gateway makes possible communication between DEcnet nodes on an Ethernet and DEC net
nodes connected to an x.2S PSDN. DEC net-vAX nodes on an
Ethernet can communicate through this gateway with nonDigital systems that support the x.2S protocol and with asynchronous terminals that are physically connected to an x.2S network
(see Chapter 8 for details on additional XEP software that the
DEC net-vAX nodes require). And terminals on the Ethernet that
are logically connected to VAX systems with XEP software can
access remote systems connected to a PSDN. Through the gateway,
every node on the Ethernet network has access to DEC net nodes on
the x.2S PSDN, minus the expense and problems of reliability
associated with direct connection of every system to the x.2S
network.
The DEC net Router/x.2S Gateway also provides users with all the
capabilities of the DEC net Router for making point-to-point
connections between systems observing the Digital Data Communications Message Protocol (DDCMP).

DECnetlSNA Gateway-This server enables users to derive the
complementary benefits of Digital and IBM computing environments. There are two DEcnetl SNA gateway products: one for wide

88· DECnet Configurations

area networks and one for local area networks. Through either of
the DEcnet/sNA Gateway products, Digital systems on an
Ethernet or non-Ethernet DEC net network can communicate with
IBM host systems in a Systems Network Architecture (SNA)
network. Through the DEcnet/ SNA gateway, users within a
DEC net network can perform Remote Job Entry, 3270 Terminal
Emulation, and Application Program Interface (for task-to-task
communication) into the IBM SNA network. These functions can
be easily installed and integrated into existing user procedures,
thereby providing a transparent and bidirectional flow of information between the Digital and IBM networking environments. A
gateway management capability enables a DEC net network manager to monitor and control gateway performance.
General purpose nodes perform a variety of high-level network
functions (file transfer, remote resource access, task-to-task communication, and network virtual terminal, for example) that are
discussed in Chapters 4 through 9 of this handbook. All Phase IV
general purpose nodes, with the exception of a PRo/DEcnet node,
can run in either a wide area or an Ethernet environment, unless
otherwise specified .
. Line and Circuit Characteristics
Network nodes in a wide area network can be connected using point-topoint or multipoint lines and circuits. Ethernet nodes are connected by
means of a bus topology. The type ofline/ circuit between two nodes can
have a marked effect on access capabilities. For instance, data can travel
faster on a point-to-point line/ circuit than over a multipoint line/circuit.

A point-to-point line connects two nodes using a single circuit. In a
point-to-point connection, nodes always have access to the communications path linking them.
A multipoint line is shared by more than two nodes (see Figure 2.1).
Each node communicates over the line using a separate circuit.

One node on the line, called the control station, controls access to
the shared communications path; the other nodes on the line are
known as tributaries.
A channel-access method known as polling determines when each
tributary can use the communications path in a multipoint
connection. In polling, the control station sends a message to each
tributary in an order prescribed by the control-station software.
The poll message asks if the tributary has any data to send. If the
tributary does have data to send, the control station allows the
tributary to transmit. When transmission is completed, the
communications path is freed, and the control station polls the
next tributary. If a tributary has no data to send when it is polled,
the control station passes it and polls the next tributary. If data is
sent to a tributary from another tributary or system in the
network, it must first pass through the control station, which
delivers the data (using the routing mechanism) to the specified
tributary when the communications path is freed.
A multipoint connection can be more econornical than a point-topoint one, as there are several stations communicating over a single
line. However, a node may not always have access to the communications path when needed.
Chapter 2 discusses the access method and transmission and
reception of data over Ethernet links .

. The Total Picture
Figure 3.9 shows an example of a large network that includes a
wide area network and a local area Ethernet network. In such a
network environment, there are a variety of configurations that
can be implemented to serve a wide range of application needs. All
nodes in the network can communicate with each other and
combine their resources to perform network activities as an active,
cohesive unit.

90· DECnet Configurations

DEcnet
-VAX

Terminator

Ethernet Coaxial Cable

Figure 3.9
The Total Picture

DEcnet
-VAX

DEcnet

-RSX

-VAX

DEcnet/E

Phase III

DEcnet-20

Phase III

DEcnet/E

Phase III
End Node

Chapter 4· Common Mechanisms in DEC net Functions

Once a DEC net system has been configured in a network (Chapter
3) and installed (Chapter 10), it is available to users for a variety of
high-level network functions, including:
• Task-to-task communication (Chapter 5)
• Remote file and record access and file transfer (Chapter 6)
• Terminal communications (Chapter 7)
Chapter 8 describes how users can perform these functions over
foreign-vendor networks, and Chapter 9 describes application
environments that implement these functions.

All of the high-level network functions listed above have common
mechanisms at work within them, including logical links, segmentation and reassembly of data, error control, flow control, and
access control. This chapter discusses these mechanisms.
· Logical Links
Every node runs multiple programs and processes. Whether they are in
the same node or in different nodes, processes or programs that have to
communicate with each other need some way of establishing contact and
exchanging data, unhindered by the confusion and interference of other
network traffic. To facilitate the orderly flow of information between
two programs or processes, DEC net implements a mechanism called a
logical link. It is a temporary conversation path established between
two communicating programs (or processes) in a DEC net network.
The communicating programs can reside on nodes located anywhere in the network, since DEC net establishes logical links
between nonadjacent as well as adjacent nodes. (Do not confuse
logical links with lines or circuits. A line is the physical medium
over which data travels; a circuit, which operates over lines, is a
logical communication path between adjacent nodes. See Chapter
2.) Figure 4.1 shows typical logical links.

NSP

Node A

NodeB

_-------------\,:-r:gical

Node X

----------------\t!~~~- ------t ---------------------

;-.----7

",'

NSP

NodeY
;--,..,.......,._.,.---~

Figure 4.1
Logical Links

When Are Logical Links Required?
Almost all network functions, including those discussed in Chapters 5, 6, 7, and 8, require the services of a logical link between
programs. The exceptions are some maintenance functions, such
as loopback tests and the downline-loading of certain systems
(downline-Ioading of an Rsx-lls DEC net node, for example).
Logical links operate in the following types of connection:
• A user program connected to another user program
• A user program connected to a DEcnet module
• A DEcnet module connected to another DEcnet module

In the first type of connection, the application programmer
specifies a series of DEC net calls in each user program to establish
and direct the operation of the logical link. In the second type of

94 • Common Mechanisms in DECnet Functions

connection, the programmer includes calls that initiate and control the link in the user program only; the DEC net module
automatically handles its end of the link. In the last case, the
creation and operation of the logical link is a level removed from
the user; user programming is not involved in the connection at
all. The two DEcnet modules exchange messages based on information supplied by their local system.

How To Create A Logical Link
Logical links are created by an interaction of DEC net calls specified
in the two cooperating programs. This interaction is referred to as
the "handshake procedure." In this procedure, DEC net calls are
passed between the two programs to establish an agreement to
communicate. Unless both programs agree, a logical link cannot be
created.
The handshake procedure takes place in a prescribed order. First,
one program, the source program, issues a call to request communication with a second program, the target program. DEC net
software in the source program's node carries the connection
request to DEC net software in the target program's node. (If the
two programs are in the same node, the DEC net software in that
node alone is involved in establishing the connection.)
The target program has the option of accepting or rejecting the
source program's request for connection. If the target program
returns a call accepting the connection request, a logical link is
established between the two programs. At this point, source and
target distinctions become irrelevant, and both programs can issue
calls to send and receive data until either program decides to exit or
abort the connection. If the target program returns a call rejecting
the connection request, no logical link is created.

The calls required to establish a logical link and to exchange
data with another program are described in the programmer's
reference manual for each DEcnet system. Also listed in the manual
are the programming languages that can be used to to issue these
calls. Chapter 9 discusses logical link processing in an applications
environment.

Behind The Scenes-DEC net Software And Logical
Links
The DNA levels primarily responsible for affecting process-toprocess communication through logical links are the End-to-End
Communication and Session Control layers. End-to-End Communication handles the system-independent aspects and Session
Control the system-dependent aspects of process-to-process
communication.
End-to-End CotDtDunication Layer and NSP
The modules of the End-to-End Communication layer are responsible for the creation and destruction oflogicallinks. The information passed between two programs in the handshaking procedure
and in all communications activity is intercepted and evaluated by
the End-to-End Communication modules in each node. Information passes from the local progralTI to the local End-to-End
Communication software, from there to the remote End-to-End
Communication modules, and finally to the remote program (as
shown in Figure 4.2).
The End-to-End Communication modules in the local and remote
nodes communicate through the Network Services Protocol
(NSP). NSP functions include:
• creation and destruction of logical links .
• guaranteeing the delivery of data and control messages in sequence
to a specified destination by means of an error control mechanism.

96 • Common Mechanisms in DECnet Functions

Node A

--- ...

Data--------_

QJ----

Logical Li~k-------- ___ _

Node C

-___

Node B

-n-ata------------- --~-- ------------------------d~~,:~~~~

Figure 4.2
DECnet and Logical Links

• managing the movement of interrupt and normal data (see under
Data Types, below) from transmit buffers, using flow control
mechanisms (see below, under Flow Control) .
• breaking up normal data messages into segments that can be
transmitted individually and reassembling these segments in
proper sequence upon reception (see Segmentation and Reassembly
of Data, below).
NSP Messages-There are three types of NSP messages-data,
acknowledgment, and control. Table 4.1 summarizes the functions that each NSP message type performs.
NSP establishes, maintains, and destroys logical links by exchanging control messages with other NSP modules or with itself (if the
communicating programs or processes are in the same node).

Type

Message

Description

Data

Data Segment

Carries a portion of a Session Control
message. (This has been passed to Session Control from higher DNA layers
and Session Control has added its own
control information, if any.)

Data (also called
Other Data)

Interrupt

Carries urgent data, originating from
higher DNA layers. It also may contain an optional Data Segment acknowledgment.

Data Request

Carries data flow control information
and optionally a Data Segment acknowledgment (also called Link Service message).

Interrupt Request

Carries interrupt flow control information and optionally a Data Segment
acknowledgment (Link Service
message).

Data
Acknowledgment

Acknowledges receipt of either a Connect Confirm message or one or more
Data Segment messages, and optionally an Other Data message.

Other Data
Acknowledgment

Acknowledges receipt of one or more
Interrupt, Data Request or Interrupt
Request messages, and optionally a
Data Segment message.

Connect

Acknowledges receipt of a Connect
Initiate message.

Connect Initiate
and Retransmitted
Connect Initiate

Carries a logical link connect request
from a Session Control module.

Connect
Confirm

Carries a logical link connect acceptance from a Session Control Module.

Disconnect
Initiate

Carries a logical link connect rejection
or disconnect request from a Session
Control module.

No Resources

Sent when a Connect Initiate message
is received and there are no resources
to establish a new logical link (also
called Disconnect Confirm message).

Disconnect
Complete

Acknowledges the receipt of a Disconnect Initiate message (also called Disconnect Confirm message).

No Link

Sent when a message is received for a
nonexistent logical link (also called
Disconnect Confirm message).

No Operation

Does nothing.

Acknowledgment

Control

Table 4.1
NSP Messages

98· Common Mechanisms in DECnet Functions

Figure 4.3 shows a typical message exchange. In this figure, an NSP
module first initiates a connection, then sends data, and finally
disconnects the link based on commands from the Session Control
layer. The figure does not show the messages that control data
flow (these are illustrated in Figure 4.7, below).

NodeB

Node A
1 User at Node A requests connection.

connect

IInitiate

----..

2 User at Node B accepts connection.

3 User at Node A requests that a message (that is two segments long)
be transmitted. NSP at Node A sends the first data segment.

4 NSP at Node B acknowledges receipt of the first data segment.

~~
5 User at Node A sends the second data segment.

6 NSP at Node B acknowledges receipt of the second data segment and
gives the message to the user at Node B.

----chlJ
7 User at Node A requests a disconnection and NSP at Node A sends
a Disconnect Initiate message.

I Di.s~onnect I

•

.Inltlate
.
8 NSP at Node B receives Disconnect Initiate message, sends a Disconnect
Confirm message, and informs the user at Node B.

Figure 4.3
Typical Message Exchange between
NSP Modules in Two Nodes

Other logical-link functions performed by End-to-End Communication modules on each node include:
• Establishing link addresses to identify each end of a logical link (see
below, under How the Program Identifies Logical Links)
• Formatting outgoing data for transmission by communication
hardware
• Allowing multiple logical links to be created over a single physical
line by separating data into logical-link streams (see under Multiple
Logical Links Within a Program, below)
• Removing incoming data from logical-link streams and delivering
it to the appropriate program.
Session Control Layer
The modules in this layer provide system-dependent process-toprocess communication functions, thereby bridging the gap
between the End-to-End Communication layer and the logicallink functions required by processes running under an operating
system.

Session Control functions include:
• Mapping node names to node addresses-A Session Control module maintains a node-name mapping table or database that lists the
names of nodes with which the local node expects to communicate, as well as the unique network address of each node name in
the table. When a local end-user process requests connection to an
end-user process on a remote node, the Session Control module
consults the table to select the node address of the target program
or the number of the physical channel on which the connect
request will be transmitted.
• Requesting logical links on behalf of end-user processes (see
below, under Requesting a Connection for details)-The Session

100· Common Mechanisms in DECnet Functions

Control layer passes destination node address and channel number
to the End-to-End Communication layer when making logicallink requests for end-user processes.
• Receiving connect requests addressed to end-user processes (see
under Receiving a Connect Request, below)-For incoming connect
requests from the End-to-End Communication layer, a Session
Control module uses the node-name mapping table to identify the
node from which the request originated. The identity of a node is
important in determining the validity of a connect request (see
below, under Access Controlj.
• Sending and receiving logical link data-Requests by end-user
processes to send and receive data are passed directly by the Session
Control layer to the End-to-End Communication layer.
· Disconnecting and aborting a logical link-As in the case of
sending and receiving data, requests by end-user processes to
disconnect and abort logical links are passed directly by the Session
Control layer to the End-to-End Communication layer. When a
logical link has been disconnected or aborted, the Session Control
layer notifies the end-user process of the event.
• Optionally, monitoring logical links-This function can be used at
either end of the logical link to detect a probable network
disconnection. The Session Control layer can also optionally
detect failure on the part of the End-to-End Communication layer
to deliver received data in a timely manner.
The Session Control layer defines a database containing the states
of Session Control and optional default connection timers.
· Identifying end-user processes-A Session Control module executes a system-dependent algorithm to determine if an existing
end-user process corresponds to the destination end-user process
specified in an incoming connect request. It also performs additional functions related to passing a connect request to an existing
end-user process.

101

• Activating or creating processes-A Session Control module can
create a process or activate an existing one to handle an incoming
connect request.
• Validating incoming connect requests-A Session Control module
uses access-control information included in an incoming connect
request to perfornl system-dependent validation functions (see
below, under Access Control).
Figure 4.4 shows the relationship of the Session Control modules
to end-user processes and to the End-to-End Communication
modules.
Requesting a Connection
When it receives a logical-link connect request from an end-user
process, the Session Control layer performs four tasks:

· Identifies the destination node address or channel number for the
End-to-End Communication layer by using the node-name mapping table.
• Formats connect data for the End-to-End Communication layer.
• Issues a connect request to the End-to-End Communication layer.
• Optionally, starts an outgoing connection timer. If the timer
expires before the End-to-End Communication layer issues an
acceptance or rejection of the connect request, Session Control
disconnects the logical link for the source end-user process.
Receiving a Connect Request
When the Session Control layer detects an incoming connect
request from the End-to-End Communication layer, it performs
six tasks:

• Parses connect data to obtain such information as source and
destination end-user process names and access-control information.

102 • Common Mechanisms in DECnet Functions

A Network Node

Comments

User
Processes

End User End User End User End users are User, Network Application, and
2
3
Network Management
modules.

System
Dependent
Modules

Operating
Session
..-- Control
System

--@

terface to the End-to-End
Communication layer for
end user processes. It
functions in conjunction
with the operating system. CD and ® are data
bases used by Session
Control.

System
Independent
Modules

End-to-End Communication
Layer

The End-to-End Communication layer provides
the logical link service to
Session Control. Its functions are not dependent
on individual operating
systems.

The Network

-<D Session Control is an in-

* Network Management interfaces with Session Control in two ways:
(1) to obtain access to the logical link service and (2) to monitor and
control Session Control operations.

Figure 4.4
Session Control Interaction with
End-User Processes and End to End
Communication

103

• Validates any access-control information.
• Identifies, activates, or creates a destination end-user process.
· Maps the source node's address or channel number to a node
name, if there is one.
· Delivers the incoming connect request to the end-user process.
• Optionally, starts an incoming timer when the connect request is
delivered. If the timer expires before the end-user process accepts
the connect request, the Session Control layer issues a reject to the
End-to-End Communication layer.

Session Control Messages
Session Control's message protocol defines messages sent on a
logical link as connect data, reject data, and disconnect data.
Figure 4.5 shows the formats for the Session Control messages.
The numbers below each message field indicate its maximum
length in bytes.
How The Program Identifies Logical Links
A program can establish and use more than one logical link at a
time. The maximum number oflinks that can run simultaneously
in a single program is usually determined by the programmer or by
system restrictions. (The programmer's reference manual for a
system includes details on the logical-link requirements of that
system.) Since a program can use multiple logical links, there must
be some way for it to differentiate among the many logical links
that may be running wi thin it. DEC net effects the differentiation by
having each program specify a unique link identifier during every
handshake procedure to identify the link that may be created. The
link identifier defines a link only within the context of a particular
program; two cooperating programs do not have to specify the
same link identifier for the same link.

104 • Common Mechanisms in DECnet Functions

Connect Data Message Format
19

Reject/Disconnect Data Message Format

REASON

DATACTL

DSTNAME
SRCNAME
MENUVER
RQSTRID
PASSWRD
ACCOUNT
USRDATA
REASON
DATACTL

I
the destination end user name
the source end user name
the field format and version format
the source user identification for
access verification
the access verification password
the link or service account data
the end user process connect data
a reason code
user data (length of field determined by
the total length of reject or disconnect
data received from the End-to-End
Communication layer)

Figure 4.5
Session Control Message Formats

End-to-End Communication software on each node takes the link
identifier specified by its local program, and both sets of End-toEnd Communication modules agree on a pair of link addresses
that they associate with a particular link. Link addresses are
meaningful only to the End-to-End Communication modules;
they are transparent to the end user.

Multiple Logical Links Within A Program
A single program can establish logical links that communicate
with several programs, or establish several logical links with the

105

same program to exchange data intended for different purposes.
For example, two prograllls can establish two logical links
between them; one link can be used to transmit transaction data,
while the other can be used to transmit control information.

Data Types Within a Logical Link
A logical link can be considered to comprise two separate data
subchannels, each carrying messages in both directions (a logical
link operates in full-duplex mode). One data subchannel carries
normal data, which is the subject matter of most data exchanges
between two programs. User data, such as a list of messages and
the responses sent by the receiving program, would constitute
normal data. The other data subchannel carries interrupt messages,
data request messages, and interrupt request messages.
To interrupt the flow of normal data, either program can send
interrupt data. Interrupt data is high-priority information signaling the occurrence of some event requiring immediate attention.
Interrupt data breaks through the normal data flow. The means of
delivering interrupt data are system- and program-dependent, but
software in the receiving node usually reads it before accepting any
normal data that may be waiting to be delivered.

· Segtnentation and Reassembly of Data
The End-to-End Communication modules that manage each end
of a logical link guarantee:
• that all transmitted data is delivered to the destination node in
proper sequential order.
• that all data received by the NSP module on the destination node is
given to the target program in the proper sequence.

106 • Common Mechanisms in DECnet Functions

The Routing layer limits the amount of user data that an NSP
module can send in one datagram (data without routing information is known as a datagram). Taking normal data from Session
Control buffers, the transmitting NSP module breaks it up, if
necessary, into segments. To guarantee proper segment sequencing, the NSP module numbers the segments and transmits them
along with other control information in a Data Segment message
over a link. The receiving NSP module uses the sequence numbers
to reassamble the data segments in correct sequence in the
receiving node's Session Control buffers.
End-to-End Communication software assigns a separate set of
transmit numbers to interrupt messages. (Interrupt messages are
not segmented, however, because interrupt data is limited in size
and always fits in a single datagram). The discrete sets of numbers
for normal and interrupt data logically divide them into separate
data streams within the logical link.
Error Control
The receiving NSP module acknowledges delivery of the segments
by using the transmit numbers for identification. Optionally, a
receiving NSP module can negatively acknowledge normal data it
has received, if the data has to be discarded (from being irreparably
out of order, for example). If the transmitting NSP module receives
a negative acknowledgment or fails to receive a positive acknowledgment within a certain period of time, it retransmits the data.

There are variations of detail among the several implementations
of DEcnet in the execution of segment acknowledgment and
retransmission. However, despite the differences, all the DEC net
implementations use these mechanisms to guarantee delivery of all
transnlitted segments and to ensure that the segments are delivered
in the proper sequence. Figure 4.6 illustrates data segmentation
and acknowledgement.

107

The data-transmitting NSP assigns a transmit number to the message, and starts
a timer.
transmit number = n
---. Data
Segment Message

---

...

Data-transmitting
NSP

Data-receiving
NSP
f---

transmit number = m
Other Data Message

--- .....

Data Subchannels

2 If the timer times out, the message is retransmitted.

3 If the timer does not time out, and the flow control mechanism allows another
message to be sent, the data-transmitting NSP assigns the transmit number plus
one to the next data message transmitted in that subchannel.
1- __ ....

transmitnumber=n+l
Data Segment Message

Data-transmitting
NSP
r-- .... transmitnumber=m+l

Other Data Message
Data Subchannels

___ •

----

Data-receiving
NSP

4 When the message with the first transmit number is received by the data-receiving NSP, it returns that number as an acknowledgment number within the first
acknowledgment.
5 If the next data message transmit number received is equal to the current acknowledgment number plus one, the data-receiving NSP accepts the data message,
incrementing the acknowledgment number. It then sends the new receive acknowledgment number back to the data-transmitting NSP within an acknowledgment message.

,-----------------,

I receive ack. number = n

,----4 Data Acknowledgment ~I i
Message*
I
L __________________ J

receive ack. number = n + 1
Data Acknowledgment
Message*
Data-receiving
NSP

Data-transmitting
NSP
receive ack. number = m
-- Other Data-Acknowledgment
Message
Data Subchannels

Figure 4.6
Acknowledgement Operation

---

108· Common Mechanisms in DECnet Functions

(Continued)
6 However, if the data-receiving NSP receives a data message transmit number
less than or equal to the current receive acknowledgment number for that sub channel,
the data segment is discarded. The data-receiving NSP sends an acknowledgment
back to the data-transmitting NSP. The acknowledgment contains the receive
acknowledgment number.
7 If the data-receiving NSP receives a data message transmit number greater than the
current receive acknowledgment number plus one for that subchannel, the data segment
may be held until the preceding segments are received or it may be discarded.
*The data-receiving NSP might not send an acknowledgment for each data
message received. The receive acknowledgment number implies that all previous
numbers were received .

. Flow Control
Network programs and DEC net software both require a certain
amount of buffer space for temporary message storage. For
example, a transmitting NSP module keeps a copy of every message
it sends over a link until the receiver acknowledges receipt of the
message. At the program level, buffer space is necessary to hold
inbound messages waiting to be processed. DEcnet software and
programs need buffer space for other purposes as well, depending
on the application and the DEcnet implementations.

Without some kind of control, message traffic can easily cause
available buffer space to overflow. If this happens, communications overhead is incurred, because a message must be present at
specified intervals until the destination node can accept it. To
prevent this, the programs and NSP modules exercise flow controL
NSP'S flow-control mechanisms ensure that data is not lost for lack

of buffering capability and that deadlocks do not occur. Both
normal and interrupt data are subjected to flow controL

109

The data receiving part of NSP controls data flow. When a logical
link is formed, the NSP modules in each node inform one another of
the manner in which they, as data receivers, want to control the
flow of normal data. The receiving NSP module can choose not to
implement flow control or it can choose one of two types of
normal data flow control:
• Segment-the receiver sends a request count of the number of
segments it can accept. See Figure 4.7 on Segment Flow Control.
• Message-the receiver sends a request count of the number of
Session Control messages it can accept. (Message flow control is
obsolescen t and will be eliminated at a future date).
The transmitting NSP module uses the receiver's segment request
count to determine whether to send data. As an added control, the
receiver can always tell the transmitter to stop sending data
unconditionally or to start sending data under the normal request
count conditions. The receiver also controls interrupt data flow
with an interrupt request count.
· Access Control

The Session Control layer in most DEC net nodes performs accesscontrol functions by examining logical-link requests and preventing unauthorized access of node resources. A program or process
on one node that wishes to communicate with a program or
process on a remote node must specify information similar to the
information needed to log onto a remote system. For example, a
program may have to specify an authorized user name or account
and password in order to gain entrance to a task on a remote node.
Session Control software at the remote node verifies the information and decides whether or not to allow access. In task-to-task
communication, users must specify access-control information in

NSP/NodeB
Data-Receiving

NSP/NodeA
Data-Transmitting

1 NSP/Node A sends a Connect Initiate to NSP/Node B:
Connect

IInitiate

2 NSP INode B, having received the Connect Initiate message, returns a
Connect Confirm message. A field in the message indicates that NSP I
Node B expects segment flow control:
"[ want segmentflow control"

..-1 g~~fi~~ I!

3 NSP/Node 1\s data base has the initial value of 0 for its request count
variable for normal data segment flow control:
IFLOWrem~
4 NSP INode B sends a Data Request message containing a flow control
value that indicates the number of Data Segment messages NSP INode B
can receive. (NSP INode B executes an implementation-dependent algorithm to determine this value):
"[ can receive n segments"
~
I Data
I_Request

5 NSP INode A increments FLOWrem_dat by n:

IFLOWrem dat=n I

6 NSP I Node A executes algorithm to determine if Data Segment number
1 can be sent (highest acknowledged Data Segment message plus the
current request count must be greater than or equal to the number of the
next Data Segment message sent):
NO-can't send
YES-send

7 The answer to the above is YES, so NSP INode A sends Data Segment
number 1:

Ife~~ent 1 I

8 NSP INode B acknowledges receipt of first data segment:
Data Acknowledgment
oJsegment number 1

9 NSP/Node A subtracts 1 from its flow control request count:

IFLOWrem dat=n-ll

Figure 4.7
Segment Flow Control Shown in One
Direction on a Logical Link

111

a special parameter when they request a logical link connection
(see Chapter 5). In remote file access (see Chapter 6) and in
terminal communications, this information must be included in
the file specification. (File specifications differ for each type of
DEcnet system; the programmer's reference manual for a specific
DEC net system describes its correct file specification syntax.)
Access-con trol inform.ation may not be required for every network
function. To determine what information is needed and when,
users can consult the system manager at the remote node to which
they are seeking access. Access-control information is initially
established during installation, and can be dynamically modified
on most systems by the system manager .
. User Transparency
A network user is not aware of the many activities that take place
within systems and across links when two programs communicate. The connection between the two programs appears to be
direct and without mediation. However, if some problem occurs
in the connection, the DEcnet software notifies the user by means
of an error or status message.

Chapter 5 . Task-To-Task Communication

DEcnet enable's two programs or tasks running on different nodes
in a network to exchange data over a logical link. For example, a
program in an Rsx-llM node can use local DEcnet-Rsx facilities to
communicate with a program running in a DEC net-vAX node.
PRo/DEcnet does not provide a user interface for task-to-task
communication. This chapter may, therefore, not be relevant for
PRO/DEcnet users.

In most DEcnet implementations, performing task-to-task communication is similar to performing input/output (II 0) operations. The logical link between the two communicating programs
or tasks is like an II 0 channel over which both programs can send
and receive data. (Refer to system-specific documentation for
details on how a particular operating system implements II 0
operations. )

. DECnet Task-To-Task Calls
Programmers include DEC net calls in cooperating programs to
enable them to communicate with each other. These calls activate
routines requesting the End-to-End Communication modules in
each node to perform specific functions, such as the creation and
control of a logical link (see Chapter 4 for details on logical links ).
The form of the programmer-specified DEC net calls depends on the
programming language being used; calls can be actual calls or
macros or system directives. However, the DEcnet task-to-task
capability translates each of these call types into the same set of
DEC net messages, regardless of the language in which the calls are
programmed. For example, whether a source program issues a
connect request in FORTRAN or MACRO, the End-to-End Communication modules in the source node send the same type of message
to the End-to-End Communication modules in the target node.
Thus, when programmers write task-to-task communication calls
into communicating programs, they do not have to be concerned

113

about what programming languages the remote node supports.
They do, however, have to knO\v what languages the local node
supports, since not all operating systems support the same languages and not all languages support task-to-task communication. The programmer's reference manual for a particular DEcnet
system lists the programming languages available on that system
for task-to-task communication.
DEC net task-to-task calls in communicating programs perform
several functions, including requesting a logical link, receiving a
logical-link request, and accepting or rejecting a logical-link
request. In addition, task-to-task calls send data, receive data, send
interrupt data, and receive interrupt data. A task-to-task call also
terminates a logical link.

In some cases, a program must issue three separate calls to create a
logical link; in other cases, a program needs to issue only one call.
The number of calls required to effect any of the task-to-task
functions listed above is dependent on the operating system
running on the communicating nodes and the programming
language being used .
. Requesting a Logical Link
The first step in establishing task-to-task communication is
requesting a logical link. A programmer does so by issuing
connect request calls in accordance with the language conventions
of the source node's operating system (issuing a connect request is
the first stage in the handshake procedure described in Chapter 4).
The connect request call must include network addressing information that identifies the node and the program sending the
message and the node and the program that should receive it.
Programmers can specify this addressing information in the source
program either by building a special data area (called a connect block)
or by including an ASCII string called a network specification.

114 • Task-to-Task Communications

Building a Connect Block-Most DEC net task-to-task call sequences
include one or more call to build a connect block. A programmer
specifies parameters to the connect block call(s) that identify the
source and target programs. These parameters include all or part of
the following information:
Link identifler-This identifier differentiates the requested logical
link from any other logical link currently being used by the source
program. At any given time, a program can have multiple logical
links, either with a single program or with several programs. Each
active logical link of a program has a unique identifier. End-to-End
Communication modules in the communicating nodes use the
link identifiers to ensure that data is transmitted over the correct
logical link and arrives at its intended destination. See Chapter 4
for further details on logical-link identifiers. If the connect request
succeeds, the source program uses the link identifier to address
data to be sent over the link. The different operating systems
implementing DEC net refer to the source program's link identifier
by different names. Check the programmer?s reference manual for
a specific system to determine what the link identifier is called on
that system.
Target node identifler-A target node can be identified either by a
unique node address that distinguishes the node from all others in
the network or by a node name defined by the source node. A
Session Control module on the source node translates the target
node's name into a unique node address (see Chapter 2 for
additional information on translating node names to node
addresses) .
Object type or name-Object is another term for a network program.
A network program or object has a special identifier by which it
makes itself known to the local End-to-End Communication

115

modules. This identifier consists of an object type and/ or an object
name (see below under the section on object types and object
names for details). Figure 5.1 shows how objects are addressed.

Access-control inJormation- This information describes the source
program and includes a user identification, a password, and,
optionally, an account number. Access-control information is
equivalent to the data a user supplies when logging onto a system.
In most DEcnet implementations, this information is a factor in
the target node's decision to accept or reject the connect request.
Optional user data-A source program usually has the option of
sending 16 bytes of data to the target program as part of a logicallink connect request.
When programmers finish building the connect block using the
call(s), macro(s), or directive(s) appropriate for the operating
system and language being used, they can specify the label or
address of that connect block in the source program as a parameter
to a connect request call. DEC net requires this network addressing
information to respond to logical link requests.

Network Specifications-If programmers do not build connect
blocks to provide network addressing information for the connect
request, they must provide that information in a network specification. (Some DEcnet systems require a network specification
instead of a connect block. Refer to the documentation for a
DEC net system to find out if and how network specifications must
be used.) A network specification is an ASCII string specified in the
connect request itself. Like the connect block, it includes a target
node identifier, access-control information, and a target object
type or name. (DEcnet-20 network specifications can include
optional data.)

116 • Task-to-Task Communications

. Object Types and Names
A program in a node makes itself known to the local End-to-End
Communication software by declaring its object type and name.
In most DEC net implementations, a program must declare its
object type and name to the local End-to-End Communication
software in order to be eligible to receive link requests. (In some
implementations, a system manager can use a DEC net command at
a terminal to declare a program's object type and name. The
programmer's reference manual for a system describes how objects
are specified on the system.) The object name can be a special
network name for the program or it can be the same name by
which the program is known to the local operating system.
In a connect request, the source program uses object types and
names to specify the target program with which it wants to
communicate. Object types can take one of two formats: an object
type equal to 0 and an ASCII name, or an object type equal to a
positive integer (from 1 to 255) and a null name. The first format
identifies a program by name, and the second format identifies a
program by numeric object type.
To address a target program, a connect request specifies either a
name or an object type, but not both. The first format-object
type 0 plus name-is commonly used to address user-written
network programs. To use this format in a connect request, a
source program must know the target program's declared object
name. Note that the maximum length allowed for a name is
dictated by the operating system in which the target program
runs.
The second format-an object type equal to a positive integerprovides an abbreviated means of identifying a frequently used
network function, usually a DECnet module. A specific object type
always represents the same generic function within a network,

117

even if the program that actually performs the function has a
different name at each node. For example, the DEe net module that
performs the file access function (FAL) might have a different name
at every node, but it can always be identified by object type 17
(decimal) or 21 (octal). Thus, a source program can address the FAL
function without knowing the FAL program's name in the target
node.
Digital reserves a range of object types for DEenet system programs. Types within this range are used consistently across all
DEe net implementations to refer to the same functions. The
programmer's reference manual for each DEe net implementation
lists the object types reserved for DEenet use.
Unreserved types can be used to identify user-written network
programs. For example, in a user-written transaction-processing
application, each node might have a resident program for recording statistics on transactions performed within the last 24 hours.
An application designer could choose an unreserved object type to
identify all such programs throughout the network.
Figure 5.1 illustrates the object identifiers for several programs in
Nodes A and B .

. Accepting/Rejecting a Logical Link Request
End -to-End Communication software in the source node forwards the connect request to· the node specified in the connect
block or network specification. At the target node, End-to-End
Communication software first checks that the program addressed
in the connect request is a valid object and then verifies the source
program's identification. If the target program is a known object
and any required verification checks out, the target End-to-End
Communication software delivers the connect request to the
target program.

118· Task-to-Task Communications

NodeB

Node A

@
I

YAyandBEEareuser_written
programs identified by name
0, AY. I ~ in connect requests addressed
to them.

CEE and DEE are user-written
programs that perform the
same function in their respective nodes as part of a network
~ application. They are both
\L-_ _ _---'j
identified by object type 200 in
connect requests addressed to
them. 200 is in the range of
types reserved for customer use.
The two FAL programs are
Digital-supplied modules that
~ receive all file access requests
from remote nodes. a FAL
module is always identified by
object type 17.

Legend:

§
I

~ Network program

I ~ The object type and name by
which the local NSP module
knows the program

Figure 5.1
Addressing Network Objects

119

After examining the incoming connect request, the target program either accepts or rejects it. (See system-specific documentation for details on how each system handles connection requests.)
End-to-End Communication software in the target node forwards
the appropriate response back to the source node. The target
program usually has the option of sending 16 bytes of data along
with the acceptance or rejection of the link. For example, a
response rejecting a connection might include data that tells the
source program why the connect request was not accepted. If the
target program agrees to the connection, it specifies its own
logical-link identifier in a call accepting the connect request .

. Sending and Receiving Data
Once the target node accepts the connect request and a logical link
has been established, either progran'l can send and receive data
using the assigned logical-link identifiers as address parameters in
each message. They can send and receive data one at a time or
simultaneously.
Programs can send two types of data over a logical link: normal data
and interrupt data. Normal data comprises the subject matter of the
dialogue between the programs, and interrupt data conveys special
high-priority information.
To convey normal data over the link, a source program issues one
or more call to send the data, and a target program issues one or
more call to receive it. End-to-End Communication software at
the source node will not transmit data unless the target program
has already issued a receive call. Each receive call allocates the buffer
space that the target program needs to store the data. Figure 5.2
illustrates the various stages of this procedure.

120· Task-to-Task Communications

Some DEC net implementations allow the source End-to-End
Communication software to transmit data over any logical link, as
long as the target End-to-End Communication software has access
to enough system buffer space to hold the data. End-to-End
Communication software at the target node then delivers the data
it has received when the appropriate target program allocates its
own buffer space by issuing one or more receive call. Regardless of
the implementations involved, End-to-End Communication software in the source and target nodes exchanges link service
messages to determine whether the target program is prepared to
receive a message. This precaution is part of DEcnet's flow-control
mechanism (see Chapter 4 for details on flow control).

Source
Requests
Connection

Target Sends
Rejection
lnformation

Target Gets
Connection
Information

Target
Completes
Connection

Target
Requests
the Data

Figure 5.2
Transmitting Normal Data

Source
Transmits
the Data

Source
Disconnects
the Link

121

Interrupt data can consist of up to 16 bytes of information to be
delivered immediately to the target program. If End-to-End
Communication software in the target node has a queue of normal
data that it has already received but not yet processed, the interrupt
data is placed either at the head of the queue or in a separate queue
that the target program can access without first reading the
normal data .

. Tenninating the Link
Either program can issue a call at any time to terminate the link in
one of two ways. One method, which disconnects the link in an
orderly fashion, is normally used by a program to terminate a
session that has proceeded as expected. All transmissions are
completed before the link is dissolved. The programmer must
decide which of the two programs disconnects under normal
circumstances. The various DEC net user manuals refer to orderly
link disconnections as disconnects or synchronous disconnects.
The second method of disconnecting a link forces termination of
the link, regardless of whether or not the remote DEcnet software
has acknowledged previously transmitted data. In most programmer manuals, this method of disconnecting is called aborting the
link. When the source node's End-to-End Communication software receives notification to abort a link, it cancels all messages
waiting to be transmitted over the link. A program may choose to
abort in response to some unusual system event, such as an
impending emergency shutdown. When disconnecting or aborting a link, a program can send up to 16 bytes of data to the other
program.

Chapter 6· Rern.ote File and Record Access

Using DEcnet, a program in one node can access a file in another
node, despite differences in the operating and file systems of the
two nodes.DEcnet's remote file access capability enables userwritten programs that incorporate appropriate DEC net II 0 calls to
open and close a remote file, create or delete a remote file, and read
from or write records to a remote file.
A second application of the remote file access capability is that
people who use terminals can run a DEcnet utility or issue a system
command to manipulate remote files. From their terminals, users
can transfer (copy) files to or from remote nodes, delete remote
files, submit and execute command files at a remote node, append
files to local or remote files, list remote directories, and queue files
to a remote printer. This chapter describes program- and usermanipulation of remote files.
Since programs and users of terminals can access and manipulate
remote files, a system manager need maintain only one copy of a
file. Ensuring that a file is up to date is simpler when there is only
one copy of it, rather than several copies on several network nodes.
Like other DEC net functions, remote file access requires the
cooperation of two network programs which exchange a series of
DEC net messages. A program in one node issues a call requesting
remote file access. This call is translated into one or more DEC net
message by the local DEC net software, then sent to a DEC net
program in the remote node. The program in the remote node
receives the file access request messages and translates them into a
form recognizable to its file system. As in task-to-task communication, the two programs must establish a logical link before they
can begin communicating.

In the context of remote file access, the program that requests
remote file access is called the source program, and a DEC net system

123

program that receives the request is called the target program.
Depending on the application, the source or accessing program
can be a version of the Network File Transfer (NFT) utility (see
below for details), a system command such as the copy command,
or a user-written program that accesses remote files via calls to
DEcnet subroutines or to file access subroutines provided by the
local file system.
The target program is always a version of a DEC net utility called the
File Access Listener (FAL), whose role is to receive remote access
requests from the network. FAL completes connections initiated by
remote accessing programs and translates the incoming requests
into calls to the file system at FAL'S node. FAL then sends the
resulting file data back to the accessing program, at whose node
special routines reformat the data to make it conform to local file
structures.
File data can flow in either direction between the source program
and a target program. The file accessed can reside on a massstorage device like a disk and can be directed to an II a device such
as a lineprinter or a terminal. The actual file operations that a
source program can perform depend on the file systems local to
both nodes. The programmer reference manuals for the particular
systems involved in remote file access include details on file
operations .
. The Data Access Protocol (DAP) Interface
The File Access Listener at the target node and the accessing
program exchange Data Access Protocol (DAP) messages to perform remote file access operations. DAP resides in the Network
Application layer of the DNA architecture and uses the logical-link
services of the End-to-End Communication layer. DAP defines a set
of Inessages that controls the execution of remote file access and
outlines procedures to accomplish specific file operations. For

124 • Remote File and Record Access

instance, to create a file, an accessing program and a remote FAL
must exchange a subset of DAP messages in a prescribed sequence.
The Data Access Protocol supports the transfer of files between
heterogeneous file systems and supports sequential, relative, and
indexed file organizations. It retrieves a file from an input device
like a disk, cardreader, or terminal. It stores a file on an output
device like a magtape player, lineprinter, or terminal. It permits
sequential, random, and indexed sequential (ISAM) access of
records. DAP lists directories of remote files and supports the
deleting and renaming of these remote files. It allows multiple data
streams to be sent over a logical link, recovers from transient
errors, and reports fatal errors to the user. DAP also submits and
executes remote command files. It permits wildcard file specification for command file execution, sequential file retrieval, file
deletion, and file renaming. And it permits an optional file
checksum to ensure file integrity.
In a typical DAP dialogue, the first message exchange consists of
configuration information regarding the operating and file systems of the source and target nodes and the buffer size. File
A ttributes messages then supply information about the file to be
accessed. Following these messages is the Access Request message
to open a particular file. If data is to be transferred to the accessing
program, a data stream is set up at this point. For both sequential
and random access file transfer, one control message sets up the
data stream (DAP has a file transfer mode that eliminates the need
for DAP control messages once file data flow begins). After the file
transfer is completed, Access Complete messages terminate the data
stream. Figure 6.1 shows a DAP message exchange for sequential
file retrieval.
DAP has been designed to minimize protocol overhead. Defaults are

specified for fields whenever possible. And relatively small file
records can be blocked together and sent as one large message.

125

User Node Message
Description

Configuration Information (e.g., Buffer
Size, OS, File System,
DECnet Phase No.,
and DAP Version
Number)

Messages

Configuration
Message -----I.~
Configuration
Message

......
1------

File Characteristics
(e.g., Type, Block
Size and Record Size)

Attributes Message ~

Access Request

Access Message _
___ Attributes Message

Set up Data Stream

Request Start of Data
Transfer and declare
mode of Transfer

Request to Terminate
file access

Acknowledge
..
Message
Control (Initiate Data
Stream) Message --.Acknowledge
..
Message

Configuration
Information Returned

Actual File Characteristics Returned
File Opened

Data Stream
EstablIshed

Control (Get)
Message -----I.~

..._---1
.

Recorr 1
Record N
____ Status Message
Access Complete
Message
•
Access Complete
..
Response

Figure 6.1
DAP Message Exchange (Sequential

File Retrievalj

Remote Node
Message Description

Data Sent in Records

End of File Detected

Request Completed
Successfully

126 • Remote File and Record Access

A user program does not handle DAP messages directly. All DEC net
implementations include system software that sends and receives
DAP messages on behalf of user programs. DAP-speaking DEC net
modules use the messages listed in Table 6.1 to accomplish remote
file access and transfer .
. Programming Remote File Access
To gain access to a remote file, a user program incorporates DEC net
II 0 calls that activate remote file access subroutines. The function
of the subroutines is to build, send, and interpret DEC net file access
messages on behalf of user processes. The File Access Listener acts
on behalf of the accessing or source user program by receiving 110
requests at the target node and translating them into calls to the file
system at that node. Other file access subroutines currently
supported by DEC net implementations include:
Network File Transfor (NFT) , an interactive utility operating at the
User layer of the Digital Network Architecture. NFT interfaces to a
DAP-speaking accessing process and provides DAP functions for
people who wish to manipulate files on any node in the network
from their terminal.
Record Management Services (RMS), the standard file system for
many of Digital's operating systems. In DEcnet-vAX, DAP functions are incorporated into RMS. Remote file access functions are
therefore transparent to accessing user programs on DEcnet-vAX
systems. A program uses the same II 0 calls to access a local file as it
does to access a remote one. The only difference between the two
access requests is that, to access a remote file, a program includes a
node identifier in the specification of the RMS file to be accessed.
RMS converts the access request into equivalent DAP messages.
These are sent to a remote FAL to complete the access request. To a
DEcnet-vAX user, RMS remote and local file access are almost
identical, the only difference being the specifation of a node name
and access-control information for a remote file access.

Message

Function

Configuration

Exchanges system capability and configuration information between DAP-speaking processes. Sent
immediately after a logical link is established, this
message contains information about the operating
system, the file system, protocol version, and buffering capability.

Attributes

Provides information on how data is structured in
the file being accessed. The message contains information on file organization, data type, format,
record attributes, record length, size, and device
characteristics.

Access

Specifies the file name and type of access requested.

Control

Sends Control information to a file system and
establishes data streams.

Continue-Transfer

Allows recovery from errors. Used for retry, skip,
and abort after an error is reported.

Acknowledge

Acknowledges access commands and Control messages used to establish data streams.

Access Complete

Denotes termination of access.

Data

Transfers file data over the logical link.

Status

Returns status and information on error conditions.

Key Definition
Attributes
Extension

Specifies key definitions for indexed files.

Allocation
Attributes
Extension

Specifies the character of the allocation when creating or explicitly extending a file.

Summary Attributes Returns summary information about a file.
Extension
Date and Time
Attributes
Extension

Specifies time-related information about a file.

Protection
Attributes
Extension

Specifies file protection codes.

Name

Sends name information when renaming a file or
obtaining file directory data.

Table 6.1
DAP Messages

128· Remote File and Record Access

Network File Access Routines (NFARs), a set of FORTRAN-callable
subroutines that become part of the user process. NFARS cooperate
with a remote FAL, via DAP, to access remote files for user
applications. RSX-DEcnet uses NFARS to provide remote file access
functions.
VAX/VMS Command Language Interpreter, which enables users of
terminals on VAX/VMS systems to enter VMS commands interactively, in order to access and manipulate remote files. VMS commands pertaining to remote file access and manipulation interface
with RMS to provide network-wide access. No separate NFT utility
is therefore required in DEC net-vAX systems.
Network Management Modules use DAP services to obtain files for
downline-Ioading other remote nodes and to transfer upline
dumps for storage.
Figure 6.2 shows the location of the various file access subroutines
in the Digital Network Architecture.
Regardless of the file access subroutine that is used, all DEcnet
systems allow people who use terminals and user programs to
open an existing remote file, create a new file on a remote node,
read records from a file, write records to a file, close a file, and
delete a file. Besides this, specific DEC net systems allow user
programs to manipulate remote files in other ways. The programmer's reference manual for each system discusses what operations
are available and how to specify DEC net calls for remote file access.
The rest of this chapter discusses aspects of remote file access that
apply to all DEC net implementations.

File System Capabilities
A programmer needs to be familiar with the file system resident at
the target node. File organization, access modes, and other

129

Local Node

Remote Node

User Layer
The accessing program:
NFT Utility
A VAX/VMS Command
A user-written Program

Local
File
System
(e.g., FCS
orRMS)

Network Application
Layer
DAP-speaking routines:
NFARs, or Routines
in the local file
system (e.g., RMS)

Network Application
Layer
DAP-speaking server:
File
Access
Listener

Session Control Layer

End-to-End
Communication Layer

Routing Layer

Data Link Layer

Physical Link Layer

Remote
File
System
(e.g., FCS
or RMS)

.......

DECnet Network

Legend:
NFT
FCS
RMS
NFARs
DAP

Network File Transfer Utility
File Control System
Record Management Services
Network File Access Routines
Data Access Protocol

Router

-z..-

..-. --.

Figure 6.2
File Transftr and DECnet File Access
Subroutines

DECnet node performing Switching
A Physical Communication Link
A path through the network
DAP message flow

130 • Remote File and Record Access

characteristics are dependent on the type of file system that each
operating system supports. When the local and remote nodes have
the same operating and file systems, programming remote access is
similar to programming local I/O operations. However, when the
access operation bridges different types of operating systems, the
programmer faces certain variables and restrictions.
Generally, the theory of the lowest common denominator applies.
A source program can perform those functions that its source
language provides and that the remote file system supports. For
example, a program on a VAX/VMS node cannot use all the remote
access functions provided by DEC net-vAX if the target node runs
Rsx-llM.

By comparing the file system characteristics with the available
remote access calls, a programmer can find out the kinds of file
operations that are possible between two different DEC net implementations. This information is available in the programmer's
reference manual for each DEC net system.
Initiating ReDlote Access

Like task-to-task communication, remote file access requires a
handshake sequence at the beginning of the operation. Not only
do NSP modules in the End-to-End Communication layer of the
source and target nodes set up a logical link bet'ween the source and
target programs, they also enable the exchange of DAP file access
messages to prepare for the file operation that will be performed
over the link.
This extended handshake, which is transparent to the source
program, occurs automatically when the program issues a call to
open a remote file. The form of the call to open a file varies from
system to system and from language to language, but the call

131

always provides much of the information exchanged in the
handshake. Using the information supplied by the call, as well as
system-supplied data about the local file system, the remote file
access subroutines (described above) generate DAP file access
messages addressed to the remote FAL. In response, FAL, the target
program, sends back DAP messages to define characteristics of its
local file system.
For most remote access operations, the call to open that the source
program issues contains a file specification, access-control information, and characteristics of thefile to be accessed. Through the
call, the source program passes this information to the subroutines
in the local file system. Depending on the open call's particular
function, the call passes additional information. For example, the
call might specify if a file is to be opened for reading or opened for
appending.
The File Specification
The file specification identifies the remote file to be accessed.
Because the remote file system actually carries out the requested
file operation, the programmer must know how the file is
identified by users on the node where the file resides. Refer to the
programmer's reference manual for the source and target systems
for file specification syntax.
Access-Control InforInation
Access-control information identifies the accessing program to the
remote system. It consists of a user identification code or name,
a password associated with the user identification, and additional
accounting information required by the remote system. If this
information matches an account or guest account entry in the
remote system's user file, the program gains access to that system's
resources. As in specifications of remote files, the syntax of the
access-control information must be recognizable to the remote
system.

132· Remote File and Record Access

Gaining access to the remote system does not guarantee that
requested file operations will succeed. In most Digital operating
systems, each file has a protection code that determines the types
of access allowed to defined groups of users. The user
identification- a code or a name-specified by the accessing
program determines the program's group category and, therefore,
determines the types of access it can make to each file.
File Characteristics
The characteristics of the file to be accessed include access method,
file organization, record attributes, data type, and record format.

Access Method (sequential or random): The source program indicates how it will access the file. All Digital file systems support
sequential access, the method in which the records in a file are read
or written one after the other. Selected file systems permit random
access, which allows the source program to access a specific record
without having to read all the records that come before it.
File Organization (sequential, relative, or indexed): Files that are
organized sequentially have records arranged one after the other.
All types of storage devices support sequential file organization. In
files that are organized by the relative method, records within a file
are identified by a relative record number. This number identifies
the record's position relative to the beginning of the file. Indexed
organization (available for RMS files only) is a complex file structure
that allows both sequential and random access and uses record keys
for identification (using a record key to retrieve a record from an
indexed file is similar to consulting the index in a book to find the
pages that discuss a particular subject). The keys used to identify
individual records are defined at file creation. Only disk devices
support relative and indexed file organization.
Record Attributes-This characteristic indicates the type of ~ertical
format control that applies to the file.

133

Data Type (ASCII or image): Depending on the data's record
attributes, DEC net software converts the format of ASCII data.
Image data is a stream of bits to which the DEC net software applies
no interpretation.

Record format (fixed-length, stream, variable-length, or variablewith-fixed length control-vFc): This characteristic indicates the
manner in which records are formatted within the file. A VFC
record supported by RMS files only includes a fixed-length control
field in addition to the variable-length data portion .
• Accessing Remote Files From a Terminal
All DEC net implementations support some kind of terminal-based
access to remote files. In some DEC net implementations, system
commands, such as COPY, TYPE, and APPEND, enable remote file
access from a terminal. Other DEC net implementations use the
Network File Transfer (NFT) utility. The programmer's reference
manual for a specific DEC net system tells which access method the
system supports for terminal-based file access.
All DEcn'et implementations allow people using terminals to
perform several operations on remote files. Users can transfer
(copy) a file to or from a remote node or between two remote
nodes. They can delete a remote file, and they can submit a local
command file for execution at a remote node or execute a
command file already existing at a remote node (command files
cannot be submitted to or executed at a DEcnet-RT node). In
addition, users can append one or more local or remote file to an
existing local or remote file (on DEcnet-20 nodes, supported only
for R]E-20 stations). Users of terminals can also obtain listings of
remote directories (a directory file lists all the files that reside on a
device and that belong to a specific user or category). Further,
users can queue one or more file toa lineprinter. The files can be
remote and the printer local, or the files can be local and the printer
remote (on DEcnet-20 nodes, this operation is supported only for
RJE-20 stations).

134 • Remote Pile and Record Access

Implementations of DEC net that support system commands rather
than NFT also enable users to create, open, and close remote files as
well as read to and write from them. Check system-specific
documentation for details.
On behalf of the person using the terminal, the access method (NFT
or system commands) creates a logical link between itself and the
remote File Access Listener. From the input a user types at the
terminal, the access method formulates the appropriate Data
Access Protocol messages. End-to-End Communication software
at the source node then sends these over the link. In turn, FAL at the
target node interprets the messages it receives and interfaces with
its local file system, in accordance with the terminal user's request.
FAL then returns information and any requested file data to the
source program (NFT or system commands) by sending back DAP
nlessages. See Chapter 9 for examples of this procedure in an
applications environment.
Access Control
To access a remote file, a person using a terminal must supply the
same access-control information that an accessing program must
include. This information-user identification, password, and
optional account data-must be available whenever a file access
command refers to a remote node. Refer to system-specific
documentation to see how a system handles access information in
terminal-based file access.
File Protection
The access-control information that a user at a terminal supplies at
the local node determines the user's access rights at the remote
node. The remote file system compares the user's identification
with the protection code associated with the file to be accessed.
The file system carries out a requested access only if the protection
code for the file grants access to the identified user.

135

Refl10te File Specifications
A specification that describes a file on a remote node must conform
to the remote node's syntax rules. If the remote node has a
different operating system than the source node, the remote file
specification may contain fields or conventions that the source
node does not recognize. To prevent the source node from being
confused by a foreign syntax, some implementations require the
user to enclose the foreign specification in double quotes; some do
not. Refer to the programmer's reference manual for a system for
specific guidelines.
Refl10te COfl1fl1and File Subfl1ission
Most terminal-based file access implementations allow a terminal
user to execute a command file in a remote system. A command file
contains ASCII command lines equivalent to the command lines a
user enters at a terminal. The remote node reads and executes these
commands when a user submits the file. See system-specific
documentation for details on how a system implements this
capability.

Remote command file submission and execution are valuable
means of using the resources of a remote system. At the local node,
a user can run a text editor to create an ASCII file consisting of
commands that conform to the syntax of the remote node's
command language. Subsequently, the user can invoke the NFT
utility (or system commands) to copy the file to the remote node
for execution there.

Chapter 7 • Network Terminal Facilities

Digital offers several terminal facilities that enable users at one
node to communicate interactively with remote nodes and with
terminal users at remote nodes. These facilities include:
• Interactive terminal-to-terminal communication
• Network virtual terminal, allowing direct access to a remote
node's operating system
• Remote file access from a terminal
A fourth terminal facility, the Local Area Transport (LAT) protocol,
enables terminals connected through a Terminal Server to an
Ethernet LAN to establish links with any host on the same
Ethernet. In addition, the terminals can connect to remote hosts
not on the same Ethernet via an intervening DEC net host on the
Ethernet. LAT is not a DEC net protocol.
This chapter discusses terminal-to-terminal communication,
direct access to a remote node's operating system, and the LAT
protocol. For information on how to access remote files from a
terminal, see Chapter 6.
· Interactive Terminal-To-Terminal ComInunications
Users at two DEC net terminals can communicate interactively
through the Phone or TLK utilities:
Phone-.:..Phone is a utility of the VAX/VMS operating system. DEcnet
can operate on Phone to allow a terminal-to-terminal conversation
over the network between two DEC net-vAX terminal users. Phone
allows a terminal user to "dial" another terminal user on the same
or on a different DEcnet-vAX node. In a DIAL command, users
specify the name of the target node and the name of the person
with whom they wish to talk. If that person is logged on and
available, the Phone utility activates the bell function on the
remote terminal and displays messages on the terminal screen to
notify the person of the awaiting phone call. The person being

137

called can then elect to answer the call by issuing an ANSWER
command. Once the call has been answered, the two people can
engage in typed conservation by entering messages at their
terminals. At the end of the conversation, either user can issue a
command to hang up, thereby disconnecting the connection at
both ends.
Phone offers several options that can be implemented to provide
terminal users with additional services and information. For
example, there is a directory service that lists available users on each
DEC net-vAX system supporting Phone.
Terminal users on two nodes can communicate via Phone only if
both nodes support the utility. Nodes supporting different utilities cannot communicate at this interactive level.
TLK-TLK is a DEe net utility supported by DEcnet-RSX. It enables
terminal users on two DEcnet-Rsx systems to communicate.
Through TLK, users can send one-way messages contained in one
line (single-line mode) to another terminal user or can engage in an
interactive conversation (dialogue mode). To send a message in
either mode, users issue a TLK command specifying a node name
and a terminal identifier for the terminal with which they wish to
"talk." The user at that terminal receives a copy of the message sent,
along with the identity of the sending terminal. All mess~ges
received are displayed on the terminal screen. In dialogue mode,
TLK establishes an interactive connection with the target terminal
so that a user there can respond instantly to the TLK messages
received.

A single-line-mode message is just one line long, and it's terminated when the sender types a carriage return to transmit the
message. To end a "conversation" of dialogue-mode messages,
either user can type a [CTRL/z] or EXIT command.

138· Network Terminal Facilities

A system manager can use TLK to send messages read from a TLK
command file. TLK command files are useful for sending many
messages at once and for storing and sending sets of messages that
need to be sent more than once (reminders to users of routine file
backup procedures at the end of every day, for example).
All DEcnet systems do not support TLK. Refer to system-specific
user documentation (accompanying a system) to see if the system
supports TLK and, ifit does, for details on how to use TLK. Refer to
Chapter 9 for an example of interactive terminal communications
in an applications environment.

. Network Virtual Terminal Facilities
All DEC net systems, with the exception of DEcnet-RT, have a utility
that logically connects a local terminal to a remote node. With
DEC net Phase III, such connections are possible only if the local
node runs the same operating system as the remote node. A
DEcnet-vAX terminal user could use the SET HOST terminal facility
to log on to another DECnet-vAX node; a DEcnet/E terminal user
could communicate with a remote DEcnet/E node using the NET
utility. Intermediate routing nodes do not have to run the same
operating system as the nodes involved in the terminal-to-remotenode connection.
With DEC net Phase IV, logical-terminal-to-remote-node connections can be established even if the local and remote nodes are
running different operating systems. The Phase IV utility enabling this is called network virtual terminal (NVT). NVT also
enables terminals that are physically connected to a host on an
Ethernet to connect logically to any other host node on the
same Ethernet and to remote host systems (by using the CTERM
protocol).

139

Programmers are the primary beneficiaries of the network
command-terminal facilities. If a node does not have adequate
system and storage resources to support program development,
programmers at that node can log onto a remote node that does
have the adequate resources. They can then develop programs
interactively, as if they were local users of that remote node. For
exam pie, programmers on a DEcnet-11S node would access a larger
remote node, since Rsx-lls nodes do not have any mass storage.
In Phase IV networks, NVT enables programmers to use the
resources of any system in the network. Depending on the
applications they are developing, programmers could log onto to
the remote nodes best suited to their needs, to develop their
programs with increased efficiency.
Terminal users at a node do not see the complex activities that
DEcnet performs to make it possible for them to log onto a remote
node. All that a user needs to know to gain access to a remote
node's operating system is its node name. A simple login command gives the user access to the resources of the remote node.
Refer to the user documentation accompanying a system for
details on how to use available network command terminal
utilities. Although different DEcnet systems support different
network virtual terminal utilities, all the utilities allow a terminal
user issuing appropriate commands to log onto a specified remote
node and perform most of the functions that the remote node
allows its local users to perform (see Figure 7.1 for an example).
Network Virtual Ter:minal
The DNA layer responsible for providing NVT services is Network
Application, the layer above the Session Control layer and immediately below the User layer. In the Network Application layer, the

140· Network Terminal Facilities

Figure 7.1
Remote Terminal Processing

protocols effecting virtual terminal communications are organized in two sublayers:
• terminal communication module, which controls the connections
between applications and terminals .
• command terminal module, which enables terminals to communicate
with Command Language Processors and application programs.

In an Ethernet local area network, NVT enables terminals to
connect logically to any host node on the same Ethernet. In a wide
area network, terminals on a node can use NVT to connect logically
to any remote node in the same network. NVT protocols provide
the following addi'tional services:

141

• Distribution of terminal-handling functions between the two
communicating systems.
• Management, at the operating-system level, of terminal input/
output functions and terminal characteristics. The description of
the command-terminal protocol (see below) includes a list of these
functions.
• Offering standard terminal services, and featuring good performance and device independence. Optionally, offering methods of
controlling the terminal's behavior in considerable detail.
• Support of high-availability implementations. The protocols contain functions that allow the restart of interruptedcommunications.
Figure 7.2 depicts how the NVT modules are organized within
DNA.

Terminal Communication Protocol-This protocol is the Foundation
layer, the lower of the two sublayers, of the NVT service, and is the
base protocol for NVT. The terminal communication protocol is
responsible for making and breaking connections between applications and terminals. It extends DNA Session Control layer
services by establishing logical links between endpoints that are
specific to terminal services. The endpoint in the host system is
called a portal; the other end, in the server system, is called a logical
terminal.
In the context of the NVT function, the host node is the node at
which the terminal attempting to make the connection is located;
the server system is the remote node with which the terminal
establishes the connection.

142· Network Terminal Facilities

Host Node

Server Node

User Layer
Command Processor
(e.g., DCL)
User application

Server
Terminal
Services

Terminal

Network Application
Layer
Command
Terminal Module

Terminal Communication Module

Session Control Layer

End-to-End
Communication Layer

Routing Layer

Data Link Layer

Physical Link Layer

Host
Terminal
Services

Legend:
Router

-r...-

DECnet node performing Switching
A Physical Communication Link

Figure 7.2
Network Virtual Terminal Service

.-.-+-

A path through the network
NVT message flow

143

A portal corresponds to a remote terminal identifier in the host,
and a connection binds that identifier to an actual terminal. The
NVT connection is therefore referred to as a binding.
In the future, it may become desirable to add new modules to the
Network Application layer as alternatives to the commandterminal module. These modules might require their own protocols. In anticipation of such developments, the current
implementation of the NVT modules has been designed so that
portal, logical terminal, and binding share a state called a mode. In
the future, mode management will enable the terminal communication protocol to select one of the several available higher-level NVT
modules and protocols (currently, the only available higher-level
module is the command-terminal module) Table 7.1 shows the
terminal communication protocol messages and their functions.
Command Terminal Protocol-The functions of this protocol primarly involve command-line input and output. They are general
enough to support a broad range of video and hardcopy terminal
applications. The input/output functions include:

• Reading a line of input. Line terminators can be specified individually. Although the server system is responsible for echoing and
operator editing, the host software has considerable control over
these features, including the enabling and disabling of each
feature.
· Accepting input even if the host program has not issued a read
request and saving input until a read request arrives from the host
program (this capability is known as typeahead).
· Inputting, or taking action in response to, certain characters
immediately as the keys are struck (out-ofband character processing). These characters can be specified individually by users.
• Writing a string of output characters. Writing can proceed
concurrently with reading, subject to certain synchronization
options. The terminal operator can cause output to be discarded.

144 • Network Terminal Facilities

Message

Function

Bind Request

Requests a Binding; identifies version and type of
sending system.

Rebind Request

Requests a rebinding (reestablishes broken communications, for high availability implementations).

Unbind

Requests that a binding be released.

Bind Accept

Accepts a Bind request.

Enter Mode

Requests entry of a new mode (the only mode
currently defined is command mode). This selects
the command terminal protocol as the higher level
protocol.

Exit Mode

Requests that the current mode be exited.

Confirm Mode

Confirms the entry of a new mode.

No Mode

Indicates that the requested mode is not available or
confirms an exit mode request.

Data

Carries Data (i.e. command terminal protocol
information) .

Table 7.1
Terminal Communication Protocol
Messages

· Recognizing ANSI standard escape sequences on input and
output.
· Cancelling a read request.
• Reading and setting terminal device characteristics.
• Determining how many characters there are in the typeahead and
input buffers combined.
• Clearing the typeahead and input buffers.

145

Message

Function

Initiate

Carries initialization information, as well as protocol and implementation version numbers.

Start Read

Requests that a READ be issued to the terminal.

Read Data

Carries input data from terminal on completion of
a read request.

Out-of-Band

Carries out-of-band input data.

Unread

Cancels a prior read request.

Clear Input

Requests that the input and typeahead buffers be
cleared.

Write

Requests the output of data to the terminal.

Write Complete

Carries write completion status.

Discard State

Carries a change to the output discard state due to a
terminal operator request (via an entered outputdiscard character).

Read Characteristics

Requests terminal characteristics.

Characteristics

Carries terminal characteristics.

Check Input

Requests input count (number of characters in the
typeahead and input buffers combined).

Input Count

Carries input count as requested with Check Input.

Input State

Indicates a change from zero to non-zero or viceversa in the number of characters in the input and
typeahead buffers combined.

Table 7.2
Command Terminal Protocol Messages

Table 7.2 shows the command-terminal protocol messages and
their functions.

146 • Network Terminal Facilities

Network Virtual Terminal Operation
In a typical network command-terminal session, a terminal
management module in a server system with an active terminal
requests a binding to some host system. When the terminal
operator issues a SET HOST command, the terminal management
module invokes a terminal communication services function.

The terminal communication services module initiates a logical
link to its counterpart in the host system, using DNA Session
Control layer services. On discovering the incoming logical link,
the host module allocates a portal, thereby beginning the formation of a binding. The host module then accepts the logical link.
Once the link has been formed, the server module sends a Bind
Request message to the host. A terminal management module in
the host, acting on behalf of the host's login process, recognizes
the binding request and causes the terminal communication
services module to send a Bind Accept message.
Once the biQding has been formed, the host takes the initiative and
prepares the binding for the command terminal protocol (in other
words, enters command mode) with a further exchange of
messages. This process takes place in connection with the first
terminal I/O request from the login process. The respective
protocol modules in the host· and server can now begin speaking
the command terminal protocol by sending an Initiate message to
one another. After initialization, the dialogue of remote terminal
service requests and responses ensues.
Terminal service requests normally originate with an application
program in the host system. The application program issues
requests to the host operating-system terminal services. Host
terminal services issue corresponding requests to the host protocol
module. The server protocol module reproduces those requests
remotely and reissues them to the server-terminal services.

147

Termination of the session can occur in a number of ways. A
typical, orderly shutdown begins with a logout by the application
in the host. As a result, the host terminal communication services
module sends an Unbind request. The server responds by releasing
its resources. Finally, the host disconnects the logical link. Figure
7.3 shows an NVT message exchange.

Host Node Message
Description

Accept Binding
Establish Command
Mode

Initialize command
terminal protocol

Messages

Bind Request
Message
Bind Accept
Message ------I~~

Server Node
Message Description

......
1------

Request a binding
with a specified Host

Enter Mode
Message
~
Confirm Mode
..
Message

Server enters commandmode

Initiate Message -----.____ Initiate Message

Initialize command
terminal protocol

Start Read Message~
Read Data
..
Message

Accept and process a
line of input

Out-ofband
Message

User types an out-ofband character

Begin terminal service
dialog
Issue Read request

Host terminal software notified of outof-band character

....

Dialog continues

Dialog continues
Application logs off.
Release host resources

Unbind Message ~

Figure 7.3
NVT Protocol Message Exchange

Release server
resources

148· Network Terminal Facilities

. Local Area Transport (LAT) Protocol
LAT is an Ethernet-based virtual circuit protocol that does not use
DEC net as a message transport facility. A Terminal Server on an
Ethernet LAN implements the LAT protocol to enable terminals that
are connected to it to establish a logical link with any host node on
the same Ethernet. The host node must implement the host side of
the LAT protocol. Figure 7.4 shows the LAT protocol implementation in an Ethernet local area network.
The terminal server and a host node on the Ethernet exchange LAT
messages at regular intervals dictated by the circuit timer, which is
a parameter that a network manager can set. Characters that a user
inputs at a terminal are buffered until the circuit timer expires.

VAX/VMS Host

Application
Terminal Class Driver

Ethernet Coaxial Cable

Figure 7.4
LAY Protocol Implementation

149

When this happens, a message containing the buffered characters
is transmitted to the host.
The actions of the Terminal Server are transparent to users. With
simple commands, users can establish establish logical connections
(sessions) with host nodes and proceed with their work. Terminal
Server software enables a user to maintain several sessions at one
time. With a user-defined switch character on the terminal
keyboard, the user can move quickly from one session to another
without repeating the login dialogue each time. One of the
primary advantages of the Terminal Server is that it frees host
nodes from terminal processing, thereby enabling them to process
user applications more efficiently.
Terminal Server software enables the Terminal Server to multiplex
access requests from many terminals into one message for each
host node on the Ethernet, thereby virtually eliminating blocking
(a problem with terminals that are physically connected to host
nodes, wherein the number of terminal access requests exceeds the
number of host ports).

Relationship Between LAT and CTERM
To establish a logical link with a remote DEC net node that is not on
the Ethernet LAN a terminal connected to a terminal server must go
through an intervening DEcnet host node on the Ethernet. This
host node converts the LAT protocol that operates over the
Ethernet link to the NVT CTERM protocol that operates over the
DEC net link between the host and remote nodes. See Figure 7.5 for
an illustration of the relationship between the LAT and CTERM
protocols. The dotted line in the figure indicates the communication path established between a terminal on the terminal server and
the remote DEcnet node.

150· Network Terminal Facilities

Node

TERM
I

: CTERM
.r-;----r-----------~--_r:,_--~----------~----,:r---~~

Ethernet Coaxial Cable

I
I
I

Figure 7.5
LAY and CYERM Operation

Chapter 8· Internetwork Communications

Systems in a DEC net network can communicate with systems in
non-DEC net networks. Specifically, DEe net nodes can exchange
data and share resources with DEcnet and non-DEc net nodes over
an x.2S packet-switched data network (PSDN) and with systems in
an IBM SNA network.
A DEcnet node can communicate over a PSDN with other DEC net
and non-DEcnet systems if it has been installed either with
Packetnet System Interface (PSI) software or with the x.2S/x.29
extension package that enables the DEC net node to be connected
logically to a DEC net Router/x.2S Gateway. (See below for details
on these additional software components.)
. X.2S Communications

x.2S is a recommendation of the Comite COllsultatif Intertl:1tioll:l1
Telephonique et Telegraphique (CCITT), which defines :1 stand:1rd
means for computers to interface with packet-switched data
networks. A PSDN is a data comnlllllic:1tions service offered hy
common carriers, such as thc Postal 'Idephonc and '1elegraph
Authorities (PTT). Packet-switched data networks are widespread,
with their number increasing every year. An organization can
realize substantial cost savings by implementing network configurations in which widely dispersed systems communicate over a
PSDN (the tariffs for data communications over a PSDN are based on
volume of data sent rather than on connect time or distance
between communicating systems).
It is mandatory for all public packet-switched data networks to use
the x.2S interface. Private carriers, while not required to do so, can
choose to run a network service using x.2S standards.

A packet-switched data network is one that receives addressed
packets of data from network users and conveys them through
internal switching routines to specified receivers. The way the

152· Internetwork Communications

network delivers the packets is completely transparent to the end
user who has no influence over the path the packet takes through
the network. The x.2S recommendation specifies the manner in
which the data packets get from the user to the PSDN.
The x.2S recommendation defines standards that govern the
relationship between users and the PSDN network on the following
levels:

Levell, the physical level- This level defines the mechanical, electrical, functional, and procedural characteristics of the physical link
between the user's equipment and the PSDN equipment. In DNA,
Levell resides in the Physical layer.
L{'I/('/2, theIrame /evc/- This level defines the link-access procedure
for data exchange over the communications link between the user
and the PSDN. Level 2 resides in the Data Link layer of DNA.
l.e·I'el .1, tIlt' racket IClJcl- This level defines the packet format and
procedures for the exchange of packets containing control information and user data between the user and the PSDN. Like Level 2,
this level resides in DNA'S Data Link layer.

In accordance with the x.2S recommendation, a header containing control and destination information is included as part of each
data packet submitted to the PSDN network. (In DEcnet implementations, x.2S software, described below, creates this header.) The
header information identifies to the PSDN the data sender and the
data receiver. Thus, regardless of the kind of system or language
used to generate the data message, the PSDN recognizes the header
information and routes the packet to its proper destination.
The PSDN interleaves packets from many users over its transmission
lines, thereby providing a fast and economical communications

153

service. The network differentiates between multiple packets on a
line by means of a numbering scheme that ensures that the packets
get sent to the proper destination in the correct order .

. DTEs and DCEs
Each PSDN consists of a number of geographically separated
switching nodes that are connected by high-speed links. When an
organization leases a circuit from the PTT or other carrier, the
circuit physically connects a computer in the organization to one
of the PSDN switching nodes. The PSDN switching nodes are called
network interfaces or data circuit terminating equipment (DCE).
User computers or terminals connected to DCES are called data
terminal equipment (DTE). Computers and block-mode terminals
are packet-mode DYEs. Asynchronous terminals are start-stop mode
DYEs. Figure 8.1 illustrates these components.

X.2S Circuits
Two DTES (for example, system A and system B, with which A
wants to communicate over the PSDN) communicate by means of
one or more virtual circuit. A virtual circuit is a permanent or
temporary logical association set up by the PSDN. A logical channel is
an association between a DTE and its DCE for a given virtual circuit.
Each virtual circuit handles the exchange of data between two
specific DTES. The DTE at each end of a circuit assigns a logical
channel number. This number is allocated independently at each
end of the circuit (at each DTE-DCE interface). When sending data,
a DTE includes a logical channel number to identify the channel and
the corresponding circuit to which the data belongs.
When a DTE uses more than one virtual circuit at a time (for
example, multiple circuits to the same destination or circuits to
several different destinations), the circuits are multiplexed over the
physical link between the DTE and the DCE.

154· Internetwork Communications

Packet
modeDTE

Packet
modeDTE
Start-stop
modeDTE

PSDN = Packet Switched Data Network
DeE = Data circuit-terminating equipment,
a PSDN switching node
DTE = Data terminal equipment, a user computer (packet mode)
or a terminal using the PSDN (start-stop mode)

Figure 8.1
Components oja Packet-Switched Data
Network

Virtual circuits are either permanent or temporary (switched). A
permanent virtual circuit (PVC) is analogous to a leased line between a
local and a remote DTE. Either DTE can send data over the pvc at any
time without issuing calls to set up or break the circuit. When a
user subscribes to a PSDN, the administrators of the PSDN allocate
the logical channel number for the pvc. A temporary association
between two DTES is called a switched virtual circuit (SVC). A DTE sets

155

up this type of circuit only when it wants to send data. The
sending DTE assigns a channel, identifies the target DTE, and then
obtains that DTE'S agreement to communicate. When the DTES
have finished exchanging data, one or the other initiates a clearing
sequence to terminate the SVC.
If a system uses an x.2S native-mode user interface, discussed
below, under x.2S Access Methods, application programs running
on the system must issue calls to specify or create a virtual circuit
with a remote DTE. If the system uses data link mapping (DLM)
mode, also described under x.2S Access Methods, users need merely
include the node and task name of the destination DTE; the DEC net
routing mechanism handles the selection of circuits as it would
over any DEcnet-to-DEcnet connection. (With DLM, the user is not
aware that the connection with the destination node is occurring
over a PSDN.)
X.25 Gateway Access Protocol
This protocol, which resides in DNA'S Network Application layer,
permits user-written programs in a DEC net node to communicate
with user programs in a non-DEc net system across an x.2S-based
PSDN. The communicating DEC net user program does not have to
reside in the same DEcnet node that is physically connected to the
PSDN.

Table 8.1 lists the messages that the x.2S Gateway Access protocol
uses to accomplish remote access to the PSDN.
Gateway Access Operation-Three processes are involved in the x.2S
Gateway Access operation: the user process, the x.2S Gateway
Server process, and the foreign x.2S DTE process. The user process
accesses x.2S packet-level functions by issuing calls to the x.2S
Gateway Access module (see below, under x.2S Gateway Access
Modules). These calls are translated into x.2S Gateway Access

Message

Function

Open

Requests the use of a permanent virtual circuit.

Open Accept

Accepts a permanent virtual circuit Open request
and allocates the PVC to the Gateway Access user.

Open Reject

Rejects a Permanent Virtual Circuit Open request,
refusing allocation of the PVC to the Gateway
Access user.

Outgoing Call

Requests that an outgoing Virtual Call be placed by
issuing a Call Request Packet to the X. 25 network.

Call Reject

Indicates rejection of an outgoing or incoming
virtual call for lack of resources.

Incoming Accept

Indicates acceptance of an outgoing virtual call.

Incoming Call

Indicates an incoming virtual call has been received.

Outgoing Accept

Accepts a previously received incoming virtual call.

Clear Request

Requests that a virtual call be Cleared.

Clear Indication

Indicates a clear request packet has been received
from the X.25 network.

Clear Confirm

Indicates a clear confirmation packet has been received from the X.25 network.

Reset Request

Requests that a virtual circuit be reset.

Reset Indication

Indicates that a reset indication packet has been
received from the X. 25 network.

Reset Confirmation

Indicates a reset confirmation by either the user or
the X.25 network.

Reset Marker

Marks the place in the stream of Data messages
where a reset occurred.

Data

Contains Data packets outbound from or inbound
to a user.

Interrupt

Contains Interrupt packet data outbound from or
inbound to a user.

Interrupt
Confirmation

Confirms the receipt of an Interrupt message (used
for flow control).

No Com

Indicates that a PVC entered a no-communication
state.

No Com Seen

Indicates that the user of a PVC acknowledges a No
Com message, and thus has been notified that there
has been one or more failures of the X.25 network
(e. g., Restarts).

Table 8.1
X.25 Gateway Access Protocol

157

Protocol messages and transmitted over a DEcnet logical link to the
x.2S Gateway Server process. The x.2S Gateway Server (any node
that functions as the DTE to a PSDN) transmits x.2S packets to and
receives x.2S packets from the PSDN. These packets flow over an
x.2S virtual circuit to the foreign x.2S DTE process.
In a typical Gateway Access Protocol dialogue, the first message is
usually an Open Message requesting that a pvc be opened, or an
Outgoing Call message indicating that the user process wishes to
place a virtual call. In the case of a virtual call, if the foreign DTE
accepts the call and issues a Call Accept packet, the user process
receives an Incoming Accept message. The user process can then
exchange data with the foreign DTE process by sending and
receiving Data and Interrupt messages. When the user process
wishes to terminate the virtual call, a Clear Request message is
sent to the x.2S Gateway Server, which clears the call by issuing a
Clear Request packet to the x.2S network. The user process is
then informed of the termination of the call through a Clear
Confirm message.
Figure 8.2 shows x.2S Gateway Access Protocol exchange for
switched virtual circuit operation, where the virtual call is initiated by the user process on the DEC net system.

X.2S Gateway Access Modules
DEC net utilities that provide the x.2S Gateway Access serVIce
include:
X.25 Gateway Access Module-This module receives user x.2S

requests and communicates with the DEC net DTE that is directly
connected to the x.2S PSDN.
X.25 Gateway Server Module-This module resides in the DEcnet DTE
(also called the gateway system), communicates with the x.2S

Gateway Access module over a DEcnet logical link, and acts on

158· Internetwork Communications

User Node Message
Description

Request to place a
Virtual Call

Gateway Node
Message Description

Messages

Outgoing Call
Message -----I.~

Incoming Accept
......
1 - - - - - - - Message
Request to send a Data
Packet

Data Messqge ~

Gateway Server process issues Call Request packet to X.2S
network
Call Accept Packet
received from X.2S
network
Data Packet sent on
virtual circuit

Data Message ~
User receives Data
sent by foreign DTE
process

~ Data Message

Data packet received
from virtual circuit

User receives Interrupt Data

Interrupt
......___- - - - Message

Interrupt Packet received from virtual
circuit

Interrupt Confirmation
Message
..
Request to terminate
(clear) Virtual Call

Virtual Call is
terminated

Clear Request
Message

Clear Confirm
Message

......
1------

Figure 8.2
X.25 Gateway Access Message
Exchange

•

Clear Packet issued to
X.2S network
Clear Confirmation
Packet received from
X.2S network

159

requests made by the x.25 Gateway Access module and the PSDN.
This module uses the x.25 packet-level module to gain access to the
PSDN.

X.25 Packet-Level Module-This module resides in the Data Link
layer of the gateway system and acts as the DTE on the x. 25 PSDN. It
uses the x.2S frame-level module for its connection to the x.2S
DCE. The DTES at either end of a virtual circuit access the circuit by
exchanging x.25 Level 3 packets with their local DeES.

Table 8.2 summarizes the x.25 Level 3 packet messages and their
functions.

In response to a user request, the x.25 packet-level module initiates
the establishment of a virtual circuit to the specified DTE. The
module assigns a logical channel number and transmits a Call
Request packet, specifying the logical channel number and the
address of the remote DTE.
Once the remote DTE has accepted the call and a virtual circuit has
been established, data packets can be transferred. Each data packet
is numbered sequentially. A Send Sequence Number carried in
each data packet specifies the position of the packet in the
sequential stream. A Receive Sequence Number from the receiving DTE acknowledges received packets and authorizes the transmission of additional packets. This number is carried in Data
packets, Receive Ready packets, and Receive Not Ready packets.
The flow-control mechanism (see Chapter 4) operates separately for
each direction of data transfer on a logical channel. The mechanism is based on the concept of a window, which is a range of
packets authorized to be transmitted across the DTE/DCE interface.

160· Internetwork Communications

Packet Type

Direction

Description

Call Request

DTE toDCE

Assigns a logical channel to the DCE and
establishes a virtual circuit to the DTE
addressed in the packet.

Incoming Call

------

DCEtoDTE

Indicates that a remote DTE wishes to
establish a virtual circuit with the receiving DTE and assigns a logical channel.

Call Accepted

DTE toDCE

Indicates that the DTE accepts the establishment of the virtual circuit.

Call Connected

DCEto DTE

Indicates that the remote DTE has accepted the establishment of the virtual
circuit.

Clear Request

DTEtoDCE

Indicates that the DTE wishes to deassign
the logical channel and destroy the virtual
circuit.

DCE Clear
Confirmation

DCEtoDTE

Informs the DTE that the logical channel
is deassigned.

Clear Indication

DCEtoDTE

Indicates that the remote DTE destroyed
the virtual circuit.

DTE Clear
Confirmation

DTEtoDCE

Informs the DCE that the virtual circuit
has been destroyed and deassigns the logical channel.

DTE Data

DTEto DCE

Carries user data from the DTE over the
logical channel.

DCE Data

DCEto DTE

Carries data from the remote DTE over
the logical channel.

DCERR

DCEto DTE

Updates the DTE transmit flow control
information.

DCERNR

DCEtoDTE

Indicates a temporary inability of the
DCE to accept data packets. This condition is cleared by a DCE RR packet.

Table 8.2
X.25 Level 3 Packet Messages

161

(Continued)
DTERR

DTEto DCE

Updates the DCE flow control information, authorizing the transmission of additional DCE Data packets.

DTE Interrupt

DTEto DCE

Carries user interrupt data over the logical
channel.

DCE Interrupt
Confirmation

DCEtoDTE

Acknowledges receipt of a DTE Interrupt
packet.

DCE Interrupt

DCE toDTE

Carries Interrupt data from the remote
DTE.

DTE Interrupt
Confirmation

DTEtoDCE

Acknowledges receipt of a DCE Interrupt
packet.

Reset Request

DTEto DCE

Requests that the logical channel and virtual circuit be set to an initial state.

DCE Reset
Confirmation

DCEtoDTE

Acknowledges that the logical channel has
been reset.

Reset Indication

DCEto DTE

Indicates that the logical channel has been
reset.

DTE Reset
Confirmation

DTEto DCE

Acknowledges the resetting of the logical
channel.

DTE Restart
Request

DTEto DCE

Requests that all logical channels for SVCs
be deassigned and that all PVCs be reset.

DCE Restart
Cofirmation

DCEtoDTE

Acknowledges the Restart Request.

Restart Indication

DCEtoDTE

Indicates that all logical channels for SVCs
should be deassigned and all logical channels for PVCs should be reset.

DTE Restart
Confirmation

DTEto DCE

Acknowledges a Restart Indication.

Diagnostic

DCEto DTE

Indicates to the DTE an error on one of its
logical channels.

162· Internetwork Communications

The window size is selected independently for each direction of
data transfer on the logical channel. The range of packets in the
window changes as the packet-level module in the destination DTE
receives and acknowledges packets.
Flow control can also use Receive Ready and Receive Not Ready
packets. DNA uses this method of flow control when transmitting
data to a DeE. This method is never used when receiving data from
a DeE.

A DTE can construct a logical message from several consecutive
packets, by using the More Data bit in the data packets. This bit is
set in each packet, with the exception of the last packet in a
sequence of packets composing a logical message. Data packets
also contain another control bit, the Q bit, which allows a
transmitting DTE to define two categories of data-normal data
and interrupt data.
Each direction of transmission on a virtual circuit can have an
outstanding Interrupt which allows a DTE to transmit one byte of
information not subject to the rules of flow control that apply to
normal data packets. Interrupt data is transmitted via an Interrupt
packet and acknowledged by an Interrupt Confirmation packet.
The x.2S packet-level module of a DTE can use a Reset procedure to
initialize a virtual circuit. When a DTE wishes to reset a logical
channel, it transmits a Reset Request packet. A reset of one logical
channel for a virtual circuit will cause a reset of the corresponding
remote logical channel. A DeE confirms a Reset Request packet by
transmitting a DeE Reset Confirmation packet.
The PSDN can also initiate a reset. At the time of a reset, the PSDN
discards all data and interrupt packets in transit. When the

163

network wishes to indicate that the virtual circuit is being reset, it
transmits a Reset Indication packet containing a reason for
resetting. A DTE confirms a Reset Indication packet by transmitting a Reset Confirmation packet.
Either of the DTES associated with a SVC can destroy the circuit by
Clearing. The DTE'S x.2S packet-level module transmits a Clear
Request packet. The DCE returns a DCE Clear Confirmation
packet, which deassigns the logical channel. The Clear Request is
reported to the remote DTE in a Clear Indication packet. The
remote DTE replies with a DTE Clear Confirmation packet. The
logical channel at the remote node is then deassigned.
After a catastrophic failure, all SVCs associated with a DTE can be
cleared, and all pvcs can be reset by means of the Restart procedure.
A DTE issues a Restart Indication packet whenever the DTE Level 2
module connects to the DCE. A DCE can transmit a Restart
Indication packet to force a DTE to clear all its virtual circuits.
X.25 Frame-Level Module-Like the x.2S Packet-Level Module, this

module resides in the Data Link layer of the gateway system. It
provides the x.2S frame-level protocol that enables the gateway
system to communicate with the PSDN'S DCE.
The x.2S frame-level specification defines two alternative link
access procedures (LAP and LAPB) for frame-level communication
between a DTE and aDCE. DNA supports the LAPB procedure, which
defines the control of full-duplex communication over a synchronous line between a DTE and DCE. LAPB provides an error-free link
by detecting and recovering from transmission errors. It also
detects procedural errors and reports them to the user of the link.
In addition, LAPB provides synchronization, to ensure that the
frame-level entities in the DTE and DCE are in step.

164· Internetwork Communications

Frame-level modules in the DTE and DeE communicate by sending
and receiving frames to provide link control, data exchange, flow
control, and error control. There are three types of frames:

Information Frames (I Frames) are used to transmit packet-level
information and are also used to acknowledge previous correctly
received information frames.

Supervisory Frames (S Frames) are used to perform channel control
functions like acknowledging correctly received information
frames, requesting retransmission of information frames, and
indicating temporary inability to receive information frames.

Ummmbered Frames (U Frames) are used to provide additional link
control functions like connecting and disconnecting the link.
Frames are divided into commands and responses. Table 8.3 offers
a summary of the frame types.

Format

Commands

Responses

.-------

Information Transfer (I)

Receive Ready (RR)
Receive NOT Ready (RNR)
Reject (REJ)
Receive Ready (RR)
Receive NOT Ready (RNR)
Reject (REJ)

Connect Link (SABM)
Disconnect (DISC)
Unnumbered
Acknowledgment (UA)
Disconnected Mode (D M)
Frame Reject (FRMR)

Table 8.3
X.25 Level 2 Frame Types

165

Figure 8.3 depicts the relationship of the x.2S Gateway Access
modules to the PSDN and to the foreign DTE.

X.2S Access Methods
Access methods refer to how a DEC net system establishes communications over a PSDN. For x.2S communications, a DEC net node

Local Node
User Layer
The accessing program:
A user-written program. I
Network Application j
Layer
I
X.25 Gateway
j
Access Module
.
. ControI Layer Ij
SessIOn
End-to-End
j
Communication Layerj
Routing Layer
Data Link Layer

Physical Link Layer

Gateway Node (DTE) Foreign Node (DTE)
Foreign DTE Process
The cooperating program
Network Application
Layer
_ ..... X.25 Gateway
Server Module!
Session Control Layer
End-to-End
Communication Layer
I
Routing Layer
I
Data Link Layer
j
I
X.25 Packet j
jAnyDNA
. Data Link
Level
i
! Protocol
X.25 Frame·
I
I
Level
Physical Link Layer

Network Layer

X.25 Packet Level
Data Link Layer

X.25 Packet
Switching Network

Legend:
DTE
Data Terminal Equipment
DCE
Data Circuit Equipment
Router
A DNA node performing Switching
Incoming Call mapping Database
----c---

Figure 8.3
X.25 Gateway Access Operation

A Physical Communication Link
A path through a network
Data Flow through X.25
Gateway Access

166 • Internetwork Communications

needs to run specific software and sometimes needs to be able to
access dedicated hardware products. What a system needs for
communication over a PSDN depends on the type of system (DEC net
or non-DEc net) with which it wishes to communicate.
If a DEC net node communicates with DEcnet systems only,· it may
be able to use a DEC net facility called data link mapping (DLM).
(Check system-specific documentation to see if a system supports
DLM. Some systems, DEcnet-10 and DEcnet-20, for example, do
not.) DLM allows two DEC net nodes to communicate as though the
PSDN were not there. The x.25 interface in this situation is
completely transparent to users at both nodes. DEcnet node A can
use DLM if it has access to a DEC net node installed with PSI software.
The PSI node must be positioned as a DTE in relation to the PSDN. If
node A does have access to a PSI node, it can send data through a
DEC net Router/x.25 Gateway node on an Ethernet. To use the
DLM interface (either through a PSI node or DEcnet Router/x.25
Gateway), node A can be located anywhere on the network and
does not require any special software other than DEcnet.
If users want the option of being able to communicate with 11onDEC net as well as DEcnet systems over the PSDN, their DEC net nodes
need PSI software. Rsx-ll, VAX/VMS, and Tops-20 operating systems support system-specific PSI software. This software allows a
DEcnet system to conlmunicate in x.25 native mode ifit is directly
connected to a PSDN. x.25 native mode provides the complete x.25
programming interface which any DEcnet or non-DEc net system
connected to the PSDN can interpret. Regardless of vendor, all
systems connected to a PSDN must support the x.25 protocol.
. Note:
Tops-20 software is divided into two categories: (1) PSI gateway
software installed in the DN20 front-end processor, which provides
gateway functions and (2) PSI access routines installed in the

167

DECSYSTEM-20 main processor. The access routines communicate
with the gateway software to set up a connection with the PSDN.
Because the DN20 is the DTE in these connections, it must be
connected directly to the PSDN.
Native-mode PSI software enables users on the PSI DEC net node to
write application programs to access resources on remote DTES.
This process involves issuing calls to create or specify virtual
circuits, send and receive data, and terminate communications.
Refer to PSI documentation for specific operating systems for a list
of the supported calls, functions, and languages.
In DEC net-vAX and DEcnet-RsX nodes, PSI software includes the
DLM interface, thereby enabling user-transparent communication
over the PSDN between these and other DEC net nodes. DEC net nodes
containing PSI software do not require the services of intervening
DTES or DEcnet Router/x.25 Gateway nodes.
Instead of PSI software, a DEcnet-vAX node on an Ethernet can be
installed with a VAx-11 x.25/x.29 Extension Package (XEP). The
extension package provides subroutines that enable users to
employ DEcnet communication facilities to create and transmit
data packets containing x.25 information. XEP cannot be used
without the services of the DEC net Router/x. 25 Gateway. The XEP
data packets that users send are received by corresponding x.25
software in a DEcnet Router/x.25 Gateway. The DEC net Router/
x.25 Gateway is connected directly to the PSDN. It translates the
XEP messages it receives into x.25 native-mode messages that can
be understood by DEcnet and non-DEcnet systems connected to the
PSDN. The combination of XEP software and the gateway facilities
provides users with the complete x.25 native-mode programming
interface. Multiple DEcnet nodes can share a single x.25 connection to a PSDN through the Gateway or by using DLM.

168· Internetwork Communications

CCITT Recommendations X.3, X.28, and X.29
Thus far, the discussion in this chapter has involved computers or
terminals that send "packets" of data to the PSDN. These devices
are referred to as packet-mode DTES. There are, however, start-stop
mode devices, such as asynchronous terminals, which can also
communicate as DTES over a PSDN. CCITT has defined three
recommendations describing the PSDN interface for start-stop
mode communication: x.3, x.28, x.29.
X.3-For terminals operating in start-stop mode in a PSDN, x.3
describes the basic functions and user-selectable capabilities of the
packet assembly/disassembly facility (PAD).
X.28-Specifies how to control the user-selectable functions of the
PAD from the terminal.
X.29-Specifies procedures for the exchange of control information
and user data between a packet-mode DTE and a PAD facility.
A PAD is a facility provided by the PSDN. It acts as an interface
between the start-stop mode DTE and the network DCE. The
terminal user connects to the PAD and commands it to establish a
virtual circuit to a destination DTE. The PAD receives the characterby-character data generated by the user terminal, reformats it into
packets, and dispatches the packets through the network to the
remote computer. At the remote computer, a pseudodevice called
the x.29 Virtual Terminal Driver handles information sent by the
terminal user in much the same way that a regular DEC net terminal
device would. The x.29 Virtual Terminal Driver sets terminal
characteristics, such as typeahead options and line terminators,
that the remote terminal specifies. It delivers the information to
the appropriate application on the DTE. The PAD and the x.29
Virtual Terminal Driver operate in a manner that is transparent to
the user. The terminal appears to the terminal user to be connected
directly to the remote computer.

169

Built-In PADs-Asynchronous terminals connected to a DEcnetVAX node with vAx-ll PSI or vAx-ll x.2S/x.29 XEP can initiate a
circuit with another DEcnet or non-DEcnet system over the PSDN
without buying the services of the PSDN PAD focility. VAx-ll PSI and
VAx-ll x.2S/x.29 XEP each contain a program that emulates the
functions of a PAD on the DEC net system. (The standard DEcnet
terminal interface is used in these communications.) In addition to
the cost savings entailed in having a built-in "PAD," as opposed to
leasing the option from a PSDN, VAx-ll PSI and VAx-ll x.2S/x.29
XEP software enable a DEcnet-vAX node to initiate and to respond
to start-stop mode communications. Without the vAx-ll PAD
capability, an asynchronous terminal connected to a DEC net-vAX
node can only respond to messages initiated by 'terminals connected to a PAD on a PSDN.
Figure 8.4 shows the relationships between x.3, x.28, x.29, and
the PSDN.
The x.3, x.28, and x.29 recommendations are layered on top of
x.2S. These recommendations define the standards required for
start-stop mode systems to communicate over a PSDN in addition to
the protocols defined by x.2S. Digital PSI products include provisions for x.2S, x.3, x.28, and x.29 recommendations. Information in this chapter describing x.2S communications applies to
these other standards, as well .

. DECnet/SNA Communications
DEcnet/sNA communications involve sending data from one or
more DEcnet-vAX or DEcnet-Rsx node through a Gateway to an
IBM system in an SNA network. The protocol differences between
the IBM network and the Digital network are resolved by interceding functions in the Gateway. The IBM SNA network environment
sees the Gateway as a PU (Physical Unit) Type 2 cluster controller.

170· Internetwork Communications

Packet
modeDTE
Packet
modeDTE
(x.29
support)

-----~"'-XJ.2::~~~~~""

'\.?)(If----t'

procedures for
exchange of
information
across here

Packet
modeDTE

Start-stop
modeDTE

Figure 8.4
Relationships between X.3, X.28, X.29,
andthePSDN

The term Gateway in the discussion that follows refers to all
implementations of the DEcnet/sNA Gateway node.
The functions performed at the Gateway are not visible to the end
user. As far as the user is concerned, the Gateway is a black box that
receives DEcnet data at one end and sends out SNA data at the other.
The DEcnet node(s) and the Gateway involved in these communications can be located either in a wide area network or on an
Ethernet. See Figure 8.5.
DECnet/SNA Gateway Access Functions
DEcnet/ SNA Gateway Access functions reside in the Network
Application layer of a DEC net node. They are designed to make the

171

Ethernet Coaxial Cable

DEenet

-RSX

-VAX

DEenet

-RSX

-VAX

Figure 8.5
DECnetlSNA Environments

full capabilities of the SNA Data Flow Control, Transmission
Control, and Path Control layers available to user programs
residing anywhere in a DEcnet network. The DEcnetl SNA Gateway
Access functions of a DEC net node communicate over a DEcnet
logical link with the DEcnet system that is the pu2 on the SNA
network. The DEC net system and the pu2 system exchange
DEcnetl SNA Gateway Access Protocol messages over the logical

172· Internetwork Communications

link. Table 8.4 lists the SNA Gateway Access Protocol messages that
the DEcnet/sNA Gateway Access modules use to accomplish
remote access to the IBM SNA network.

DECnet/SNA Gateway Access Operation-Three processes are
involved in the DEcnet/ SNA Gateway Access operation: the user
process, the DEcnet/sNA Gateway Server process, and the IBM host
application process (IBM hosts are called puS nodes in SNA terminology). The user process issues calls to the DEcnet/sNA Gateway
Access Routines. These calls are translated into DEcnet/ SNA

Message

Function

Connect

Requests that a Session be established with a Primary Logical Unit (PLU) in an SNA host.

Listen

Waits for a Session to be established by a PLU.

Bind Data

Indicates that a BIND has been received for a
session solicited with Connect or Listen.

Bind Accept

Requests that the session being established with the
BIND be accepted and placed in the running state.

Normal Data

Carries data on the SNA session's normal flow to
and from the PL U.

Interrupt Data

Carries data on the SNA session's expedited flow to
and from the PL U.

Disconnect

Requests orderly session termination.

Disconnect
Complete

Indicates normal session termination has been completed successfully.

Abort

Requests abnormal termination of a session, or
indicates an UNBIND has been received from
thePLU.

Table 8.4
DECnet/SNA Gateway Access
Messages

173

Gateway Access protocol messages and transmitted over a DEC net
logical link to the DEcnet/sNA Gateway Server process. The
DEcnet/sNA Gateway Server transmits and receives SNA Basic
Information Units by accessing the functions of SNA protocol
emulation modules (see below, under DECnet/SNA Gateway
Access Modules), which connect directly to the SNA network as a
pu2 and which contain the functions of the SNA Path Control and
Data Link Control (SDLC) layers.
In a typical DEcnet/sNA Gateway Access protocol dialogue, the
first message exchange consists of Connect and Bind Data messages to establish a session with a Primary Logical Unit (PLU) in
the IBM host. The session is completely established when the Rind
Accept message is sent successfully. The user process can then
exchange data with the PLU by sending and receiving Normal
Data and Interrupt Data messages. When the user process wishes
to terminate the session, it informs the PLU by using the
appropriate Data Flow Control Request Unit. The PLU call thcll
terminate the session by sending an Unbind, which results in
session termination accompanied by an appropriate reason code to
the user process. Figure 8.6 shows DEcnet/sNA Gateway Access
Protocol exchange for a session requested by the user process (the
Secondary Logical Unit-sLu).

DECnet/SNA Gateway Access Modules
DEcnet utilities that provide the DEcnet/ SNA Gateway Access
Service include:
DECnet/SNA Gateway Access Module-This module receives user
SNA session requests and communicates with the gateway system
(the Physical Unit-pu-connected directly to the IBM SNA network).

174 • Internetwork Communications

User Node M.essage
Description

Messages

Gateway Node
Message Description

Request to establish
an LU-LU session

Connect Message ~

Gateway System issues an INIT-SELF
message to the SNA
network
BIND received from
PLU

User accepts session
as proposed by BIND

Bind Data
__
Message
Bind Accept
Message ------l.~

Normal flow I hta

Normal Data
Messa,l(,c

User Sends normal
flow RU to PLU

Normal Data
Message
Normal Data
Message
•

User Receives

Gateway accepts
session to PLU
Normal Flow RU
received from PLU

.....

Interrupt Data
Message

User sends expedited
flow Data to PLU

Interrupt Data
Message ------l.~

User senses session
termination

-------- Abort Message

Expedited flow RU
received from PLU

PLU sends UNBIND
to terminate session

Figure 8.6
DECnetlSNA Gateway Access
Message Exchange

DECnetlSNA Gateway Sewer Module-This module resides in the
DEC net system that is the PU on the SNA network (the gateway
system) and communicates with the DEcnetl SNA Gateway Access

175

module over a DEcnet logical link. It acts on requests made by the
DEcnet/sNA Gateway Access module and the PLU. The DEcnet/
SNA Gateway Server module uses the facilities of a SNA protocol
emulator to gain access to the IBM SNA network.
DECnet/SNA Protocol Emulation-This set of modules resides in the
gateway system and performs the functions of SNA Path Control
and Data Link Control. These modules connect to the IBM SNA
network as a Physical Unit Type 2 node. They also perform the
functions necessary to communicate with the IBM SNA network's
System Service Control Point (sscP).
Figure 8.7 depicts the relationship of the DEcnet/sNA Gateway
Access modules to the IBM SNA network and to the IBM host.
application (the PLU).
DEC net-vAX or DEcnet-RsX nodes can communicate with systems
in an IBM SNA network if they have been installed with appropriate
DEcnet/sNA software and have access to a Gateway. (For information on the operating requirements for the IBM system to communicate with the Gateway, such as supported subsystems and
component parameters, refer to the DECnet/SNA Gateway Installation Guide which is available through a Digital sales representative when a DEcnet/ SNA Gateway is purchased.)
The DEcnet/sNA Gateway implements IBM SNA protocols and
works with function-specific gateway access software on a DEcnetVAX or DEcnet-Rsx node to enable users to interact with the IBM
SNA network in the following ways:
Write application programs that communicate with programs in
IBM SNA network. DEcnet/sNA Application Interface software on
the DEcnet-vAX or DEcnet-Rsx node enables this internetwork
function.

176·Internetwor Communications

IBM Host Node

Gateway Node

Local Node
User Layer
The accessing program
(SLU):
SNA RcmoteJob Entry
3270 Emulation
A user-written program_,
Network Application i
Layer
i
SNA Gateway
i
Access Module
i
Session Control Layer i
~------~---------I

End-to-End
COlllmunication
Layer
Routing Layer
Data Link Layer

Physical Link Layer

PLU
The Cooperating program:
(e.g., CICS, IMS,JES2,
.J' JES3, etc.)
Network Application
Layer
SNA Gateway
._~___S_e_r_ve_r_M_o_d_u,lc_ _ _
!Session Control
Path i
!___--:-L_a-=-y_er__-c-__--I Control i
! End-to-End
(PU2) i
! Communication
!
Layer
! Routing Layer
Data
I
Data Link
Link I
.____
L_a.=...y_er______--I Controlj
! Physical Link
!
Layer

NAU Services
Layer

--fr

Data Flow Control
Layer
Transmission Cd.
Layer
Path Control Layer
Data Link Layer
\ Physical Link Layer
\

Router

..........

DECnet Network

Legend:
Secondary Logical Unit
SLU
Primary Logical Unit
PLU
Physical Unit Type 2
PU2
Physical Unit Type 4
PU4

Figure 8.7
DECnet/SNA Gateway Access
Operat.ion

Router

-:t-

A DNA node performing Switching
A Physical Communication Link
A path through a network
Data Flow through SNA
Gateway Access

177

• Access IBM applications using a Digital terminal as though it were
an IBM 3270 Information Display System (IDS) terminal. OEcnet/
SNA 3270 Terminal Emulator software on the Digital node makes
this possible .
• Submit jobs to an IBM system and receivejob output. DEcnet/sNA
Remote Job Entry (RJE) software enables this function.
Depending on the functions users wish to perform between the
DEC net and IBM SNA environments, they can install a DEcnet-vAX or
DEcnet-Rsx node with any or all of the software packages described
above. There are also gateway management tools that enable an
operator to manage the DEcnet/ SNA Gateway from a DEcnet-vAX
or DEcnet-Rsx node.

DECnet/SNA Circuits
Communications between a DEcnet node and the Gateway node
occur over DDCMP or Ethernet circuits. Communications between
the Gateway node and the IBM system occur over SOLC circuits. To
perform certain oEcnet/sNA functions, users may have to identify
particular circuits. Circuit ID specifications are different for DDCMP
and SOLC circuits. The DEcnet/sNA Gateway management guide
for a system contains information on specifying circuit IDS for each
type of circuit.
Some system management actIvItIes related to SDLC lines and
circuits require that the Gateway node be set up as the command
executor. The oEcnet/sNA Gateway management guide includes
information on this process, as well.

178 • Internetwork Communications

Loading the Gateway Node Software
Figure 8.5 illustrates two implementations of the Gateway node:
one resides on an Ethernet cable; the other does not. Functionally,
both nodes are the same. However, if the Gateway is attached to an
Ethernet cable, it must be downline-Ioaded from a DEC net-vAX or
a DEcnet-Rsx host node on the same Ethernet. If the Gateway node
is not located on the Ethernet, it can be directly loaded from floppy
disks and bootstrapped either manually or by means of a TRIGGER
command from a host node.

Chapter 9· How DECnet Supports Applications

To illustrate the DEC net capabilities introduced in the preceding
chapters, this chapter describes the user environment of a fictional
company called Trident Manufacturing Corporation. Several
interacting application subsystems support the company's daily
operations. These subsystems include:
• Marketing subsystem, with order-entry, sales analysis, and market
research applications.
• Manufacturing subsystem, with production scheduling, inventory, quality control, and purchasing applications.
• Accounting and Financial susbsystems, with billing, accountsreceivable, and short-term cash management applications.
Figure 9.1 illustrates the interaction among these subsystems and
applications within the context of the following discussion. The
figure does not show application interactions that are irrelevant to
a discussion of DEC net capabilities (the short-term cash management application interacts with other applications at Trident
Manufacturing, for example).
The order-entry application captures a basic set of order-related
information, including customer name, address, products
ordered, and delivery requirements. If the products ordered come
out of finished-goods inventory, this basic information is then
input to order-assembly and dispatch applications. If the ordered
goods have to be manufactured, (as in the example to be presented), the basic order-entry information goes to a production
scheduling application, which then interacts with raw materials
inventory and/ or component inventory. If required, the inventory
applications interact with purchasing. The production scheduling
system updates a corporate database that records historical and
scheduled utilization of production facilities.

180· How DECnet Supports Applications

Marketing
Order
Entry
Sales
Analysis
Market
Research

Manufacturing

Production
Scheduling
Inventory
Quality
~
Control
Purchasing
!~
" - -_ _ _ _---i

Accounting

Billing
Accounts
Receivable

Financial

Short-term
Cash
Management,
I
I

Figure 9.1
Subsystem and Application Interaction
at Trident Manufocturing Corporation

The billing and delivery-scheduling applications (not shown in
Figure 9.1) also use the basic order-entry information, as well as
other information retrieved from various databases that are
accessed and updated by the complex of application subsystems.
Accounts receivable, which interacts with credit and billing
applications, is updated on a regularly scheduled basis.

181

Since Trident Manufacturing is an international company, its
application programs interact both within and over national
boundaries. The nodes in Trident Manufacturing's DEcnet network communicate over packet-switched data networks (PSDNS)
in the United Kingdom, the United States, Canada, France, the
Netherlands, and West Germany. These cost-effective communications facilities implement the x.25 recommendation of the
CCITT. At Trident Manufacturing, DEC net Router/x.25 Gateway
nodes and x.25 Packetnet System Interface (PSI) capabilities resident in general purpose DEC net nodes (see Chapter 8) meet
requirements for communicating over PSDNS within and over
national boundaries.
The organizational structure of Trident Manufacturing Corporation dictates a network configuration that includes wide area and
local area networks. The processors, terminals, and personal
computers that make up the network are at functionally defined
locations such as sales offices, manufacturing facilities. warehouses, distribution centers , and corporate headquarters. [)ECllet
nodes at these locations are linked by leased lines and x.2S
networks. In addition, some applications require interaction
between DEC net nodes and IBM/sNA systems.
DEcnet capabilities discussed in this chapter include task-to-task
communication, file transfer, remote file access, and terminal
communications .

. DECnet's Task-To-Task Capability in an Application
Environment
DEcnet's task-to-task capability handles all the network processing
that enables application systems at Trident Manufacturing to
interact with one another over the network. (Chapter 5 describes

182 • How DECnet Supports Applications

how the task-to-task communication capability enables interaction between user application programs, between user programs
and DEC net modules, and between DEC net modules in different
nodes. The example here illustrates only the interaction between
user programs. )
The only requirement that network communication imposes on
programmers at Trident Manufacturing is that they write applications to issue calls to network services and to cooperate with one
another in regard to data formats and function-specific data
transfer operations. Logical-link processing by DEC net is completely transparent to terminal operators. End-to-End Communication layer software in the communicating nodes automatically
sets up the logical links (see Chapter 4) over which the applications
communicate.
Since all DEC net implementations are designed in accordance with
the specifications of the Digital Network Architecture, there is a
standard application-to-network interface across all implementations. All of Trident Manufacturing's applications, regardless of
the Digital operating systems under which they run, access the
network by means of the same mechanism. This common network interface encourages an open-ended design of application
systems. Thus, future applications that programmers atTrident
Manufacturing develop will be compatible with the existing set at
the networking level.

Task-to-Task Communication: Order Entry And
Production Scheduling
The order entry application has three levels of communication:
• Between local and regional sales offices (order-entry application
and sales analysis application)
• Between regional sales offices and corporate headquarters (sales
analysis application and market research application)

183

• Between local sales offices and corporate headquarters (order-entry
application and production-scheduling application)
Basic information captured at the local level includes customer
name, billing and delivery addresses, products ordered, and
delivery requirements.
The production-scheduling application is centralized at corporate
headquarters. This application system matches order requirements against available production facilities (as recorded on the
production-scheduling database) and against raw-materials-andcomponent inventories. It then posts an indication back to the
order-entering sales office as to whether or not delivery requirements can be met. If they can, the production-scheduling system
notifies the appropriate manufacturing locations and updates the
production scheduling database .
. Network and Application Interaction
Logical-link processing between the applications encompasses
three distinct operations: The establishment, control,· and
destruction of the link.
Establishment of the Link
At the sales offices, data entry operators at VT100 terminals invoke
the order-entry transaction system and complete the order-entry
forms displayed on the terminal screens. From the opera tors' point
of view, the procedure is straightforward. They enter order
informa~ion in the displayed forms; the completed forms are then
ENTERED; and a new form appears. What happens at the
applications level is transparent to the operators. At the applications level, the entered orders are checked locally against customer
and product files, and the operators are notified via the screen of
errors.

184 • How DECnet Supports Applications

The activities at the network level are also transparent to the
operator. At this level, the order-entry system opens communication with DEC net. Then, by issuing connect requests, it tries to
establish logical links with a sales analysis system at the regional
sales office and with the production-scheduling system at corporate headquarters.
If the other applications are ready to communicate with order
entry, the programs are logically linked, and they begin exchanging data in accordance with the rules established by the cooperating applications. The production-scheduling and sales analysis
systems could be receiving order-entry information simultaneously from several terminals at one sales office or at different sales
offices. DEC net End-to-End Communication layer software sets up
unique logical links for each pair of applications bound in a task-totask session. Any communications from production scheduling
back to order entry during a session, such as notification that the
delivery requirements specified cannot be met, are transmitted
over the logical link that had been established between the two
programs at the beginning of the session.

Control of the Link-In the course of the sessions between the
programs, DEcnet monitors and controls the links, making sure
that all transmitted data is received and that all data received at the
data link level is turned over to the receiving program in proper
sequence. Order data records are of variable length. End-to-End
Communication layer software at the sending node segments the
record and specifies a sequence number for each segment. End-toEnd Communication layer software at the receiving node
acknowledges receipt of each segment by sequence number and
assembles message segments in the order specified by the sequence
numbers.

185

To make certain that the order-entry and production-scheduling
applications interact efficiently, End-to-End Communication
software requires that the receiving program issue RECEIVE calls
that set up buffers for the receipt of data. End-to-End Communication layer software at the sending node will not transmit data
unless it has been notified that a buffer is available at the destination
node for the receipt of message segments. DEcnet employs this
flow-control technique to make sure that buffer overflow conditions do not arise. In this manner, message retransmissions, which
increase communications overhead, are almost entirely eliminated.

Destruction of the Link-When all order data has been entered, the
application initiating the logical link issues DISCONNECT and
CLOSE calls to the network. The logical link binding the
programs is destroyed, and the network resources that were
employed in maintaining that link are released .
. DECnet's File Transfer Capability in an Application
Environlllent
At Trident Manufacturing Corporation, terminal operators and
programs initiate file transfer operations over DEcneL Below, are
examples of each.
User-Initiated File Transfer: Market Research And Sales
Analysis
Trident Manufacturing's market research system runs at corporate
headquarters. This system accepts input from the sales analysis
applications running at the regional sales offices and from market
research analysts located worldwide.

186· How DECnet Supports Applications

Input from sales analysis provides historical information that is
used to plot product acceptance trends. Input from market
research analysts quantifies market trends in specific geographical
areas and projects the impact of new products on existing and
forecasted customer bases.
Market research analysis is supported by a series of charts produced
by the sales analysis application. The charts are written to files that
are then made available to the research analysts. Terminal operators
at the regional sales offices use electronic mail over DEC net to notify
the analysts when a new set of charts is available.
The analysts edit the charts, supplying additional input and
modifying existing data. Some use DEC net file transfer capabilities
to transfer the chart files from the market research application at
corporate headquarters to their local nodes. Others use the same
DEcnet facility to copy the chart files to their accounts on the node
running the market research system at corporate headquarters.
Others use the editing feature available at their local nodes to
update the charts online.
Analysts working at DEcnet-vAX nodes, for example, use the
command $ COPY CVAX::CHART.CUR *. * to obtain the
chart files.
With this command, a research analyst copys a file called CHARTCUR that resides on node CVAX (at corporate headquarters).
CHART.CUR on node CVAX is written to a file of the same name on
the analyst's local node. The analyst can specify a different file
name.
As Figure 9.2 shows, the copy request (solid line) and response
(dotted line) can be routed through multiple nodes. To the user

187

r-------,
Copy

Copy Request

--------~

Node

MYVAX

Node

CHART. CUR

(File Name)

: Node
1_ _ _ _ _ -

CHART. CUR

Figure 9.2
COPY Request/Response Is Shown
Routed through Multiple Nodes

making the copy request, this network activity is transparent,
however. The preferred path for each transmission is selected by
DEcnet's routing capability.
The DEcnet-vAX node at corporate headquarters runs ALL-IN-l
software (an integrated menu-driven set of office applications that
includes word processing, calendar and desk management, graphics, mail, spreadsheet, and other user-specified capabilities). Analysts with accounts on this node frequently use the integrated
spreadsheet, graphics, and word processing capabilities to prepare

188 • How DECnet Supports Applications

reports and presentation materials. They can then distribute the
reports over the network to specific individuals. The timeliness
with which the reports can be prepared and distributed enables
management to react to market shifts and adjust production,
development, and advertising direction appropriately.
The transfer of the files to or from the research analysts' systems is
made possible by the DEC net file transfer mechanisms described in
Chapter 6. Operation of the mechanisms is completely transparent
to the terminal users.

ProgranI-InitiatedFile Transfer: Quality Control And
Vendor Analysis
At Trident Manufacturing, programmed file transfer operates in
the interaction between quality control and purchasing systems.
The quality control applications run at the company's manufacturing locations, and the purchasing systems operate at regional
purchasing centers.

In one of the quality control operations, incoming components
used in product assembly are screened by sensors that analyze
dimensional characteristics. The sensors are multidropped from
DEcnet-Rsx nodes that control sensor operation and that translate
sensor input to formatted data.
Results of quality control analyses are accumulated by component,
shipment, vendor, and quality control location. On a daily basis, a
program that runs in a DEcnet-vAX node at each of the plants
accesses this information and builds files by manufacturing location, date, vendor, shipment, component, and quality control
location. The same program also develops information on rejected

189

items in terms of units and dollar value for each vendor, shipment,
component, and quality control location.
During the third work shift, when systems and lines are less busy,
a data acquisition program in the purchasing subsystems uses the
DEC net file transfer capability to move these files to the regional
purchasing centers, at which the data is input to a vendor analysis
and evaluation program .
. DECnet's Remote File Access Capability in an
Application Environment

The file transfer capability (user- and program-initiated) described
above is a subset of DEcnet's remote file access capability. This
capability enables programs and users to transfer files from one
node to another, to manipulate files on remote nodes, and to
display remote files at local nodes. The DEcnet mechanisms used to
perform these operations are identical to those used in file transfer.
Network Virtual Terminals
Another DEcnet feature that supports extensive remote file access
operation at Trident Manufacturing Corporation is the network
virtual terminal capability. This capability enables a user at a
terminal at one node to log on to a remote node and operate on the
remote node as if the terminal were directly connected to it. People
using this capability to access remote system resources must, of
course, have proper access to the remote nodes and files.

Although there are differences in the details of how the various
DEcnet implementations require users to exercise the network
virtual terminal (NVT) capability, the overall NVT procedures are

190· How DECnet Supports Applications

similar. A user at a terminal on a DEcnet-vAX node would use the
SET HOST command to log on to a remote node to use resources
located there:

$ SET HOST ATLANTA
Username: RAINER
Password: (not echoed)
Welcome to VAX/VMS Version 3.4 on node ATLANTA
$
In the above exanlple, a user identified as RAINER and located at a
terminal connected to a DEcnet-vAX node entered a SET HOST
command to log on to node ATLANTA. The system then asked for
Rainer's name and password. The welcome message that printed
out indicates that Rainer has a valid account on node ATLANTA.
The prompt ($) that follows, shows that Rainer is logged onto
node ATLANTA and can use its files and other resources as ifhe were
at one of ATLANTA'S local terminals. If Rainer did not have a valid
account on that node, an "Unauthorized User" message would
print out and Rainer would be returned to his local system.

Short-Term Cash Management
DEcnet's remote file access capability is the foundation for a shortterm cash management application currently under development
at Trident Manufacturing Corporation. This application is
expected to optimize the corporate treasurer's ability to identify
and invest idle funds.
Corporate funds are kept in many accounts at foreign and domestic
banks. A short-term cash management application system will
enable the corporate treasurer to check the status of these funds,
identify idle cash, and then transfer these nonworking assets to the

191

portfolio management group for investment, specify the funds for
debt reduction, or retain them in the current account. DEcnet's
remote file access capability is a principal component in ensuring
the success of this application.
The facets of the application development effort include developing appropriate security measures, identifying parameters and
processing requirements that will help define "unencumbered
funds," cross-referencing regional databases for relevant cash and
local currency value data, and developing algorithms that equate
investment opportunities with available cash. Through Gateways
and through DEcnet' s remote file access capability over local area,
wide area, x.2S, and IBM SNA networks, the treasurer's office will
be able to isolate and examine cash records maintained in regional
databases worldwide. Analysis of this information will then
enable the treasurer to make timely transfer and investment
decisions.
DECnet's Terminal Facilities in an Application
Environment
Interactive terminal communications (see Chapter 7) enable users
to communicate with one another over the network in realtime.
Network virtual terminal capabilities enable terminal users to
access any remote node and to use its resources as ifit were the local
node. In addition, the ability to include personal computers in a
DEC net configuration and to use them in both processing and
terminal-emulation modes adds a new and powerful dimension to
the terminal facilities available through DEcnet.

The following paragraphs describe some of the ways in which
Trident Manufacturing Corporation uses these capabilities in its
daily operations.

192· How DECnet Supports Applications

Order Entry And Production Scheduling
Production scheduling frequently involves some online communication between a sales office and a manufacturing site. If, for
example, the sales office is notified by the production scheduling
system that customer delivery requirements cannot be met, the
sales office might consult with the customer and negotiate another
acceptable date for delivery of ordered products. A terminal
operator or the salesperson will then invoke a DEcnet interactive
terminal facility (Phone or TLK, described in Chapter 7) at a sales
office terminal and communicate revised delivery instructions to
the manufacturing location.
When working at a terminal on a DEcnet-vAX node, for example,
the salesperson would invoke the Phone utility. In the following, a
dialogue is established with a user at another DEcnet-vAX node at a
manufacturing location named MFG7 .

. $ PHONE MFG7::SCHED
This command places a call to a user named SCHED at the
manufacturing site. If SCHED is currently logged on to the DEcnetVAX node, a message will be displayed on the screen indicating that
a communication has arrived. If the sender is a user named
GREEN at a node called SALES4, the message will read as
follows:

SALES4: : GREEN is phoning you on MFG7
To receive and respond to the call, SCHED enters:

. $ Phone Answer
This reply causes both terminals to display a split screen. The top
half of the screen displays the local user's input, and the bottom
half displays the remote user's input. Each participant in the
dialogue can enter text at the same time.

193

After a salesperson and a user at the manufacturing site have used
this facility to arrive at a mutually acceptable schedule for production and customer delivery, the manufacturing-location user
updates the production-scheduling database at corporate headquarters.
Trident Manufacturing finds that such informal interaction
between users frequently resolves customer service issues more
quickly than do predefined procedures. DEcnet interactive terminal communications support efficient user interaction throughout
the network.
Market Research
Because effective market research must be responsive to current
market trends, market research analysts at Trident Manufacturing
develop forecasts and test new product marketing strategies in real
time. Field input is played against market models; results are
analyzed by statistical processing capabilities; and' 'what if" input
is developed to refine and extend the models.

Research analysts use DEcnet's terminal facilities extensively in
performing this work. Field analysts provide input via VT100
terminals that are connected to nodes at regional locations.
Analysts at corporate headquarters use VT125 terminals connected to a DEC net-vAX node with node-resident ALL-IN-1 software
capabilities to develop reports and presentation material.
In developing reports, the analysts use the integrated ALL-IN-1
software capabilities to combine spreadsheet computation, graphics development, and word processing. When required, they use
ALL-IN-1 's mail capability to distribute reports and analyses over
the network.
In addition to the integrated mail capability that ALL-IN-1 provides,
several similar capabilities are available to DEC net users who might

194 • How DECnet Supports Applications

not need all of the capabilities of the integrated package. The VMS
MAIL capability, for example, enables users at DEcnet-vAX nodes to
exchange correspondence. To send a message, the user simply
enters the MAIL command, specifies· the subject of the correspondence, and either specifies a file to be sent or types the message in
response to a prompt. For example:

$ MAIL/SUBJECT = ZONE12 UPDATE.MKT
LONDON::GRAHAM
This command line delivers the contents of the file, UPDATE.MKT,
to a user named Graham at node London. If Graham is logged on
to node London, a message such as the following appears on
Graham's screen:

New mail from YORK: :PEEK
This line indicates that the waiting message is from a user named
Peek at node York. to read the mail, Graham enters the MAIL
command and presses RET. A message is displayed:

You have 1 new mail message
MAIL
Graham presses RET in response to this prompt, and the mail sent
by Peek is displayed on the screen.
Users who are not logged on when mail arrives for them are
notified of the messages received when they next log in.

Short-Term Cash Management
The PRo/DEcnet node in the treasurer's office is connected with an
Ethernet local area network cable that serves the department and

195

other adjacent groups. This personal computer serves the treasurer
in several ways:
· As a stand-alone processor with its own self-contained file and disc
serVIces.
· As a node on the DEcnet network that is capable of interacting
with computing capabilities and other resources located elsewhere
on the network.
· As a terminal on the DEcnet network that can access other nodes
and use resources located at those nodes.
The network mail services available to the treasurer through this
node or through another host node on the network will be of
significant value when the short-term cash managment application is fully implemented.
Decisions to use available cash must take into account both internal
and external factors in a very dynamic environment. Internal
factors include payables, operating requirements, and debt constraints. External factors, such as bank policies in regard to
minimum balances, for example, are important considerations in
determining thresholds that define cash as excess. Such policies,
however, are frequently conditioned by the general economic
climate, government regulations in regard to fund transfer, and
other uncertainties that govern international financial transactions.
Using the mail and dialogue facilities, the treasurer can keep
informed, in realtime, of actual or potential changes in such
conditions. Fund accumulations can be monitored along with
those conditions that can affect movement of those funds. Records

196 • How DECnet Supports Applications

of information that influence investment decisions can be retained
online for quick lookup and can also be printed if the treasurer
requires hardcopy. This kind of information will enable the
treasurer to optimize decisions in regard to existing and potential
inventories of investable funds .
. DECnet as a Consideration in Relocation Planning
In planning relocation of the corporate headquarters, Trident
Manufacturing Corporation management, along with a group of
outside consultants, carefully examined the consequences that the
projected move could have on future operations. The company
planned to move to an location outside the commuting distance of
a major metropolitan area. Real estate costs and tax advantages
were attractive; a dependable labor pool was available, as were
facilities for employee cultural and recreational activities.

The expense to be incurred-architectural fees, construction cost,
and employee relocation outlay-was substantial, however. In
addition, management harbored serious anxieties related to the
company's ability to change in the future. Senior managers were
concerned whether the location of the planned move would
confine their operational structure. Management realized that in
moving from its current urban location, Trident Manufacturing
would be cutting itself off from the numerous services-like
availability of rental space and a market for subleasing-that are
common to a metropolitan environment. The primary question
facing them was, "Would the new physical plant in its proposed
location, and built at great expense, be flexible enough to continue
serving a changing organization?" What, for exarriple, would
happen if Trident Manufacturing decided to sell a subsidiary
operation or to decentralize certain operations that were now
centralized or to realign intergroup and reporting relationships?

197

Together with the consultants, management at Trident Manufacturing determined that recent developments in computer technology, such as those offered by DEenet, facilitate change. Local area
networks, for example, permit those changes in information flow
that accompany reporting and administrative realignment to be
put in place rapidly, without disrupting ongoing business. Computer networking, Trident Manufacturing management learned,
provides a measure of protection to capital investment.

Chapter 10· Network System. Managem.ent

Managing a network involves configuring it, maintammg it,
monitoring its performance, and detecting and resolving problems before they become crises. Network management also
includes planning for the evolution of a network. A network
manager is responsible for performing all these tasks and ensuring
a smooth-running network. Systems managers, those responsible
for the operation of the individual nodes making up a network,
can greatly assist a network manager in achieving and maintaining
good network performance by closely monitoring the networkrelated aspects of their specific nodes. System managers are
responsible to both local and remote users. They should be aware
that, while local applications usually demand the greatest share of
system resources, remote users, potentially a very large group,
must be assured of each node's response to network applications.
The procedures for performing network management functions
can vary from system to system. For instance, the procedure for
generating a DEcnet/E node is different from that for generating a
DEcnet-Rsx node. RSTS/E is a timesharing system with a software
structure quite unlike the structure of the realtime, event-driven
Rsx-11M. Because each DEC net implementation (DEcnet/E,
DEcnet-vAX, and so on) is an extension of an operating system, the
different ways of managing the various nodes reflect differences
among Digital's operating systems. This chapter provides an
overview of network management functions rather than a description of the actual procedures for performing them. For procedural
information, refer to the documentation packaged with a specific
implementation.
Network system management functions in DEC net include:
• Planning for node generation to tailor DEC net software to suit a
specific node's network application. Node-generation planning
should be influenced by an overall network plan that establishes

199

basic network-wide conventions. Implementation of such conventions will assure users of communications and processing
capability throughout the network.
• Generating network software to build the tailored DEcnet software to
create an active node.
• Defining and redefining network parameters to set the various network
parameters whose definitions determine many aspects of a node's
role within a specific DEcnet configuration.
• Operating a node, such as starting up and shutting down a node and
the physical lines connected to a node.
• Monitoring node activity, including the day-to-day performance of a
node by gathering and analyzing logging data that DEC net makes
available.
• Downline loading, the procedure for loading system and task images
from an executor node to satellite nodes such as Rsx-lls and
communication server nodes.

Network system management responsibilities also include the
testing and monitoring functions described in Chapter 11.
· Network Management Utilities
Most network management modules reside in the Network
Management layer. In addition to these, there are network
management modules in the User and Network Application
layers. Also, one network management module, the Event Logger, has a queue residing in each lower layer.

The User layer of every DEC net implementation supports Network
Control Program (NCP) , a utility that interfaces with network
management modules in the lower DNA levels. NCP includes a
standard set of interactive commands that enables a network or
system manager to communicate with lower-level modules by
issuing commands from a terminal.

200· Network System Management

Table 10.1 briefly describes the function of NCP and other network
management utilities in each DEcnet implementation. NCP accepts
commands that activate network management modules in the
lower layers to perf~rm specific tasks or to request information
about the current state of the local node or the network.
The utilities or modules in the Network Management layer
include:
Network Management Listener, which receives network management commands via the Network Information and Control
Exchange (NICE) protocol from the Network Management layer
of remote nodes. This utility enables a network manager to
monitor the status and activity of remote nodes from a node at a
central location.
Network Management Access Routines, which provide generic
network management functions. These routines communicate
across logical links with the Network Management Listener using
the NICE protocol.
Local Network Management Functions, the component that takes
function requests from the Access Routines, translating the
requests into system-dependent calls.
Link Vf.i:ltcher, which senses service requests on a link from an
adjacent node and handles state changing for automatic remote
load, dump, and trigger functions.
Maintenance Functions, which provide the actual protocol operation to support system maintenance functions like downline
loading and data link loopback testing.

201

Link Service Functions, the module to which the higher-level
network management modules interface for services that require a
direct data link (bypassing the Session Control, End-to-End
Communication, and Routing layers).
Event Logger, which records significant events, like a line going
down, occurring in the lower layers. DNA specifies event types in
the Network Management interface to each layer. An event
processor in the Event Logger takes raw events queued in each layer
and records events of the types specified by the system and system
manager. Using the Event Logger protocol, an event transmitter
in one node can inform event receivers at other nodes of event
occurrences. Events travel to a node specified by the network
manager as the sink node, where they are output on console, file, or
monitor. This log of events at the sink node helps network
managers identify problem areas in the network.

Figure 10.1 shows the relationship among the network management components in a node.

In the Network Application layer, Loopback Access Routines and
Loopback Mirror enable logical-link loopback tests. The Loopback
Access Routines in a node use the Loopback Mirror protocol to
loop test messages from the Loopback Access Routines of a remote
node .
. Network Managetnent Messages
The NICE protocol handles most functions of the Network
Management layer. Table 10.2 lists and describes NICE messages.

202· Network System Management

_O_p_e_ra_to_r_ _ _ _~.1 NCpJ
User
l
Program

J
User Layer

----------1-------------------

IW~:~:er I

Network
Management
NICE
NICE
commands
Protocol
from other
r----'-_ _ _~_...,Protoc-r°.:;..l---__::_____,
Network
Network
nodes
-.- - - - - - Management r- - .. Management .... - - - - - - Network
Access Routines
Listener
Management
commands
to other
nodes

Local Network Management Functions

,----------11
Mainte?ance I~

I FunctIOns I
Link S~rvice

I FunctIons

Events
to other
nodes

... -

Eventsfrom
other nodes
~1 Event Logger ....- - - - --Network
Management
Layer

-----------t-t--------r----------

Data Link layer
communication
Control over lower level
functions (examine line ......
state, turn on routing,
etc.)
Legend:

Lower
Layers
System dependent
calls to application
layer and local operating system func.... Read event
tlons (file access,
queues
logical link loopback,
timer setting, etc.)

---~ Peer protocol interface
----... Network management interface

Figure 10.1
Relationship Among Network Management Components at a Single Node

•

Client interface

203

System

Utility Function

DECnet-VAX

NCP

Loads, controls, monitors, and tests DECnet
software; defines configuration database
parameters; downline-Ioads DECnet-11S and
communication-server nodes.

DECnet-RSX

NCP

Loads, controls, monitors, and tests DECnet
software; down-line loads DECnet-11S and
communication-server nodes.

CFE

Changes parameters in the configuration
file CETAB.MAC, which is produced at
network generation.

VNP

Changes the disk image of a DECnet-RSX
system (VNP cannot be run from a DECnet11S node).

DECnet/E

NCP

Loads, controls, monitors, and tests DECnet
software; maintains the DECnet/E parameter
file; reports the current status of active logical
links, known physical links, remote nodes,
and programs using local and remote send!
receive services.

DECnet-RT

NCP

Loads, controls, monitors, and tests DECnet
software; defines and changes configuration
database (CETAB.MAC) parameters.

DECnet-20

NCP

Loads, controls, monitors, and tests DECnet
software; defines the configuration database;
downline-Ioads system and task images.

DECnet-lO

NCP

Loads, controls, monitors, and tests software;
defines the configuration database; down line
loads system and task images.

Legend:
CFE-Configuration File Editor
NCP-Network Control Program
VNP-Virtual Network Processor

Table 10.1
DECnet Systems and Network
Management Utilities

204 • Network System Management

Message

Description

Request Down-line
Load

Requests a specified executor node to down-line
load a target node.

Request Up-line
Dump

Requests a specified executor node to dump the
memory of a target node.

Trigger Bootstrap

Requests a specified executor node to trigger the
bootstrap loader of a target node.

Test

Requests a specified executor node to perform a
node, circuit, or line loop back test.

Change Parameters

Requests a specified executor node to set or clear
one or more Network Management parameters.

Read Information

Requests a specified executor node to read a
specified group of parameters or counters.

Zero Counters

,Requests a specified executor node to either read
and zero or zero a specified group of counters.

System Specific

Requests a system-specific Network Management
function.

Response

Provides request status and requested information
in response to a NICE request.

Table 10.2
NICE Messages

The Event Logger protocol uses only one message-the Event
message. This message provides event-related information such as:
• The node at which the event is to be logged (the event sink) and
whether the event is to be logged to the sink's console, file, or
monitor.
• The event type and class.
• The data and time the event occurred at the source node.
• The name and address of the source node.

205

Message

Description

Command

Requests a loop test and sends the data to be
looped.

Response

Returns status information and the looped back
data.

Table 10.3
Loopback Mirror Messages

· Whether the event relates to a DEC net module, node, circuit, or
line.
• Specific data concerning the event.
A third protocol in the Network Management layer is the
Loopback Mirror protocol, which handles node loopback functions. Table 10.3 describes the messages of this protocol.
The Maintenance Operation Protocol (MOP) is another Network
Management protocol. MOP supports special primitive functions
between the Network Management and Data Link layers. These
functions do not require the services of any of the layers between
the Network Management and Data Link layers. Nodes that are
being downline-Ioaded or are upline-dumping cannot support
more than a minimal data link; such nodes would not therefore be
able to use the services of any layers above the Data Link layer. MOP
supports downline loading, upline dumping, loopback testing,
and system console control for a remote node and for unattended
nodes.
· Planning for Node Generation
This process entails gathering and consolidating information.
Digital-supplied DEC net software provides generalized network

206 • Network System Management

capabilities, but users must supply the data and programs that
create a complete and working DEC net application. The system
managers for the various network nodes accumulate the data and
programs relevant to their specific systems so as to incorporate the
data and programs eventually into the nodes for which they are
responsible.
To ensure a smooth-running network, system managers should
exchange information regarding their nodes that is appropriate to
network operation. Programmers responsible for network applications should cooperate with system managers. For example,
programmers must use correct addresses in calls that generate
connect requests (see Chapter 5). System managers must know the
correct names and object types that the network programs use to
identify themselves.
Configuration Databases-Every DEC net node has some form of
configuration database that defines characteristics of that node and
determines how it functions within the network. In some cases,
Digital-supplied software already includes such a database to
provide initial default values for many database entries. For other
implementations, the network generation procedure creates -the
database. Table lOA lists the term that each DEcnet implementation uses to identify its configuration database.
Depending on the type of DEcnet node, the configuration database
may need to be updated periodically to reflect changes in the
network or to tune the performance of the network. See below
under Defining Configuration and Other Static Parameters for an
explanation of the parameters typically included in a configuration
database. Before a node can participate in a network, its system
manager must define the configuration database.

207

System

Term

Comments

DECnet-VAX

Configuration
Database

The initial database is provided within
the DECnet software supplied by Digital. NCP commands are subsequently
used to modify the database.

DECnet-RSX

Configuration
File

File CETAB.MAC is created during
network generation and can be subsequently modified by the Configuration
File Editor (CFE).

DECnet/E

Parameter File

DECnet-RT

Configuration
File

File $NETPRM. SYS is created and
subsequently modified by NCP
commands.
File CETAB.MAC is created during
network generation and can be subsequently modified by network management commands.

DECnet-20

CONFIG.CMD

File CONFIG. CMD is shipped with
the TOPS-20 monitor and modified
by the installer to define the local
hardware configuration.

NETPAR.MAC

The network-configuration file containing system-specific parameters.
File CETAB.MAC is created during
network generation and can be subsequently modified by network management commands.

CETAB.MAC

DECnet-to

Configuration
Files

Table 10.4
Configuration Database Terms

Files mapping. CNF and
CETAB.MAC are created during
network generation and can be
subsequently modified by network
management commands.

208 • Network System Management

Network Generation Planning Aid-DEcnet-RT aids users in planning
for node generation. The aid consists of a command file with
questions pertaining to node generation options. A system manager runs this command file from a terminal and answers the
questions according to the requirements of the local node.
Using the system. manager's responses, the command file generates several worksheets. Each tells the system manager how to
generate an aspect of DEcnet, in accordance with the application
requirements of that node. See the network generation manual for
a DEcnet-RT system for a complete description of the command file
and the worksheets it generates .
. Generating Network Software
DEC net software arrives from Digital on such distribution media as
magnetic tapes or floppy disks. The type of medium depends on
the DEC net implementation and, in some cases, on the hardware
configuration of a specific system. DEcnet-Rsx is distributed on
several types of media to accommodate a variety of user hardware
configurations. The procedures for using the distributed software
to generate an active node are different for each implementation of
DEC net. To generate an RSX node, a system manager has to
regenerate the operating system first and then tailor and build the
network application as a second step. Other implementations do
not require a system manager to rebuild the distributed software.

The types of DEC net software modules distributed with a system
depend on the DEC net implementation and the specific network
application that the system supports. In general, distributed
DEC net software includes network device controllers, a DDCMP
module, an Ethernet protocol module, Routing layer modules, an
NSP module, and network utilities like NCP, NFT, and TLK.

209

. Defining Configuration and Other Static Parameters
The configuration database of a node contains parameters which
are internal network values that have a significant i~pact on
network operation and performance. These values are of two
types: characteristics and status parameters. Characteristics remain
constant until changed by a network or system manager. Node
name is a characteristic. Status parameters are dynamic and reflect
the condition oflines, circuits, modules, or loggers. For example,
line state is a status parameter. If a failed line comes up or a working
line goes down, the change in line state is recorded in the
configuration databases of the nodes in the network. System and
network managers can examine status parameters.

In addition to defining parameters that are part of a configuration
database, a system manager must define non-configuration
parameters, such as local node network object descriptions (see
Chapter 5).
Parameters can be defined in various ways, depending on the
DEcnet implementation involved. They can be defined at network
generation by means of NCP commands. They may be predefined
in Digital-supplied software. The paragraphs that follow offer
brief descriptions of the parameters that are typically defined for
most types of DEC net nodes.
N ode Addresses and Names
For every node in the network, the system manager must assign a
numeric address that uniquely identifies that node within the
network. See Chapter 2 for a discussion of node addresses and node
names.
Node Verification Passwords
Whenever one of its physical lines is turned on, a node exchanges
initialization messages with the remote node at the other end of

210· Network System Management

the line. (See below, under Controlling Line or Circuit State for a
discussion of starting up a physical line.) The nodes exchange
information like the version numbers of their DEC net software
modules or their node type-routing or nonrouting.
After exchanging initialization messages, a node can request that
the adjacent node supply a password to facilitate verification of its
identity. If a local node requires identity verification, adjacent
nodes must supply passwords to gain access to the local node. If
the local node does not need to verify the identity of adjacent
nodes, they do not have to supply passwords, and they automatically gain access to the local node after exchanging initialization
messages.
Within a DEcnet configuration that enforces node verification,
each node maintains a database of passwords that it sends to and
expects to receive from its neighbors. Receive password is the
password that the local node expects to receive from the adjacent
node. If the password actually received does not match the receive
password expected, the local node denies access to the adjacent
node. Transmit password is the password that the local node sends to
the adjacent node. The transmit password must match the receive
password that the adjacent node expects from the local node.
Network Object Paratneters
Most nodes maintain a database that describes all the network
objects-both user-written programs and DEC net modulescurrently residing in the node and capable of engaging in network
activity. (For a description of network objects, see Chapter 5.) The
manner in which the node's object database is maintained depends
on the DEC net implementation on that node. Typically, a node
stores object types, names, and addresses; access-control and
verification information associated with each object; and the
number of copies of a specific object that DEcnet can run to satisfy
incoming connect requests.

211

Routing Parameters
Parameters that affect the operation of the Routing module in
Phase III and Phase IV full-routing nodes include:
• Maximum path cost, a value that limits possible routes between two
nodes to paths that are equal to or less than this value.
• Individual line costs that figure in routing algorithms used by the
Routing module. Each line cost is a number from 1 to the
maximum path cost set for the node. Higher line cost can lead to
less traffic on the line because the Routing modules dispatch
packets on the least costly paths. Lowering the cost of a line does
not always increase the traffic it carries. For example, if a line leads
to an end node, line cost does not affect that line's traffic (since all
traffic arriving at an end node is intended for it). If one of a node's
lines is more expensive than others to use, the traffic on that line
can be diminished by raising the routing line cost assigned to it. In
fact, this can be done whenever the system manager wants to
decrease traffic on a line.

• A maximum number of hops per path that sets the limit on the
number of circuits a packet can traverse to arrive at its destination.
A node is unreachable ifit cannot be reached within the maximum
number of hops.
• Area maximum cost is a parameter for an area-based network. It
limits possible routes between Level 2 routers in different areas to
paths with cost equal to or less than this value.
• Area maximum hops, another area parameter that sets the limits on
the number of circuits a packet can travel between Level 2 routers
in different areas.

• The maximum area parameter sets the limit on the number of areas
a packet can traverse before arriving at its destination.
• A routing timer that determines the interval between automatic
updates of the local node's routing database.

212· Network System Management

• A buffor size for the unit of data actually transmitted over physical
lines by the Routing module .
• A buffor count that determines the size of the Routing module's
pool of available buffers.
Line Identification
Each communication line leading from a node has a unique
identification that is used in commands. A node's configuration
database usually contains the identification of each communication line. This identification has the following general format:

dev-c-u, where dev is a mnemonic for the type of device. Examples include UNA (for the DEUNA Ethernet communications
controller), DUP (for the Dupl1-DA), DMC (for the DMCl1-DA/ AR,
-MA, AL, or FA/ AR), DZ (for the Dzl1-A or -B), and DMR (for the
DMRl1).
c is a number (zero or a positive integer) designating the device's
hardware controller.
u is a unit/line number (zero or a positive integer) that is included
if the device is a multiplexer.

Examples of typical line identifications include:
Identification

Description

UNA-O

DEUNA, controller 0

DMR-O

DMRl1, controller 0

DZ-1-0

DZl1, controller 1, unit 0

Circuit Param.eters
In addition to identifying actual physical lines, most DEcnet
implementations also identify circuits, which are logical point-to-

213

point communications paths operating over physical lines (see
Chapter 2). At nodes with circuit-identifying implementations, a
network manager defines and manipulates circuits, rather than
lines, to control the flow of data between nodes. DEcnet-vAX,
DEcnet-RsX, DEcnet-20, DEcnet-10, and DEcnet/E users frequently
specify circuits, while users at other DEC net nodes specify lines in
equivalent network management commands.
Definitions of circuit parameters are based on the type of link
(DDCMP point-to-point, DDCMP multipoint, or PSI virtual circuit)
between nodes. When a circuit corresponds to a point-to-point
line, circuit IDS and line IDS are identical. For example, the string
"DMC-O" in a network management command can identify either
the line or the circuit of a point-to-point configuration. In
multipoint configurations, circuit and line identifications are not
identical. See below for an explanation of how they differ. At a
DEcnet-Rsx/PSI node, Data Link Management (DLM) circuits to be
mapped to the PSDN are not associated with specific physical lines.
The identification of such a circuit starts with the mnemonic DLM.
Multipoint Line And Circuit Para meters-A multipoint line is a single
communications line connected to more than two nodes. (A line
connecting two nodes is called a point-to-point line.) The DEcnet
implementations supporting multipoint are the Phase III and
Phase IV versions of DEcnet-vAX and DEcnet-Rsx and the Phase III
versions of DEcnet/E and DEcnet-RT. Figure 10.2 shows a multipoint line with its various components.

The control station is the device responsible for overseeing data
transmission to and from all the nodes attached to the line. The
devices attaching the other nodes to the line are called tributaries. A
DEcnet-Rsx, DEcnet-vAX, or a DEcnet/E node can support either a
control station or a tributary device. A DEcnet-RT node can
support only a tributary device on a multipoint line.

214· Network System Management

Node
A

I*l

Multipoint

Node

Node
C

B
/

Node

Node
D
/

I
E

Tributaries

* Device controller = control station for multipoint line:
DMPll synchronous link
DMVll synchronous link

Figure 10.2
A Multipoint Line

From the perspective of the control station, the multipoint line
and the tributaries connected to it constitute a single line. But the
separate paths to each tributary represent individual circuits. A
multipoint line, therefore, has more than one circuit associated
with it. (Figure 2.1 illustrates the relationship between a multipoint line and the circuits that correspond to its tributaries.)
A tributary supports only one physical link to the control station.
From the tributary's perspective, the DDCMP line that links it to the
control station is point-to-point. To the tributary, the line and the
corresponding circuit are the same.

215

In the context of full-routing implementations, whether or not a
node supports a control station or a tributary on a multipoint line
is not significant. The mechanisms for handling data transmission
on multipoint lines are transparent to the modules in the Routing
layer of the architecture.
At the control station's node, a system manager needs to define
several parameters that affect the operation of the multipoint line
and its corresponding circuits:

Tributary addresses: The database at the control station's node must
contain correct tributary addresses. The system manager must
therefore record the unique decimal line address of each tributary
on the line.
Polling ratios: The control station delivers data addressed to tributaries under its control. All data transfer between tributary nodes
takes place through the control station. In order to transmit data
originating from its tributaries, the control station polls each
tributary periodically and asks whether it has data waiting to be
transmitted. If the polled tributary wishes to transmit, the control
station permits it to send data. If a polled tributary does not wish
to transmit, the control station goes on to poll the next tributary
on the line.
The frequency with which a tributary is polled depends on the
frequency of its data transmissions. For efficient network operation, the control station polls active tributaries more often than
inactive or dead tributaries. A dead tributary is one that has not
responded within a predefined period of time.
Network or system managers can exercise control over the polling
of specific tributaries. If system managers decide that a tributary

216· Network System Management

should not be polled as often as others, they can issue a command
to assign an appropriate active polling ratio to that tributary. A
command can also be issued to set a dead polling ratio that applies
to all inactive tributaries.

A polling ratio is a number from 1 to 255. If a tributary's active
polling ratio is 5, the control station passes through the active
polling list five times before polling that particular tributary.
(DEcnet/E does not implement the polling technique discussed
here. )
Transmission Mode
The system manager for a node sets the transmission mode for
every line or circuit connected to that node. The transmission
mode is either half-duplex or full-duplex. See Chapter 2 for a
defini tion of these terms .
. Operating aNode
To start up a node, a network or system manager issues commands
from a terminal to load and activate required DEC net software and
to turn on physical communication lines. In response to the
startup procedure, the node's DEC net software initializes the
DEC net software in adjacent nodes connected by the activated lines
(see below, under Controlling Line or Circuit State). To shut down a
node, the network or system manager issues commands to halt
network activity that involves that node, to shut down physical
lines, and to unload DEC net software.
Controlling the State of a Node
For most implementations of DEcnet, a system manager turns a
node on and off by manipulating the state of that node. For
instance, the command NCP )SET EXECUTOR STATE ON activates the
DEcnet software at the node currently defined to be the executor.

217

The executor is the node at which the NCP command actually
executes. Using an NCP command, the system or network manager can determine whether the executor is the local node or a
remote node.
Most DEC net implementations define three distinct node states:
ON, SHUT, and OFF.
ON-The local node is enabled for performing network functions.
SHUT-The node maintains all existing logical links but does not
permit any new links to be created. When existing links are
disconnected, the node's state changes to OFF.
OFF- The local node cannot participate in any network activity,

and existing logical links are aborted.
Controlling Line Or Circuit State
A node cannot actively participate in a network until at least one
communication line or circuit has been turned on. By issuing an
NCP command, a system manager sets the state of a line or circuit to
ON, OFF, or SERVICE.
ON- The line or circuit is available for use by the m.odules in the

Routing layer. When a line or circuit is turned on, the local node
exchanges initialization messages and, optionally, node passwords
with the remote node at the other end of the line or circuit.
OFF- The line or circuit is not available for any kind of network

activity.
SERVICE- The line or circuit is available only for special network
functions like downline loading, upline dumping, and loopback

218 • Network System Management

testing. Some implementations of DEC net do not recognize SERVICE as a state distinct from ON; such implementations may
impose the SERVICE state on a line or circuit by their own
internal means.
When a line or circuit is in the ON state, DEC net software exercises
the data link protocol that ensures data integrity and sequentiality
for normal network transmissions. (The standard DEcnet protocol
for normal traffic is DDCMP.) In the SERVICE state, a line or
circuit embeds data in a protocol provided for special network
functions. The standard protocol for these functions is the Maintenance Operation Protocol (MOP).

· Monitoring Node Activity
At each node, DEC net continuously monitors the current state and
ongoing performance of the node. From a terminal on a node, a
system or network manager can use network management utilities to display the following information:
• The current state of the local node, remote nodes, and physical
lines.
• Values currently defined for the node's configuration database and
other static parameters.
• The contents of various counters that DEC net software maintains
to track network performance.
In addition to displaying information on request, DEcnet automatically logs certain events both at the operator's console and in a file.
Event logging records operational events such as a line starting up
or shutting down. DEcnet-RT does not support event logging.

Monitoring Remote Systems or System Console Control-The Maintenance Operation Protocol can be used to monitor and control
remote and unattended systems. From an executor node, a

219

network manager can monitor Phase IV communication servers
(the DEC net Router, the DEC net Router/x.25 Gateway, and the
Terminal Server). By means of MOP'S Reserve Console message, an
executor node can take control of the console of a remote system
and proceed through a complete dialogue with the remote system
(the executor sends Console Command and Poll messages and the
remote target node returns Console Response and Acknowledge
messages). At the end of the dialogue, the executor sends a Release
Console message.
Counters-DEcnet uses counters to track other kinds of events and
errors in addition to line startups and shutdowns. Periodically, a
system manager can record these counters in a file or display them
at a terminal to obtain detailed statistics on the node's network
activity, such as how many connect requests have been sent and
received and how many messages have been sent and received over
logical links .
In Phase III and Phase IV implementations, counters record
Routing layer activity as well. This includes how many errors of
different kinds have been found in packet headers and how many
line initialization and line verification failures have occurred.
Certain Routing counters enable a system manager to determine if
too many packets are being discarded due to network congestion~
If this is the case, the system manager can adjust the parameters
that control the paths over which a packet is routed, thereby
improving the distribution of traffic loads over network circuits.
Line counters monitor events on individual communication lines.
By periodically checking each line's-counters, a system manager
can assess its performance and be aware of potential line problems.
Line counters record statistics such as the number of data blocks
sent and received successfully, the number of blocks received with
errors, and the number of times a tributary has passed from an
active to a dead state.

220· Network System Management

After counters have been displayed or recorded in a file, a system
manager can issue Ncpcommands to set the counters to zero. In
this way, the system manager can manipulate the time span that
the counters monitor. For instance, the system manager could set
all node counters to zero as programmers begin to test a network
application. At the end of the test, the counters could be examined
to see how the application affected the node's performance.
Chapter 11 discusses some of the tools that the system manager can
use in testing network components and in monitoring network
activity.
The Network Management layer has been designed to limit as
much as possible the overhead to the network in gathering
statistics and errors. For example, there are no Routing counters
that detect bugs in the Routing code. Network Management
modules assume the code is correct. The Routing layer algorithms
themselves have been designed to detect and recover from most
failures that would disrupt network operation .

. Downline Loading
DEC net-vAX , DEcnet-RsX, DEC net-lAS , and DEcnet-20 support
downline loading, which means loading a memory or system
image from a file at one node onto a separate target node. The
target node is usually an Rsx-lls DEcnet node or a communication
server node (DEC net Router, DEC net Router/x.25 Gateway, or
Terminal Server.) These memory-only systems have no disk-based
file storage of their own.
A DEcnet-20 node loads a system image downline to a specially
adapted Rsx-11s node called a DN200, which supports a cardreader
and a lineprinter and which serves as a remote batch station. The
DN20, a communications front end, is also downline-Ioaded from a
DEcnet-20 node.

221

At a DEcnet-vAX, DEcnet-Rsx-llM, or DEcnet-Rsx-llM-PLUS node,
the system manager generates the target system image. Once this
is done, the image can be modified by Virtual Network Processor
(VNP) commands at an executor node. The load itself can be
initiated in one of two ways. An operator can issue NCP commands
to load the image downline to the target node, or an operator at
the target can initiate the load by triggering a bootstrap ROM (readonly memory). See below, under Performing a Downline Load for an
explanation of these procedures.
Upline dumping is a function that complements downline loading. Chapter 11 describes upline dumping.
Downline-Loading Definitions
The downline loading function is distributed among two or more
nodes in a DEC net configuration. The command node is the node
from which the system or network manager issues NCP load
commands. The executor node, which executes the NCP commands,
must be adjacent to the target node. The target node receives the
system image loaded down the line or dumps a system image up
the line. A single node can act as both the command and the
executor nodes.
Downline-Loading Database Param.eters
For every target node that has to be downline-loaded, the executor
has access to a permanent database. Each database contains default
parameters, such as node verification passwords and other account
information, for downline-loading a specific target node. The
system manager can override these defaults by providing parameter values in an NCP LOAD command. The parameters are defined
initially at network generation and can be redefined when
necessary.

222· Network System Management

Performing A Downline Load
Whether a remote command node or the target initiates the load,
the target must have local access to a cooperating program called a
primary loader. This loader is usually contained in a bootstrap ROM
(read-only memory) incorporated in the target. During the
downline-Ioad procedure, a series of programs can be loaded above
the primary loader. Each program calls the next until the entire
system image is downline-Ioaded.
The line or circuit between the executor and the target is in a
SERVICE state during the procedure. Either the system manager sets

the state to SERVICE, or the DEC net software sets it automatically.
How the state is set depends on the DEC net implementation on the
executor and on the manner in which a load is initiated.
The NCP LOAD command is the means of initiating a downline
load from a remote command node. As soon as the LOAD
command has been issued, an operator at the target must trigger
the bootstrap ROM manually, unless the service line's device
controller is a DMCll or DMRll device. These devices can trigger
the target's primary loader automatically if the LOAD command
passes down the correct password. Figure 10.3 illustrates a
downline load initiated by a command node.
In a target-initiated downline load, an operator at the target node
requests a load by triggering the primary loader manually. In
addition, the loader is triggered automatically at the completion of
an upline dump from the target. Target-initiated downline loads
always use the parameter values defined in the permanent database
for the target. Figure 10.4 illustrates a target-initiated downline
load.

223
2

Using NICE messages
sent over a logical link,
network management
modules in the command node forward the
request to the executor.

Executor Node
Target Node

Using MOP, the executor ships the load file
to the target node. This
completes the downline load.

An operator requests
a down-line load to a
target node from a remote executor node.

Legend:
MOP -Maintenance Operatibn Protocol
NICE -Network Informatio,h and Control Exchange
NCP -Network Control Program
NMM -Network Management Module

Figure 10.3
Downline Loading Initiated by a
Command Node

Target Node

The target's primary
loader is triggered, *
causing a down-line
load request to be sent
to the executor.

* An operator triggers the
loader manually or the
completion of an upline dump triggers it
automatically.

Executor Node

Using MOP, network
management modules
at the executor ship the
load file to the target.
Legend:
MOP = Maintenance Operation Protocol
NMM = Network Management Module

Figure 10.4
Downline Loading Initiated by a Target
Node

224 • Network System Management

Downline-Loading And Checkpointing RSX-llS Tasks
DEcnet-11M, DEcnet-11M-PLus, and DEC net-vAX support two capabilities relating to a DEcnet-llS node. The first is called downline
taskloading. Rsx-11s tasks can be stored at a DEcnet-11M, DEC net11M-PLUS, or DEC net-vAX node and loaded to the Rsx-11s node.
The second is called checkpointing, which is a standard Rsx-llM
capability. An executing Rsx-11s task can be interrupted, copied in
its interrupted state up the line, and then be replaced by a higherpriority task that is loaded downline from the executor. When the
higher-priority task is completed, the Rsx-11M executor reloads
the interrupted task downline so that the task can continue
executing.
At a DEcnet-11M or DEcnet-11M-PLUS node, the operating system
regularly checkpoints tasks to local disk storage. However, Rsx-lls
nodes are basically memory-only systems, so the only way to
checkpoint tasks is to use the executor's disk storage.
Downline loading and checkpointing give great flexibility to an
Rsx-11s node. To change the set of resident tasks at a stand-alone
Rsx-11s system, an operator has to reboot with a different system
image. See the DECnet-RSX System Manager's Guide and the
DECnet-VAX System Manager's Guide (available through a Digital
sales representative) for further information.

Maintenance Operation Protocol-Table 10.5 lists and describes MOP
messages that relate to downline-load, upline-dump, and loopback test operations.
Either the target or executor node can initiate a downline load. In
either ca5e, the target sends a Program Request message to tell the
executor what is needed and to indicate its willingness to cooperate. The major message exchange consists of Memory Load

225

messages from the executor and Request Memory Load messages
from the target. The sequence is completed when the host sends a
Memory Load with Transfer Address or a Parameter Load with Transfer
Address message (consult Table 10.5 for explanations of these
messages).
On an Ethernet data link, a target can multicast its Program
Request message and select an executor from those responding
with an Assistance Volunteer message.
Upline dumping is very similar to downline loading. The target
uses the Request Dump Service and Memory Dump Data messages.
The executor controls the process with the Request Memory Dump
and Dump Complete messages (see Table 10.5 for message descriptions).

226· Network System Management

Message

Description

Memory Load with
Transfer Address

Causes contents of the image data to be loaded into
memory at the load address, and the system to be
started at the transfer address.

Memory Load with- Causes the contents of the image data to be loaded
out Transfer Address into memory at the load address.
Request Memory
Dump

Requests a dump of a portion of memory to be
returned in a Memory Dump Data message.

Request Program

Requests a program to be sent.

Request Memory
Load

Requests the next segment of image data in a
loading sequence and provides error status on the
previous segment.

Request Dump
Service

Requests a dump to be taken.

Memory Dump
Data

Returns the requested memory image in response
to a Request Memory Dump message.

Parameter Load
with Transfer
Address

Loads system parameters and transfers control to
the loaded program.

Dump Complete

Signals completion of a requested dump.

Assistance
Volunteer

Indicates willingness to perform dump or load
service in response to a Request Program or
Request Dump Service message.

Table 10.5
MOP Downline Load and Upline
Dump Messages

227

(Continued)
Loop Data

Contains data to be sent back in a loopback test.

Looped Data

Returns the data from a Loop Data message.

Boot

Causes a system to reload itself if the verification
code matches and the system is willing to do so.
May result in a down-line load.

Request ID

Causes a system to send a System ID message.

System ID

Identifies the sending system by, for example,
maintenance version, maintenance functions
supported, processor type, etc.

Request Counters

Causes a system to send a Counters message.

Counters

Returns Data Link counter values in response to a
Request Counters message.

Reserve Console

Reserves a system's console for dialogue with
the requester.

Release Console

Releases a console reserved with the Reserve
Console message.

Console Command
and Poll

Sends command data to a reserved console and/or
polls for response data.

Console Response
and Acknowledge

Returns response data and/or acknowledges receipt
of a Console Command and Poll message.

Chapter 11 • Monitoring and Testing DECnet
Performance

A network is a highly dynamic environment. Applications and
other processes are exchanging information simultaneously over
thousands of circuits; functions at various levels are being enabled
and disabled; the status of network components is changing
rapidly; and internetwork communications could be making
multiple demands on DEC net facilities for protocol handling.
Warning conditions can arise and fatal events occur in different
regions of the network and at various levels of network function.
As the number of nodes increases, so does the volume and
complexity of this activity.
Maintaining a high level of network availability under such
conditions is a challenge to network managers. They must have
tools that report on, diagnose, and correct conditions that could
cause performance degradation or failures. Facilities must also be
available for monitoring network operation so as to detect conditions that could lead to major communications problems.
Several software tools, integrated in DEcnet and available separately, are designed to help the network manager in these assignments. The functions of the integrated testing tools include:
• Ensuring that user-specified nodes communicate properly and
that messages move over specified circuits and lines without being
corrupted.
· Testing operation of x.2S and IBM SNA Gateways to determine
whether or not protocols are being translated accurately.
• Triggering memory dumps to help system and network managers
establish the cause of node crashes.
Digital has developed Observer™, a monitoring tool that is not
integrated in DEC net and that is available as a separate product.

229

Running in a DEcnet-Rsx node, Observer alerts network managers to incipient network problems and gathers performance
statistics that can help them reconfigure and fine-tune the network
for optimal performance.
This chapter discusses the integrated testing tools and Observer.

. Integrated Testing Tools
Integrated into most DEC net implementations are tools that enable
loopback testing, tracing, and upline dumping.
Loopback Testing
This is a procedure for isolating faults in a connection between two
nodes. Loopback tests exercise software and hardware by repeatedly sending data through a number of network components and
then returning the data to its source. By changing the loopback
point, a network manager can locate problems precisely. Figure
11.1 illustrates the various hardware points from which test data
can be looped back to the point of origin. If a test completes
successfully, the data loops back to its source uncorrupted. If a test
fails, the data either does not return to its source or returns in a
corrupted state.
A system manager can run variations of the loopback tests to help
isolate the precise network component responsible for losing or
corrupting data. Digital software services personnel run loopback
tests routinely after installing DEC net software. Successful tests
verify that both the software modules and the hardware equipment on the tested path are operating correctly.
The loopback-testing capability is exercised through the NCP
LOOPBACK command in combination with other NCP commands.

230· Monitoring and Testing DEenet Peiformance

Modem

Line
Loopback
Device

Remote
Modem
Node

1 Loopback at controller
2 Loopback at modem on controller side
3 Loopback at modem on line side
4 Loopback at hardware loopback device inserted in line
5 Loopback at a remote modem

Figure 11.1
Hardware Loopback Devices

Loopback enables network and system managers to check
whether or not logical links have been established properly
between nodes, whether or not DEcnet protocols operate correctly
within and between nodes, and whether or not messages are being transmitted accurately over the circuits and lines that the managers
specify.
Figures 11.2 and 11.3 show the NCP commands used in conjunction
with command- and program-initiated node-levelloopback tests,
respectively. Figure 11.4 shows the NCP commands used in conjunction with line- and circuit-Ievelloopback testing.

231

Test data is transmitted from a source node to one of several pointsa local modem, a point on the line connecting two nodes, a remote
modem, or a remote node-and then looped back from any of these
points to the source node. Test results that are printed out indicate
whether or not the specified target network component (line,
circuit, or node) passed the test. Test messages specify the nature of
an error condition that may have caused the test to complete
unsuccessfully.
Loopback testing can be performed for general purpose nodes and
communications servers over Ethernet, DDCMP, SNA, or x.25
circuits.
Used together, loopback testing of nodes and circuits can help
network managers isolate problem areas in the network.
Node Loopback Testing-Node loopback testing consists of sending
test messages over a logical link to be looped back at a specific
point. The network manager can set a hardware loopback device
on a line, line device, or modem between the executor and adjacent
target. Alternatively, software components can loop back test
data. Different software components can be used. For example,
FAL, a user program, or the Loopback Mirror can loop data back to
their- associated access routines. Figures 11.5 and 11.6 illustrate
some types of node loopback tests.
One type of node loopback test uses the Network Management
Loopback Access Routines and Loopback Mirror residing in the
Network Application layer (Figure 11.5 B and Figure 11.6 C).
These modules communicate over logical links using the Loopback Mirror protocol.

232

NCP Setup
Commands
User
Layer

None

Set Node
CAT
(Line line-id
Circuit cir-id)

None

Loop Node
GRAHAM

Loop Node
CAT

Loop Node
BARLEY

Node GRAHAM
Network
Management
Layer
Network
Application
Layer
Session
Control
Layer
End-to-End
Communication
Layer

1
1

1
1
1

Routing
Layer
Data
Link
Layer
Physical
Link
Layer

Comm
Hardware

Node BARLEY
Physical
Link
Layer
Data
Link
Layer
Routing
Layer
End-to-End
Communication
Layer
Session
Control
Layer
Network
Application
La yer
Network
Management
Layer
User
Layer

Loopback
from
controller

1- - - - - 1
1
1
1- - - - - 1

COlllments

The command loops The command loops The command loops The command loops
data within the local data via loopback
data via loopback
data through node
node CAT. The line BARLEY where
node.
node CAT associassociatcd with
nl'twork managcated with a line
whose controller has CAT leads to adja- mcnt software in
becn set to loopback cent node BARLEY application layer
mode.
where data is looped loops data back.
back at the transport
layer. *

*The routing layer uses its routing algorithm to determine which path the test data will be looped back on.

Legend:

the network management software that accepts
command input
network management software that sends, loops, or
receives test data.
path travelled by test data
interface between modules of network management
software

Figure 11.2
Command-Initiated Loopback Tests

Other loop tests use logical links, but they do not use Network
Management software. For example, Figure 11.5 B shows a test
message traveling from one user task to another. Figure 11.6 D
shows a file transfer being used as a logical link test.
By using the NCP SET NODE LINE command, a network manager can
set up a special loop node path (Figure 11.5). In this case, if the
network manager sets no hardware device, the adjacent node's
Routing layer loops back the test.
The several methods of performing loopback tests-both
hardware- and software-based-give the network manager great
flexibility in testing network operation. In addition, the variety of
tests enable network managers to identify the exact component
causing a problem.
Trace Capability-The trace capability is useful in identifying protocol problems that can arise in the gateway products that handle
internetwork communication and in the PSI (Packetnet System
Interface) products that enable DEcnet and non-DEC net Digital
systems to communicate over PSDNS. For the DEC net Router/x.25

234

NCP Setup
Commands
User
Layer

None

Set Node
Set Node
CAT
CAT
(Line line-id (Line line-id
Circuit cir-id) Circuit cir-id)

000

None

80 0

Node GRAHAM
Network
Management
Layer
Network
Application
Layer
Session
Control
Layer
End-to-End
Communication
Layer

FAR

FAL

FAR

Routing
Layer
Data
Link
Layer
Physical
Link
Layer
Comm
Hardware

Loopback
from
controller

Node BARLEY
Physical
Link
Layer
Data
Link
Layer
Routing
Layer
End-to-End
Communication
Layer
Session
Control
Layer
Network
Application
Layer
Network
Management
Layer
User
Layer

e
13

Commcnts

Program A
sends data to
program B,
which rcturns
data back to
program A.
Target node is
localnodc
GRAHAM.

Program A uscs Same as test 1
network file
except targct
access routines node is a loop(NFARs) and
back nodc
File Acccss List- named CAT;
ncr (FAL) to
data therefore
store and thcn
loops back from
retrieve file data. an adjacent
CAT, thc target node. *
node, is a loopback node name
associated with
a con troller set
to loopback.

Same as test 2
except that
target node is
BARLEY.

Same as test 1
except Program
A and Program
B reside in separate nodes.

* The routing layer uses its routing algorithm to determine which path the test data will be looped back on.
Legend:

user program or DEenet module
path travelled by test data

Figure 11.3
Program-Initiated Loopback Tests

Gateway, the trace capability is exercised through the NCP TRACE
command. For the DEcnet/sNA Gateway, the capability is exercised through the SNATRACE utility. NCP TRACE and SNATRACE
enable network managers to capture and analyze messages that
define how protocols are being handled.
Protocol processing by the DEC net Router/x.25 Gateway can be
traced at Levels 2 and 3 of the protocol-handling software. Level 2
defines the link access procedure for data exchange between
computers and the data communications equipment (DCE) that
establishes, maintains, and terminates communications between
computers over a PSDN. The Level 3 protocol defines the procedures for formatting and exchanging packets transmitted over a
PSDN.
SNATRACE, used for identifying protocol problems that can arise in

using the DEcnetl SNA Gateway, enables network managers to
examine protocol handling at the SDLC circuit level, the physicalunit level, and the session level. At the SDLC circuit level, network
managers retrieve messages on a specified circuit as the messages
pass between the SDLC module and the device driver in the Gateway

NCP Setup
Commands
User
Layer

Node GRAHAM

Loop Line
line-id

I
I
I

Network
Management
Layer
Data
Link
Layer
Physical
Link
Layer

Comm
Hardware

Set (Line line-id
Set (Line line-id
Set (Line line-id
Circuit cir-id)
Circuit cir-id)
Circuit cir-id)
State Service/On* State Service/On* State Service/On*

I
I
I

D
Loopback
from
controller

Loopback
from -----" 1'--./
modem

or
from
line
loo£back
' - - . / ~ eVlce

Node BARLEY
Physical
Link
Layer
Data
Link
Layer
Network
Management
La yer
User
Layer
Comments

The controller must be
set manually to loop back
mode.

Both modem loop back
and linc loop back require
manual setting or insertion
ofloopback device.

The line must also be in
SERVICE or ON* state
at node BARLEY

* SERVICE or ON-depends on the implementation

Legend:

Figure 11.4
Line/Circuit-Level Loopback Tests

237

node. At the physical-unit level, network managers retrieve
messages handled by a specified physical unit as they pass between
the SNA module and the SDLC module of the Gateway. At the
session level, network managers retrieve messages being
exchanged within a specified session as they pass between the SNA
and SDLC modules of the Gateway.
Documentation for each of the gateway products describes how to
use these facili ties.

Upline Dum.p
In the context of upline dumping, a target node is either a DEcnet
RSX-11s node or a communications server node. The execu tor node
in both cases is usually the node from which the target node's
system was downline-Ioaded or a backup node of sinlilar capability, in the event that the original executor is not available.
The upline-dump capability enables a target node to initiate
automatically a memory dump upline to the executor node. This
process occurs when the target node senses an impending system
crash. By analyzing the memory dump, software specialists can
determine the cause of the system failure and take steps to ensure
that similar conditions do not arise again.

In order to keep the target node available to the network, the
executor node will automatically downline-Ioad the target node's
system image when the dump is completed. For detailed descriptions of the DEcnet facilities that operate in upline dumping and of
the NCP commands to be used to ensure that lines and circuits
specified in the procedure can accommodate the dump, see the
documentation of the systems that support upline dumping.
A system manager can trigger an up line dump from an executor
node at any time.

238 • Monitoring and Testing DECnet Performance

A. Local to Loop Node Test, Single Node, using files as test data
with hardware loopback.
User
Modules

User Task

Network
Management

I File Acc. List. I

Network
Application

f-~-----./-~/--

Session Cont.

•

-----------i-!------/-/---I!

End-to-End Comm.

- - - - - - - - - - i - ! - - - - - /- / - - - - !

Routing

/.1
/.1

-----------1-.-----

Data Link

---------i
Physical Link

i
i

_ _ _ _ __

/ - / - - - - - - - Hard·
/
Loop

l
/
/
D
----------r-iI-.1 - - - 0--/
ev. ,
! i
f
/
/:_:_o_o_._o_o_o_o_o_~:=:=:_o_
Comm.
Hardware
Legend:

L~/~ ~~ ~~.;~;~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~~J
°

__________ control flow

- . - . -~ data flow

Figure 11.5
Examples of Node-Level Testing Using
a Loopback Node

° _

° f

239

B. Node Test, Single Node, using loopback mirror and test messages
with hardware loopback.
User
Modules

Nep

, +
Net. Man. Ace. ROll.

Network
Management
Network
Application
Session Cont.

1--------/--/----

End-to-End Comm.

-----------j-I
Routing

-----------j-j

/"-/------./

/
/
/---:/
------

Data Link
1
1
/
/
----------I-i---;"-/------Physical Link
1
./ /

Comm.
Hardware
Legend:

Hard.
Loop

lJI=?&~~~~~~~=~~~~=;~.['== :

control flow

- . - . ---.. data flow

240· Monitoring and Testing DECnet Pedormance

(Continued)

c. Local to Loop Node Test, Two Nodes, using user tasks.
User
Modules

I User Task

User Task

Network
Management
I

Network
Application

I
I

------------------------ ----------1--------------

Session Cont.

------------------------ ---------1---------------

End-to-End Comm.
------------------------ --------1---------------I

Routing
I

Data Link

"------1------------------I

Physical Link

----------------.-------I-----;--~---------------

Comm.
Hardware
Legend:

control flow

-" -" -.. data flow

241

User
Modules
Network
Management
Network
Application
Session Cont.
End-to-End Comm.
1

Routing

!------

j
0--1
1

Data Link
!
---------------------------------j--!-----Physical Link
1
!
1

----------------------------:=J--~-j--!----~0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0

1
1

Legend:

- _ control flow

data flow

242· Monitoring and Testing DECnet Peiformance

(Continued)

D. Local to Loop Node Test, Two Nodes, using loopback mirror and
test messages.
User
Modules
Network
Management
Network
Application

ILoop. Mirror I
------------1/--1----

Session Cont.
------1'
1

End-to-End Comm.

./ i ';"- - - - - - -

-----/---.!.-.- - - - -

- - - - - - -_ _- -_ _ _ 1
I.

Routing
i 1
;";"
- - - - - - - - - - - - i --i----/!·-------Data Link
i i
/ /
- - - - - - - - - - - i -;---/----:1-------Physical Link
Comm.
Hardware

•

1-,-_ _ _ _ _ _ __

!! /~: -.(~.

-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..-=-..--

1.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.

°-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-.-0-0-0-0-0

Legend:

control flow

. -. -. -~ data flow

243

User
Modules
Network
Management
Network
Application
Session Cont.
End-to-End Comm.

R_o_u_t_i_n_g___________________________

0_0_0_

i-!--i
Data Link
i!
i
---------------------------------i--!---i

~~~:I~lnk~:~~:~:~~:~:.~.~:.~I~. ~ =: =--LT
_-_0-0

Legend:

- . . control flow

data flow

244 • Monitoring and Testing DECnet Peiformance

A. Normal Local to Local, using loopback mirror.
User
Modules
Network
Management

ILoop. Mirror I

Network
Application

----------!--+-----1

Session Cont.

•

1----

End-to-End Comm.
Routing

I
-0-0-0-0-0-0-0-0-0-

Data Link
Physical Link
Comm.
Hardware
Legend:

-------.. control flow

- . - . -~ data flow

Figure 11.6
Examples of Node-L eve I Logical-Link
Loopback Tests

245

B. Normal Local to Local, using user tasks.

IUser Task

User
Modules

I +User
Task
____

_ _ _ _ _ _ _ _ _ _ _ _ _ 1_ _ _ _

Network
Managenoent

1
1

Network
Application
Session Cont.
End-to-End Conono.
Routing

Data Link
Physical Link
Conono.
Hardware
Legend:

- - . control flow

-. -. -~ data flow

246 • Monitoring and 'Testing DECnet Performance

(Continued)

c. Normal Local to Remote, using loopback mirror.
User
Modules
Network
Management
Network
Application

Loop. Ace. Rou.

--------------------!-----~ --------------

Session Cont.

•

-----------------------

1-----------------

End-to-End Comm.

-----------------

Routing
1

Data Link

•

-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-

------------------------j------!----------------Physical Link
j!
--j------!;-.--------------Comm.
Hardware
Legend:

1.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.

------. control flow

-. -. -~ data flow

247

User
Modules
Network
Management
Network
Application

_S_e_s_si_o_n__C_o_n_t_·__________________________

i____ i_____

i
End-to-End Comm.
--------------------------------------1---i
~
_R_o_u_t_i_n_g_______________________________ i ____ !_____
i
~
_])_a_t_a_L
__in_k______________________________ i ____ ~ _____
1

-----

Physical Link

.=.=.=.=.=.=.=.==.==~==.=~===~=:r

Legend:

control flow

-..

data flow

248· Monitoring and Testing DECnet Performance

(Continued)

D. Norm.a} Local to Remote, using files as test data
User
Modules

1 User Task .1

Network
Management
Network
Application

I File Ace.

ROll.

Session Cont.
End-to-End Comm.
Routing

------------------------------1---------------Data Link

------------------------------1-----------------Physical Link

---------------,r------------I--,---------------

Comm.
Hardware
Legend:

- . . control flow

I. _. _. _. _. _. _. _. _. _. _. _. _. _. _. _. _. _.

-. -. -~ data flow

249

I User Task I

User
Modules
Network
Management
Network
Application

!
File Ace.!
List.

--------------------------~------------

Session Cont.
End-to-End Comm.
Routing

-------------------------------/----------------Data Link

------------------------------/
Physical Link
/

-.-.-._.-.-.-.-.-.-.-.-._.-._./

Legend:

----.. control flow

-. -. __ data flow

250· Monitoring and Testing DECnet Pedormance

. Monitoring Network Operation with Observer
Observer is a Digital-developed software product that monitors
the performance of DEC net networks. Installed in a Phase III or
Phase IV DEcnet-RSX node, it gathers, processes, and generates
information on network performance. Network events such as
changes in line, link, and node states are displayed as they occur on
a VT100 terminal dedicated to this function. Other information is
logged to files that can be processed periodically to produce
network performance reports. A command terminal that offers a
menu-based approach to network surveillance enables a network
manager to display a broad range of information concerning the
state of designated network components. Network managers can,
for example, retrieve statistical information regarding network
operation and also enter commands affecting the relationship
between Observer and specified network components or regions.

The information that appears on the event-display terminal can
alert network managers to conditions that could cause problems in
network operation. Network managers can then take steps to
correct such conditions before major communications or performance problems develop.
Consider the following example. A message from Observer
notifies a network manager of high data traffic over a specific link.
This might mean that certain nodes are processing and transmitting unexpected data loads or that the network manager may have
originally specified too Iowa data rate for that link. The network
manager could take appropriate corrective measures like changing
link and node parameters and thereby improve performance in that
part of the network.
The menu-based approach of the command terminal (shown in
Figures 11.7, 11.8, and 11.9) enables network managers to monitor
closely specific regions and components of the network and to
control Observer operation.

251

08SERVER,

13: 34: 03

20-MAY -82

COMMANo/o I SPLAY
MASTER MENU
A
8
C
o
E
F
G

NETWORK SUMMARY STATUS DISPLAY
SUMMARY STATISTICS DISPLAY
TRAFFIC STATISTICS DISPLAY
ERROR STATISTICS DISPLAY
OPERATOR COMMANDS
NETWORK ELEMENTS DATA 8ASE oEF I N IT I ON UT I LI TY
oECNET "NCP" UTILITY

E!'oITER SELECT I ON:

Figure 11.7
Observer Master Menu

Figures 11.7, 11.8, and 11.9 illustrate just one example of how
network managers can use Observer to focus on various network
areas and components. Figure 11.7 illustrates the Master Menu
that appears on the command terminal. If the network manager
selects "A" from this menu, the Network Summary Status
Display (Figure 11.8) appears. This screen, which is updated
approximately every 30 seconds, displays summary information
concerning the network being monitored. In addition to present-

252· Monitoring and Testing DECnet Petjormance

13: 37: 14

OiISERVER SUI'II'IARY STATUS
Su .... ort.d R.,ionll
CNSNET
at
13.36
NCSOVI

CATNET
at
13.38
OONJON

REACHAII I LI TY CHECK
R.ach. 8
Unr.ach. 2
Statull NORI'IAL

E-t:

TELF

TELG

REACHAII I L I TY CHECK
R.ach. 2
Unr.ach. 1
Statu •• NORI'IAL

E-t I

TELII

TELC

20-MAY-82

[CNSNET - -CATNET J

NCST3

POLLING

lIa •• :

POLLING

lIa •• ,

Statu.
WARNING
INIT
TELH

Ti ...
62.
O.

NOD505

Status
NORMAL
INI T

SlJe C fS 5
5

Fai I Eli'
1
6

DRAGON

NCSPAT

Ti ""

Success

20.
O.

2
0

NCST

Fai I El i '
2

NeSOVI'I

SI-OETAIL 1'I00E

K.y,

NORMAL, ALERT, ER'ROR, OISAIILEO

Figure 11.8
Observer Summary Status Display

ing identification information, the display specifies the number of
nodes reachable and unreachable from the Observer node, the
results Observer obtained, and other statistics that Observer
gleaned in the most recent polling cycles. The names of the nodes
in the network or network regions (which are user-defined) are
also displayed. Analysis of the information logged by Observer can
serve as a basis for reconfiguring the network and tuning it for
optimum performance.

253

OBSERVER NODE OETA I L 0 I SPLAY
REG I ON I

eNS NET

NODE I

NCSOV 1

5U""ARY STATUS I ERROR
Node ' . Oua Ac .. uisition 11 "ENABLED"
Node ' . Real-ti.e Oi ... la. is "ENABLED" • • • •
Node ' . E-T Data Ac .. uioition is "ENABLED"
Error if one of followin. i. NO
Nod. is R•• chabl.
• ••••• , ••••••••••.
Node ' . D.ata Ac .. uisition Status is "NDR"AL" • • • • • •
Alert i f on. of followin. i. NO
All Links have "NOR"AL" or "DEGRADINC" Error Rat ••
•
All Links haue Stat- •• of ·ON M
t·
•
t
••••••••

NEXT ACTION.

YES

YES
NO
YES

NO
YES
YES

51-SELECT REG S2-SEE NEXT NODE DETAIL RETURN - Actio. Di ... l u

Figure 11.9
Observer Node Detail Display

To focus-in more closely on the network, the network manager
can summon up a Node Detail Display (Figure 11.9) for a
particular node. As the figure illustrates, this screen displays status
information concerning the selected node. If network managers
need further information in order to diagnose a problem, they can
select another display directly or go back to the Summary Status
display. The Observer User's Guide (available through a Digital sales
representative) contains detailed information on the ease with
which Observer enables users to move from one screen to another.

Element

State/Status

Possible Conditions

Link

Link Error-Rate
Status

Normal/Degrading/Degraded/
Unusable

Link

Traffic-Rate Status

Normal/High/Low

Link

Realtime EventLogging Status

Enabled/Disabled

Link

State

Initialization/Online/Unknown/
Removed

Node

Base-System and
End Traffic-Enable
Status

Enabled/Disabled

Node

Data Acquisition
Status
Traffic-Rate Status

Unreachable/Normal/Unknown/
Error

Node

Normal/High/Low

Node

Realtime EventLogging Status

Enabled/Disabled

Node

State

Initialization/Normal/Warning/
Fatal Error

Region

Base-System and
End Traffic-Enable
Status

Enabled/Disabled

Node

Reachability
Check-Completion
Status

Initialization/Normal/Warning/
Fatal Error

Node

Base-System and
End Traffic-Polling
Completion Status

Initialization/Normal/Warning/
Fatal Error

Table 11.1
Observer-maintained State and Status
I nJormation

Observer monitors nodes, links, end-to-end traffic, and network
regions. It provides three kinds of information reports: state
information, status information, and statistics.
Table 11.1 identifies state and status information. Table 11.2 lists
the kinds of statistics that Observer gathers and maintains. See
Observer documentation for details on function, installation, and
operation of the product.

255

Type of Statistic

Element

Statistics

Time-related

Node

Time base in seconds, Time last data
received, Time oflast state

Link

Time base in seconds, Time last data
received, Time oflast state change
to on

Node

Buffer-resource errors, Packet-format
errors, Partial routing update loss,
Aged packet loss, Node-unreachable
packet loss, Out-of-range packet loss,
Oversize packet loss, Network errors,
Node errors

Link

Data blocks sent, Data blocks received, Collision-detect-check failure
(Ethernet only), Data errors outbound, Local-reply timeouts, Data
errors inbound, Local-buffer errors,
Unknown frame destination (Ethernet
only), Congestion loss, Selection
timeouts, Resets, Link error level,
Link down

Node

Bytes received, Bytes sent, Node
traffic level (bytes), Packets received,
Packets sent, Node traffic level
(packets)

Link

Link traffic level (bytes), Link traffic
level (packets), Bytes sent, Data blocks
sent, Data blocks received, Transit
packets sent, Originating packets
sent, Blocks deferred (Ethernet only),
Blocks sent with collision (Ethernet
only), Bytes received, Transit packets
received, Terminating packets received, Multicast blocks received
(Ethernet only)

Error-related

Traffic-related

Table 11.2
Statistics Maintained by Obseroer

Glossary

Access Control: The mechanism in a node for screening inbound
connect requests and verifying them against a local system account
file. Access control is an optional Session Control function.
Active Side: In the context of MOP loopback tests, the node that
controls a test.
Adjacency: A circuit and node pair. An Ethernet with n attached
nodes represents n-l adjacencies to a router on that Ethernet. On
DDCMP and x.2S circuits, adjacencies are in one-to-one
correspondence with circuits, since these circuits interconnect
pairs of nodes.
Adjacent Node: A node removed from another node by a single
physical line.
Aged Packet: A packet that has exceeded the maximum number
of visits.
ASCII: American Standard Code for Information Interchange.
ASCII is a seven-bit-plus-parity code established by the American
National Standards Institute to achieve compatibility among
different computers and terminals.
Asynchronous Line: A physical line on which the time intervals
between transmitted characters may be of varying and unequal
length. On an asynchronous line, the Physical Link layer provides
bit and byte framing, and the Data Link layer, message framing.
Back-off: The procedure on an Ethernet Data Link channel
whereby nodes attempting to retransmit after a collision (see
below, under collision) delay retransmissions in order to reduce the
load on the channel.
Bandwidth: The range of frequencies assigned to a channel; the
difference, expressed in Hertz, between the highest and lowest
frequencies of a band. The higher the bandwidth, the more the
data throughput.

257

Binary Synchronous Protocol: A data link protocol that uses a
defined set of control characters and control character sequences
for synchronized transmission of binary coded data between
stations in a data communications system.
Carrier Sense Multiple Access With Collision Detection
(CSMA/CD): A distributed channel-allocation procedure in
which every station can receive the transmissions of all other
stations. Each station awaits an idle channel before transmitting,
and each station can detect overlapping transmissions by other
stations.
CCITT: The International Telegraph and Telephone Consultative
Committee, the technical committee of the International
Telecommunications Union (lTv), which is responsible for the
development
of
recommendations
regarding
telecommunications, including data communications.
Channel: The data path joining two or more stations, including
the communications control capability of the associated stations.
Characteristics: Node parameters (values) that are generally
static. These reside in a permanent database on the node.
Characteristics are a Network Management information type and
can be altered by a network manager. They remain unaffected by
network activity. Node name is an example of a characteristic.
Circuit: A logical connection between protocol modules at the
Data Link layer. There is one circuit for each DDCMP point-to-point
line and one circuit for each tributary station on a DDCMP
multipoint line. There is one circuit for each x.2S permanent or
switched virtual circuit. There is one circuit for each protocol type
on an Ethernet.
Circuit Cost: A positive integer value, assigned by a network
manager, associated with using a circuit. When travelling from
node to node, messages are routed along the path with the smallest
total cost.

258· Glossary

Client: A module that requests service of another module.
Client Layer: A layer, typically the n + 1 layer, which acts as the
client of the n layer. For example, the End-to-End Communication
layer is the client of the Routing layer.
Collision: Overlapping transmissions by two or more stations on
an Ethernet. The Etherent Physical Link layer detects collisions
and the Ethernet Data Link layer retransmits affected data after a
random interval.
Command Node: The node where an NCP (the Network
Control Program, used in network management) command
originates.
Congestion: The condition that arises when there are too many
packets in the queue- waiting to be transmitted.
Congestion Control: The Routing component that manages
buffers by limiting the maximum number of packets on a queue
for a physical line. This component is also called transmit
management.

Control Station: From the perspective of the Data Link layer
protocol, a station that has overall responsibility for the orderly
operation of a circuit.
Controller: The hardware that controls a line. A multiple-line
controller device is responsible for one or more lines. The
controller identification is part of a line identification.
Cost: See circuit cost.
CSMA/CD: See carrier sense multiple access with collision detection.
Data Flow: The movement of data from a Session Control
module in a source node to a Session Control module in a
destination node. NSF (Network Services Protocol, operating in
the End-to-End Communication layer) transforms data from

259

Session Control transmit buffers into a network form before
transmitting it across a logical link. NSP retransforms the data at
the destination from its network form to its receive-buffer form.
Data flows in both directions (full-duplex) on a logical link.
Data Communications Equipment (DCE): The
packet-switched data network (PSDN) equipment that provides the
functions required to establish, maintain, and terminate a
connection between DTES (see below, under data terminal
equipment) using a physical or virtual circuit.
Data Link: A logical connection between two stations on the
same circuit. In the case of a multipoint link there can be multiple
circuits.
Data Terminal Equipment (DTE): The equipment, typically
a computer system or terminal, that consists of a data source and
sink connected to common carrier communication facilities.
Data Transparency: The capability of receiving, without
misinterpretation, data containing bit patterns that resemble
protocol control characters.
Datagram: In the context of DEenet, a unit of data passed
between the Routing Layer and the End-to-End Communication
layer. When the Routing layer adds a route header to a datagram, it
becomes a packet.
DCE: See above, under data communications equipment.
Designated Router: A routing node (see below) on an Ethernet
chosen to perform additional duties, such as informing each end
node on the Ethernet of the existence of other end nodes and of the
identity of the Ethernet routing nodes. The router chosen to be
the designated router is the one with highest priority, with
highest station address-breaking ties.

260' Glossary

Downline Load: The process of sending a copy of a systemimage or other file over a line to the memory of a target node.
DTE: See above, under data terminal equipment.

EIA: Electronic Industries Association. A standards organization
specializing in the electrical and functional characteristics of
interface equipment.
End Node: A topological description of a non-routing node (one
that cannot forward packets intended for other nodes). An end
node supports only a single active line.
End-User Module: A module or program that runs in a user
process of a network node and communicates with Session
Control software to obtain logical link service.
Error control: The protocol function that ensures the reliable
delivery of data. It typically consists of sequencing, acknowledgment, and retransmission mechanisms.
Ethernet: A local area network protocol using a carrier sense
multiple access with collision detection (CSMAI CD) scheme to
arbitrate the use of a 10 megabits-per-second baseband coaxial
cable.
Events: Occurrences that the Network Management layer logs
for recording.
Executor Node: The node in which the active local network
management function is running (that is, the node actually
executing a Network Management command).
Flow Control: The protocol function that coordinates the flow
of data between two protocol modules to ensure that data is not
lost, buffer deadlocks do not occur, and communication overhead
is kept to a minimum.
Frante: A Data Link layer message as sent or received by an
Ethernet Data Link module or an x.2S frame-level module.

261

Frame Level: LevelZ of the CCITT x.ZS recommendation, which
defines the link-access procedure for data exchange over the link
between the DTE and the DCE.
Framing: The Physical or Data Link layer component that
synchronizes data at the bit, byte, and message level.
Full-Duplex Channel: A channel that provides concurrent
communication in both directions (to and from a station).
Full-routing Node: An implementation of the DNA Routing
layer which contains the full complement of Routing components. A full-routing node performs route-through functions by
forwarding packets intended for other nodes.
Gateway: A module, or set of modules, that transforms the
conventions of one network into the conventions of another.
Half-duplex Channel: A channel that permits two-way communication, but only in one direction at any given time.
Hierarchical Network: A computer network in which computers perform process-control functions at several levels. The
computers at each level are specially suited to perform the functions of that level. In a laboratory application, for example,
computers at the lowest level collect data, which they transmit to
the computers at the next highest level. Here, the collected data is
reduced for analysis, and sent up to the next level of the hierarchical
configuration.
Hop: To the Routing layer, the logical distance between two
adjacent nodes in a network.
Host Node: A node that provides services for another node.
Interface: The relationship between different software modules
that are usually in the same node.
Intra-Ethernet Bit: A special bit, set by the designated router on
an Ethernet, in a packet forwarded from a source end node to a
destination end node that indicates to the destination end node
that the source end node is on the same Ethernet.

262 • Glossary

ISO Reference Model: The International Standards Organization Reference Model for Open System Interconnection, ISO draft
proposal DP7498. A proposed international standard for network
architectures that defines a seven-layer model, specifying services
and protocols for each layer.
Jam: A bit sequence that an Ethernet Data Link module transmits
when it experiences a collision to ensure that all affected stations
detect the collision.
Leased Line: A non-switched circuit that a user (company or
other institution) leases from a public utility company (common
carrier) for exclusive use.
Line: A physical path that provides direct communication among
a number of stations.
Link-Level Loopback: The network management process of
testing a specific data link by sending a message directly from the
Network Management layer of a node to its Data Link layer
(bypassing the intervening layers) and over a circuit or line to a
device that returns the message to the source.
Link Management: The DDCMP component that controls transmission and reception on links connected to two or more transmitters and/ or receivers in a given direction. Also, the Ethernet data
link component responsible for channel allocation (collision avoidance) and contention resolution (collision handling).
Logging: The network management process of recording information from an occurrence that has potential significance in the
operation and/ or maintenance of a network. Logging maintains a
permanent record of the occurrence so that persons or programs
responsible for making realtime or long-term decisions about
network operation can access the record.
Logging Sink Node: A node to which logging information is
directed.
Logical Channel: An association between an x.2S DTE and its
DCE for a given virtual circuit.

263

Logical Link: A virtual circuit between two end-user processes
in the same node or in separate nodes. The Session Control layer
mediates between an end-user process requiring logical link service
and the End-to-End Communication layer, which actually creates,
maintains, and destroys logical links.
Loop Node: A special name for a node that is associated with an
adjacency for loop-testing purposes. The NCP SET NODECIRCUIT
command sets the name of the loopback node.
Master Station: A station that has control of a channel at a given
instant for the purpose of sending data messages to a slave station.
Maximum Cost: An operator-controlled Routing layer parameter that defines the value at which the routing decision algorithm
in a node declares another node unreachable because the cost of the
least costly path to the other node is excessively high (more than
the defined value). For correct network operation, this parameter
must not be less than the maximum path cost of the network.
Maximum Hops: An operator-controlled Routing layer parameter that defines the value at which the routing decision algorithm
in a node declares another node unreachable because the length of
the shortest path between the two nodes is too long (more than the
defined value). For correct network operation, this parameter
must not be less than the network diameter.
Maximum Path Cost: The value of the path between the two
nodes of the network with the greatest routing cost, where
routing cost is the cost of the least costly path between a given pair
of nodes. In Figure 2.11, the maximum path cost is 9.
Maximum Path Length: The routing distance between the two
nodes of the network with the greatest routing distance, where
routing distance is the length of the least costly path between a
given pair of nodes. In Figure 2.11, the maximum path length is 9.

264 • Glossary

Maximum Visits: An operator-controlled Routing layer parameter that defines the value at which the packet-lifetime-control
algorithm discards a packet which has traversed too many nodes.
For correct network operation, this parameter must not be less
than twice the maximum path length of the network.
Message Exchange: The DDCMP component that transfers data
correctly and in sequence over a link.
Monitor: A logging sink that is to receive a machine-readable
record of events for possible realtime decision-making.
Multidrop Line: See multipoint line.
Multiple Line Controller: A controller that can manage more
than one line unit. (Digital's multiple line controllers are also
called multiplexers.)
Multiplex: Simultaneously transmit or receive two or more
datastreams on a single channel.
Multipoint Line: A line linking more than two stations, where
one station is responsible for channel control. Also referred to as
multidrop.
Network: A collection of nodes interconnected by lines.
Network Diameter: The distance between the two nodes of the
network having the greatest reachability distance, where reachability distance is the length of the shortest path between a given
pair of nodes. In Figure 2.11, the network diameter is 3.
Node: An implementation of a computer system that supports
the Routing layer, the End-to-End Communication layer and the
Session Control layer. Each node has a unique node address (see
below, under node address).
Node Address: The unique numeric identification of a specific
node.

265

Node-Level Loopback: The network manageInent process of
testing a logical link by using messages that flow, along with
normal data traffic, through the Session Control, End-to-End
Communication, and Routing layers within one node or from one
node to another and back. In some cases, node-level loopback
involves using a loopback-node name associated with a particular
circuit.
Node Name: An operator-assigned optional alphanumeric identification associated with a node address. There is a strict one-toone mapping between node name and node address. A single node
can be known by different names, but no name maybe used more
than once. The node name must contain at least one alphabetic
character.
Node-Name Mapping Table: A table that defines the correspondence between node names and node addresses or adjacencies.
The Session Control layer of a a node uses the table to identify
destination nodes for outgoing connect requests and source nodes
for incoming connect requests.
Non-Routing Node: A DEC net node that contains a subset of
Routing modules and can send and receive packets but not
perform route-through functions by forwarding packets intended
for other nodes. It is connected to the network by a single active
circuit.
Object Type: A numeric value that may be used, instead of a
process name, for process or task addressing by DEcnet modules.
Object types are normally used to identify generic services, like file
access.

OSI: See ISO reference model.
Other Data: The NSP Data Request, Interrupt Request, and
Interrupt messages. These are NSP messages other than the Data
Segment message. All Other Data messages move in one data
subchannel.

266 • Glossary

Packet: A unit of data to be routed from a source node to a
destination node. When stripped of its route header and passed
to the End-to-End Communication Layer, a packet becomes a
datagram.
Packet Level: Level 3 of the CCITT x.2S recommendation, which
defines the packet format and control procedures for the exchange
of packets.
Packet Lifetime Control: The Routing layer component that
monitors adjacencies to detect if they have gone down, and
prevents excessive looping of packets by discarding those that have
exceeded the maximum-visit limit.
Packet Switching: See route through.
Parallel Data Transmission: A data communication technique
in which more than one code element (a bit, for example) of each
byte is sent or received simultaneously.
Parameters: DNA values to which the Network Management
layer has access for controlling and monitoring purposes.
Passive Side: In the context of MOP loopback tests, the node that
loops back the test messages.
Path: A possible route fQr a packet from source node to destination node. The path can comprise a sequence of connected nodes
between the source and destination nodes.
Path Cost: The sum of the circuit costs along a path between two
nodes.
Path Length: The number of hops along a path between two
nodes.
Peer Protocol: A protocol for communication between modules
in the same DNA layer.
Permanent Virtual Circuit (PVC): A virtual circuit between
two x.2S DTES that is always established. A logical channel is
permanently allocated at each DTE/DCE interface to a pvc.

267

Physical Link: A hardware-addressable communication path.
Piggybacking: The mechanism by which the Data Link layer of
a node of sends an acknowledgment (of data received) to another
node within a returned data or control message.
Pipelining: The mechanism by which the Data Link layer of a
transmitting node sends messages without waiting for individual
acknowledgment of each successive message.
Point-To-Point Circuit: A link connecting only two stations.
Protocol: A set of messages with specific formats and rules for
exchanging the messages.

Raw Event: A logging event as recorded by the source process,
incomplete in terms of total information required
Reachable Node: A node to which a routing node believes it can
direct a packet.
Reassembly: The mechanism by which NSP places multiple
received data segments into a single Session Control receive buffer.
Remote Node: To a node, any other network node.
Request Count: This term has two definitions in the context of
NSP functions: (1) Variables that NSP uses to determine when to
send d;ta and (2) values sent in link-service (Data Request and
Interrupt Request) messages.
To determine when to send data, the flow-control mechanism adds
the request counts received in Data Request and Interrupt Request
messages to the request counts it maintains.
Retranslllission: The process of resending messages that have
not been acknowledged within a certain period of time. This is
part of a protocol's error-control mechanism.

RMS: Record Management Services. This file system is used on
all m<tior systems from Digital, except where space is limited (for
example, RT-ll). In addition to access modes provided by previous
file systems, RMS provides random access for direct and indexed
files, and offers ISAM.

268 • Glossary

Router: See full-routing node.
Route Through: The process by which one or more intermediary node in the path between a source node and a destination node
directs packets froro the source node to the destination node.
Routing nodes permit route-through. Also called packet switching.
Routing: Directing data message packets from source nodes to
destination nodes.
Routing Node: See full-routing node.
Segtnent: The data carried in a Data Segment message. NSP
divides the data from Session Control transmit buffers into
numbered segments for transmission over logical links.
Segtnentation: The mechanism by which NSP divides normal
data in Session Control transmit buffers into numbered segments
for transmission over logical links .
Server: A module or set of modules in a layer that performs a welldefined service, like remote file access or gateway communication,
on behalf of another module.
Server System: A node that contains orie or more server. A server
system is often dedicated to server functions.
Sink Node: A node that receives and records Network Management events.
Slave Station: A tributary station that can send data only when a
master control station polls it or requests it to transmit.
Star Topology: A network configuration in which one central
node is connected to more than one adjacent end node. A star can
be a subset of a larger network.
Station: With regard to the Data Link layer protocol, a termination on a data link. A station is' a combination of the physical link
(communication hardware) and the data link protocol implementation.

269

Station Address: An address assigned at the data link level to a
station.
Status: Dynamic information, such as line state, relating to
network operation. Status is a Network Management information type.
Subchannel: A logical communication path within a logical link
that carries a defined category of NSP data messages. Because Data
Segment messages are handled differently from Other Data
messages, these two types of messages can be thought of as
travelling in two different subchannels.
Switched Virtual Circuit (SVC): A temporary association
between two x.2S DTES.
Synchronous Line: A physical line on which data characters and
bits are transmitted at a fixed rate, with transmitter and receiver
synchronized. On a synchronous line, the Physical Link layer
provides bit framing, and the Data Link layer provides message
framing.
Target Node: The node that receives a memory image during a
downline load, generates an upline dump, or loops back a test
message.
Topology: The physical arrangement and relationship of interconnected nodes and lines in a network. A legal topology satisfies
the requirements of the Routing layer specification.
Transparent Data: Binary data transmitted with the recognition
of most control characters suppressed. DDCMP provides data
transparency because it can receive, without misinterpretation,
data containing bit patterns that resemble DDCMP control characters.
Tributary Station: A station on a multipoint line that is not a
control station.

270 • Glossary

Unit: The hardware controlling one circuit on a multiple line
controller. A unit, a controller, and associated Data Link modules
form a station.
Unreachable Node: A node to which the cost of the least costly
path exceeds the maximum cost of the network, or the length of
the least costly path exceeds the maximum hops of the network.

Up line Dum.p: The process by which a target node sends a copy
of its memory image up a line to a file at a host node.

User: A person or module that requests service from a DNA layer.
Client is the preferred term when referring to modules.
Virtual Call: See switched virtual circuit.
Virtual Circuit: A temporary connection between a data source
and a sink in a network. Virtual circuits typically guarantee
delivery and sequentiality of client data.
Wildcard: With regard to DAP, an asterisk C) that replaces an
element in a file specification. The asterisk specifies all known
items or values in the range that its position indicates. For
example, FILE. * ; * specifies all known types and versions of all files
named FILE.
Window: A range of packets authorized for transmission across
an x.2S DTE/DCE interface.

Index

applications
DEC net file transfer capability
for, 185-89
DEcnet remote file access
capability for, 189-91
DEC net task-to-task capability
for, 181-83
DEC net terminal facilities for
191-96
'
network interaction with
183-85
'
supported by DEC net , 179-81
see also programs; software
architecture, network, 13-14
area-based networks
Levell and 2 routers in, 53
node addresses in, 21
parameters for, 211
routing in, 47-48
area routing, 17
areas (groups of nodes), 3
routing between, 17
ASCII data, 133
asynchronous terminals, 168-69

aborting links, 121
access
on DEcnetlSNA Gateway,
170-73
remote, to files and records
122-35
'
access complete messages, 124
access control
in DEC net , 109-11
in remote file access, 134
in task-to-task
communications, 115
access-control information, 131-32
access methods, 132
in X.25 Gateway Access
protocol, 165-66
access request messages, 124
acknowledgements, 106
Acknowledge Message (ACK),
32
adaptive routing, 17
addresses
node, 21-22, 209
node, mapping of names to, 99
B
tributary, 215
addressing
bandwidth, in phase IV
in Ethernet, 41-43
Ethernet networks, 76
in task-to-task
baseband local area networks, 76
communications, 113-14
in Ethernet, 17, 18
adjacencies, 59
binding, 143, 146
algorithms, routing, 52-53
bisync (binary synchronous)
protocol, 28

272· Index

characteristics (node), 209
character-oriented protocols, 28
checkpointing, 224-25
checksums ofX.25 packets, 64
circuits, 24-26
adapting to different kinds of,
54
controlling state of, 217-18
costs (measurement) of, 49-52
DEcnet, characteristics of, 88-89
DEcnet/ SNA, 177
parameters for, 212-16
in X.25 protocol, 153-55
calls
circuit timers, 148
DEC net task-to-task, 112-13
client layers (Ethernet), 44
to open files, 130-31
coaxial cable, 78
for remote file access, 122, 126
collisions, Ethernet handling of,
Carrier Sense Multiple Access
46-47
with Collision Detect
collision windows, 46
(CSMA/ CD) protocol, 27
Comite Consultatif International
in Ethernet, 39
Telephonique et TelegraCCITT (ComiteConsultatif
phique (CCITT), 18, 151
International Telephonique
X.3, X.28, and X.29
et Telegraphique), 18, 151
recommendations by, 168-69
X.3, X.28, and X.29
command files
recommendations by, 168-69
remote submission of, 135
channels
TLK, 138
collision handling on, in
command nodes, 221
Ethernet, 46-47
commands
logical, 153
in Phone utility, 136-37
in Phase IV Ethernet networks,
in TLK utility, 137-38
76
Virtual Network Processor
subchannels of, 105
(vNP),221

bit-oriented protocols, 28
blocking and deblocking Routing
layer messages, 64
buffers
for flow control, 108
managementoL54
parameters for, 212
Session Control, 106
byte-oriented protocols, 28
byte synchronization, 35

273

command terminal module, 140
command terminal protocol,
143-44
command terminals, 250
communications
Digital Network Architecture
(DNA) for, 5-6
inter-area, 48-49
logical links for, 92
network architecture for, 13-14
between networks, DEcnetl SNA
for, 169-78
between networks, X.3, X.28,
and X.29 protocols for,
168-69
between networks, X.25
Gateway Access protocol
for, 155-67
between networks, X .25
protocol for, 151-55
task-to-task, 112-21
terminal-to-terminal, 136-38
transmission modes for, 26-27
see also networks
communications servers, 18, 82,
86
loopback testing of, 231
in Phase IV, 20
configuration databases, 206, 209
configurations
of DEC net , 71-72
defining, network management
and,209-16

general purpose and dedicated
nodes in, 84-88
line and circuit characteristics
of,88-89
local area networks, 74-81
multipoint, 19
Phase III and Phase IV nodes
in, 82
topology alternatives in, 16-19
wide area networks, 72-74
congestion control, 61-62
connect blocks, 113, 114
contention, Ethernet handling of,
44-47
control messages
in Digital Data Communications Message Protocol
(DDCMP), 31, 32
routing, 56, 58
control stations, 89, 213-15
copying files to remote nodes,
122, 133, 186-87
costs (of path lengths), 49-52
parameters for, 211
counters, 219-20
CTERM protocol, 138, 149
cyclic redundancy checks (CRCS),
68

274 -Index

D
data
segmentation and reassembly
of, in DEcnet, 105-6
in task-to-task
communications,
transmitting and receiving,
119-21
Data Access Protocol (DAF)
Interface, 123-26
handshaking by, 131
databases
configuration, 206, 209
downline-loading, 221
routing, 52-53
in Routing Initialization
sublayer, 63
data circuit terminating
equipment (DCES), 153
data encapsulation/decapsulation,
40,41
data flow, 64-66
across network to destination
node, 68-69
at destination node, 69
at source node, 66-68
datagrams, 68
Data Link layer (DNA), 9
data flow through, 68
in Ethernet, 39, 44, 45
frame and packet levels (X.25)
in, 152
protocols supported by, 13,
27-28

data link mapping interface (DLM),
167
data link mapping mode (DLM),
155, 166
Data Link module, 68
data messages, in Digital Data
Communications Message
Protocol (DDCMP), 29, 31
data packets, 56, 162
data request messages, 105
Data Segment messages, 106
data terminal equipment (DTES),
153-55
in DECSYSTEM-20s, 167
start-stop mode, X.3, X.28, and
X.29 recommendations for
168-69
'
X.25 Gateway Access protocol
used by, 155-65
data types, 133
within logical links, 105
DCES (data circuit terminating
equipment),153
dead tributaries, 215
decentralization, in DEcnet, 4
decision process, in routing, 61
DECNA adapters, 76-77
DEcnet, 1-4
access control in, 109-11
applications supported on,
179-81
capabilities of, 14-16
configurations of, 71-72
data flow within, 64-70

275

DDCMP supported by, 28-37
environments for, 72
error control in, 106-8
Ethernet supported by, 37-49
file transfer capability of,
185-89
flow control in, 108-9
general purpose and dedicated
nodes in, 84-88
internetwork communications
on, 151
line and circuit characteristics
of,88-89
lines and circuits in, 24-26
local area networks, 74-81
logical links in, 92-105
monitoring performance of,
228-55
network and application
interaction in, 183-85
nodes in, 21-23
Phase III, 19-20
Phase III and Phase IV nodes
in, 82
Phase IV, 20
Phone utility used with, 136-37
protocols supported by, 27-28
relocation planning and, 196-97
remote file access in, 122,
189-91
routing by, 47-64,82-83
segmentation and reassembly
of data in, 105-6

task-to-task communications
in, 112-13, 181-83
terminal facilities of, 191-96
topology alternatives in, 16-19
transmission modes for, 26-27
user transparency in, 111
uses of, 4-5
virtual terminals on, 138
wide area networks, 72-74
X.25 access methods for,
165-66
DEcnet-20, 20, 220
DEC net Router servers, 86
DEcnet Router/X.25 Gateway, 5,
87,167
trace capability for, 233-37
DEcnet-RSX, 20
PSI software on, 167
TLK utility on, 137-38
DEcnet-Rsx-l1M,221
DEcnet-Rsx-IIM-PLUS, 221
DEcnet/sNA Gateway, 5, 87-88,
169-78
loading of, 86
SNATRACE for, 235-37
DEcnet/SNA Gateway Access
module, 173
DEcnet/ SNA Protocol Emulation,
175
DEcnet-vAX, 20
downline loading of, 221
PSI software on, 167
Record Management Services
in, 126

276· Index

DEcsystem-10s, 3
DECSYSTEM-20s,3,167
dedicated nodes, 84-88
DELNI,77-78
DEQNA,76-77
DEUNA, 14, 76-77
Digital Data Communications
Message Protocol (DDCMP),
19,27-30
DEcnet/sNA communications
using, 177
initialization and monitoring
functions in, 62
messages in, 31-34
operation of, 34-37
Digital Network Architecture
(DNA),5-6
architecture of, 13-14
Digital Data Communications
Message Protocol (DDCMP)
designed for, 28
interfaces in, 6-10
protocols in, 10-11
directory files, 133
disconnects, 121
DLM (data link mapping) interface,
167
DN20 front-end processors, 166,
167
DNA dump/load protocol, 24
DNA remote console protocol, 24
DNA Routing Functional
Specification, 50
DNA routing protocol, 24

downline loading, 16,220-23
checkpointing and, 224-25
DTES (data terminal equipment),
153-55
in DECSYSTEM-20s, 167
start-stop mode, X.3, X.28,
and X.29 recommendations
for, 168-69
X.25 Gateway Access protocol
used by, 155-65
dumping, 237

E
electronic mail, 194
encapsulation I decapsulation, 40,
41
end hodes, 82, 83
End-to-End Communication layer
(DNA), 9, 49
accepting and rejecting logical
link requests in, 117-19
data flow through, 66-69
logical links created by, 95-99
object types and names declared
in, 116
Session Control layer connections
with,101-3
End-to-End Communication
modules
DEC net calls sent by, 112
environments
application, DEC net capabilities
in, 181-97
DEC net , 72
network,71

277

error detection
in Digital Data Communications Message Protocol
(DDCMP),36
in Ethernet, 43
error handling
by Data Link modules, 68
by DEcnet, 106-8
Ethernet, 37-40
communications servers
supported on, 18
configurations of, 79
DEcnet/ SNA communications
using, 177, 178
initialization and monitoring
functions in, 62
local area networks using, 76
Local Area Transport protocol
on, 148
messages in, 41-43
operation of, 43-47
Phase III and Phase IV nodes
connected to, 82
protocols supported by, 24-26
Ethernet coaxial cable, 77
Ethernet End node Data Packet
format, 56
Ethernet End node Hello message,
58-59,62
Ethernet local area networks, 17
Local Area Transport (LAT)
protocol for, 136
network virtual terminals in,
140
routing on, 55-56

Ethernet protocol, 27
Ethernet Router Hello message,
58,62
Event Logger, 199,201,204-5
executor nodes, 218-19, 221
downline loading by, 222
upline dumping from, 237
extended handshakes, 130-31

F
File Access Listener (FAL), 123,
126,131,134
File attributes messages, 124
file characteristics, 132-33
file organization, 132
file protection, 134
files
DEC net transfer capabilities for,
185-89
in networks, 4
remote access to, 14-15,
122-35, 189-91
file specifications, 131, 135
file systems, 128-30
fixed-length record format, 133
flow control
in DEC net , 108-9
in X.25 Gateway Access
protocol, 159-62
forwarding process, in routing,
61
Foundation layer (NVT) , 141
fragmenting and reassembling
Routing layer Messages, 64

278 • Index

Frame level (X.25), 152
frames, in X.25 Gateway Access
protocol,164-65
framing, 34-35, 40, 41
full-duplex mode, 26, 105
full-function nodes, 83

G
Gateways
DEcnet Router/X.25, 87
DEcnet/sNA, 87-88, 169-78
in Phase IV, 20
PSI software for, 166-67
SNATRACE for, 235-37
supported by DEC net , 5
4.25 Gateway Access protocol,
155-65
general purpose nodes, 84-88
loopback testing of, 231

H
H4000 transceivers, 77, 79
half-duplex mode, 26, 35
handshake procedures, 94, 130-31
hardware
adapters, 76-77
loopback testing of, 229-37
HDLC (high-level data link control),
28
headers, in packets, 152
Hello and Test message, 58, 59,
62

hops, 49
parameters for, 211
host nodes
communications servers
downline-Ioaded from, 86
on Ethernet, 80-81
in Local Area Transport
protocol, 148
in network virtual terminal
function, 141

I
IBM systems, communications

with, DEcnetl SNA Gateway
for, 5, 87-88, 169-78
identification
oflines, 212
node names, 22-23, 99, 209
of target nodes, 114
of users, 131, 132, 134
image data, 133
indexed file organization, 132
Information Frames (I Frames;
X.25),164
initialization and circuit monitor,
62
Initialization messages, 58, 59, 62
Intel Corporation, 39
interfaces
Data Access Protocol (DAP) ,
123-26
in Digital Network
Architecture (DNA), 6-10

279

International Standards
Organization (ISO), 5, 28
International Telephone and
Telegraph Consultative
Committee (CCITT), 18,
151, 168-69
interrupt data, 119, 121
interrupt messages, 105, 106
in X .25 Gateway Access
protocol, 162
interrupt request messages, 105

J
jams, 46

K
Levell routers, 17,48-49,52,53
as full function nodes, 83
Level 2 routers, 17,48-49,52,53
as full-function nodes, 83
levels, in X.25, 152
lines, 24-26
controlling state of, 217-18
cost parameters for, 211
DEcnet, characteristics of, 88-89
identification of, 212
in wide area networks, 72-74
link access procedures (LAP and
LAPB), 163
link identifiers, 114
link management, 35, 40
links, 183-85

Link Service Functions, 201
Link Watcher, 200
local area networks (LANS), 3,
74-81
Ethernet, 37-40
Ethernet, routing on, 55-56
network virtual terminals in,
140
Local Area Transport (LAT)
protocol, 86, 136, 148-49
Local Network Management
Functions, 200
local repeaters, 78-79
logging on to remote nodes, 139
logical channels, 153-55
logical links
access control for, 111
creation of, 94-95
data types within, 105
in DEcnet, 92-95
identified by programs, 103-4
loopback testing of, 233
multiple, within programs,
104-5
Network Services Protocol
and,95-99
for remote file access, 130
Session Control layer and, 100
for task-to-task
communications, 113-15,
117-19
terminating, 121

280· Index

Loopback Access Routines, 201
Loopback Mirror, 201, 205, 231
loopback protocol, 24
loopback testing, 229-37

M
MAIL command, 194
Maintenance Functions, 200
maintenance messages, in Digital
Data Communications
Message Protocol (OOCMP),
31,34
maintenance mode, 37
Maintenance Operation Protocol
(MOP), 34, 205
messages for, 224-27
to monitor node operation,
218-19
management, see network
management
mapping
data link mapping (OLM) , 166
of node names to node
addresses, 99
maximum path cost parameter,
211
memory, downline loading of,
224
message exchange, 35-37
messages
in command terminal protocol,
144-45

Data Access Protocol (OAP) ,
123-26
data flow of, 66-69
oEcnet/sNA Gateway Access,
172-74
in Digital Data Communications Message Protocol
(oocMP),31-34
in Ethernet, 41-43
Ethernet, routing of, 55-56
Event, 204-5
MAIL command for, 194
Maintenance Operation
Protocol (MOP), 224-27
network management, 201-5
Network Services Protocol
(NsP),96-98
polling, 89
routing, 56-59
Session Control, 103
in X.25 Gateway Access
protocol,155-57
MicrovAX I systems, 3
MicrovMs, 3
mode management, 143
modems, 74
modes, in network virtual
terminal service, 143
monitoring
OEcnet performance, 228-37
network operation with
Observer (software), 250-55
node activity, 218-20

281

multipoint configurations, 19
multipoint lines and circuits,
88-89
parameters for, 213-16

N
names
node, 22-23, 209
node, mapping of addresses to,
99
of objects (network programs),
116-17
NCP (Network Control Program),
199-200
loopback testing in, 230-35
negative acknowledgements, 106
Negative Acknowledge Message
(NAK),32
NET utility, 138
Network Application layer (DNA),
7
Data Access Protocol (DAP) in,
123
DEcnet/ SNA Gateway Access
functions in, 170
network virtual terminal
services in, 139-40
X.25 Gateway Access protocol
in, 155
Network Control Program (NCP),
199-200
loopback testing in, 230-35

Network File Access Routines
(NFARS), 128
Network File Transfer (NFT)
utility, 126, 133
Network Generation Planning
Aid,208
Network Information and
Control Exchange (NICE),
200
messages handled by, 201
network interfaces, 153
network management, 15-16,
198-99
defining configuration and
parameters, 209-18
downline loading and, 220-25
generating network software,
208
messages for, 201-5
monitoring node activity,
218-20
node generation planning in,
205-8
in Phase IV, 20
utilities for, 199-201
Network Management Access
Routines, 200
Network Management layer
(DNA),7
node activity monitored by,
220
Routing initialization sublayer
database maintained by, 63
utilities in, 199-201

282· Index

Network Management Listener,

200
network management modules,

128
network managers, 2, 198
diagnostics and maintenance
by, 228
Observer (software) used by,

250-53
protocols implemented by, 13
routing parameters defined by,

50
network object parameters, 210
networks
architecture of, 13-14
data flow within, 64-69
DEC net for, 3-5
DEcnet/ SNA Gateway for
communications between,

170-78
definition of, 1
environments for, 71
generating software for, 208
interaction between
applications and, 183-85
interactive terminal-to-terminal
communications on, 136-38
local area, 74-81
Local Area Transport (LAT)
protocol on, 148-49
monitoring and testing
performance of, 228-37
monitoring and testing
performance of, with
Observer (software), 250-55

nodes in, 21-23
planning node generation for,

205-8
remote file access in, 122-23
topology alternatives in, 16-19
virtual terminals on, 138-48
wide area, 72-74
X.3, X.28, and X.29 protocols
for communications
between, 168-70
X.25 Gateway Access protocol
for communications
between, 155-67
X.25 protocol for
communications between,

151-55
see also communications
Network Services protocol (NSP) ,

95-99
flow control in, 108-9
segmentation and reassembly
of data by, 106
network specifications, 113, 115
network terminal communications, 15
network virtual terminals (NVTS) ,

15, 138-47, 189-90
nodes, 1,21-23
addresses and names for, 41-43,

99,209
characteristics of, 71
collision handling by, in
Ethernet, 46-47
data flow between, 64-69

283

DEcnet, 3, 81-88
downline loading of, 220-21
lines and circuits connecting,
24-26
loopback testing of, 229-37
monitoring activity of, 218-20
in multipoint configurations,
19,89
network virtual terminals as,
138-39
operation of, network
management and, 216-18
in packet-switched data
networks (PSDNS), 153
passwords for, 209-10
in Phase IV Ethernet networks,
76
planning generation of, 205-8
protocols for communications
between, 10-11
remote file access by, 122-23
routing between, 47-49
Rsx-llS, downline-Ioading
and checkpointing tasks on,
224-25
server and host, 79-81
X.25 access methods for,
165-66
normal data, 105, 119-20

o
objects (network programs),
114-17
Observer (software), 228-29,
250-54

open systems interconnection
(osl),5
operating systems, 3
DEC net Phases implemented
on,82
file systems dependent on, 130
logical connections between
nodes under, 138 .
phase IV supported by, 20

p
packet assembly/disassembly
facility (PAD), 168-69
Packet level (X.25), 152
packet lifetime control, 62
packet-mode DTES, 153
Packetnet System Interface (PSI)
software, 166-67
PAD emulation in, 169
trace capability in, 233
packet route headers, 56-57
packets, 21-22
forwarding of, 53-54
routing of, 52
types of, 160-61
packet-switched data networks
(PSDNS)
physical lines and virtual
circuits in, 24
routing in, 56
start-stop mode communications with, X.3, X.28,
and X .29 recommendations
for, 168-69

284 • Index

X.25 access methods for,
165-66
X.25 protocolfor, 18, 151-53
parameters
circuit, 212-16
defining, network management
and,209
downline-loading database,
221
network object, 210
routing, 50, 211-12
passwords, 109, 131, 134
node verification, 209;"10
path length, 49
paths, parameters for, 211
PDP-ll systems, 3
peer relationships, 3
performance of DEcnet,
monitoring and testing,
228-37
with Observer (software),
250-55
peripherals, in networks, 4
permanent virtual circuits (pvcs),
154
Phase III, 19-20
counters used in, 219
node characteristics in, 82
nodes logically connected by,
138
Observer (software) on, 250
Phase IV, 20
communications servers
implemented on, 86
counters used in, 219

Digital Data Communications
Message Protocol supported
by, 19
Ethernet in, 17,37-39, 76
logical terminal to remote
terminal· connections under,
138
network virtual terminals in,
15
node characteristics ih, 82
Observer (software) on, 250
X.25 protocol supported by,
18
Phase IV Data Packet format, 56
Phone (utility), 136-37
Physical layer modules, 68
Physical level (X.25).152
Physical Link layer (DNA), 10
point-to-point lines and circuits,
88
polling, 89
polling ratios, 215-16
portals, 143, 146
positive-acknowledgement-withretransmission protocols,
35-36
P / os operating system, 3
Postal Telephone and Telegraph
Authorities (PTTS), 151
primary loaders, 222
Primary Logical Unit (PLU), 173
process-to-process communications Session Control layer
for, 99-101

285

PRO/DEcnet, 20
DELNI,78
in Ethernet LANS, 81
Professional 350 personal
computers, 3, 20
programming
network terminal facilities used
for, 139
of remote file access, 126-28
programs
DEC net calls in, 112-13
file transfers by, 188-89
handshake procedures between,
94
links terminated by, 121
logical links between, 92
logical links identified by, 103-4
multiple logical links within,
104-5
object types and names for,
116-17
remote file access by, 122-23
see also applications; software
protocols
command terminal, 143-44
CTERM,149
Data Access Protocol (DAP) ,
123-26
Digital Data Communications
Message Protocol, 19
in Digital Network Architecture (DNA), 10-11
Event Logger, 204-5
Local Area Transport (LAT), 86,
136,148-49

Loopback Mirror, 205
Maintenance Operation
Protocol (MOP), 205
Network Information and
Control Exchange (NICE),
200,201
Network Services Protocol
(NSP),95-99
supported on Data Link layer,
27-28
supported on Ethernet, 24-26
terminal communication,
141-43
X.25, 18, 151-55
X.25 Gateway Access, 155-65
PSI (Packetnet System Interface)
software, 166-67
PAD emulation in, 169
trace capability in, 233

Q
Q-bus structures, DEQNA for, 76

R
random access files, 132
Receive Data Decapsulation
(Ethernet),45
Receive Link Management
(Ethernet),45
receive process, in routing, 61
receiving
without contention, in
Ethernet, 45-46

286· Index

data, in task-to-task
communications, 119-21
record attributes, 132
record formats, 133
Record Management Services
(RMs),126
records, remote access to, 14-15,
122-23
relative file organization, 132
relocation planning, 196-97
remote file and record access,
122-23
access control in, 111, 134
accessing remote files from
terminals for, 133-34
in application environment,
189-91
Data Access Protocol (DAP) for,
123-26
file protection in, 134
file specifications for, 135
programming, 126-33
remote command file
submission for, 135
remote repeaters, 79
remote systems, monitoring,
218-19
repeaters, 78-79
Reply to Message Number (REP),
32
reserved object types, 117
resets, in X.25 Gateway Access
protocol, 162-63

routers, 48-49
routes, 47
routing, 17,47-55
in DEC net , 82-83
in Ethernet LANs, 55-56
messages for, 56-59
operation of, 59-64
parameters for, 211-12
in Phase IV, 20
routing control messages, 58
Routing Control sublayer, 59
Routing Initialization sublayer,
59.63
Routing layer (DNA), 9, 49
data flow through, 69
messages in, 64
segmentation and reassembly
of data by, 106
routing messages, 58
Routing modules, 52-54. 68
routing timer. 211
RSTS/E operating system, 3
Rsx-11M operating system, 3
Rsx-11M-PLUS operating system, 3
Rsx-lls operating system, 3,
224-25
RT-ll operating system, 3

s
SDLC, DEcnet/sNA communications using, 177
segmentation and reassembly of
data, 105-6

287

sequential access flies, 132
sequential file organization, 132
server nodes
on Ethernet, 79-80
in network virtual terminal
function, 141
servers, 86-87
Session Control buffers, 106
Session Control layer (DNA), 9,
99-101
data flow through, 69
extended by terminal
communication protocol,
141
nodes names in, 22-23
Session Control modules, 66, 69
SNA (Systems Network
Architecture; IBM), 5, 88
DEcnet/SNA Gateway for,
169-78
SNATRACE, 235-37
software
DEcnet, logical links and, 95
DEC net calls in, 112-13
DNA interfaces, 6-10
network, generation of, 208
for network diagnostics and
maintenance, 228-29
PSI,166-67
see also applications; programs
source programs, 122-23
Start Acknowledge Message
(STACK),32
Start Message (STRT), 32

start-stop mode DTES, 153
X.3, X.28, and X.29
recommendations for,
168-69
states
of lines and circuits, 217-18
of nodes, 216-17
status paramenters, 209
stream record format, 133
subchannels, 105
subroutines, for remote file access,
126-28
Supervisory Frames (S Frames;
X.25),164
switched virtual circuits (svcs),
154-55, 163
synchronous disconnects, 121
system commands, 133
system images, downline loading
of,224
system managers, 2, 198
loopback testing by, 229, 230
node names assigned by, 22, 23
upline dumping by, 237
Systems Network Architecture
(SNA; IBM), 5, 88
DEcnet/ SNA Gateway for,
169-78

288· Index

T
target-initiated downline loads,

222_
target nodes, 221
downline loading of, 222, 224
identifiers for, 114
target programs, 123
target system images, 221
task-to-task communications, 14,

112
accepting and rejecting logical
link requests in, 117-19
access control in, 109-11
in applications environment,

181-83
in DEcnet, 112-13
logical links for, 113-15
object types and names for,

116-17
terminating links in, 121
transmitting and receiving data
in, 119-21
terminal communication module,

140
terminal communication protocol,

141-43
terminal communication services
module, 146
terminal management module,

146
terminals
access control for, 111
accessing remote files from,

133-34

communications between, 15,

136-38
DEC net facilities for, in
application environment,

191-96
Local Area Transport (LAT)
protocol for, 148-49
remote file access from, 122
virtual, 138-47
Terminal Servers, 86-87
Local Area Transport (LAT)
protocol for, 136, 149
terminating links, 121
for network virtual terminals,

147
in X.25 Gateway Access
protocol, 163
testing DEC net performance,

228-55
throughput, in Phase IV Ethernet
networks, 76
TLK (utility), 137-38
topology of networks
adapting to changes in, 54
DEcnet, 16-19
TOPs-l0 operating system, 3
TOPs-20 operating system, 3,

166-67
trace capability, 233-35
transceivers, 77
transmission
without contention, in
Ethernet, 44-45

289

of data, in task-to-task
communications, 119-21
in Ethernet, 44-45
modes of, 216
over wide area networks, 74
transmission modes, 26-27
Transmit Link Management
(Ethernet), 44, 46-47
transmit management process, in
routing, 61-62
tributaries, 89, 214-16

u
UNIBUS structures, DEUNA for, 76
Unnumbered Frames (U Frames;
X.25),164
update process, in routing, 61
upline dumping, 16, 237
User layer (DNA), 7
Network Control Program
(NCP) in, 199
Network File Transfer utility
in, 126
users, 1-2
access control of, 109
identification of, 131, 132, 134
user transparency, 111
utilities
NET, 138
network management, 199-201
Phone, 136-37
for remote file access, 126-28
TLK,137-38

v
variable-length record format,
133
variable-with-fixed length control
record format (VFC), 133
vAx-ll PSI software, PAD
emulation in, 169
vAx-ll X.25/X.29 Extension
Package (XEP), 167, 169
VAXELN,3
VAX systems, 3
VAX/VMS Command Lanaguage
Interpreter, 128
VAX/VMS operating system, 3,
136-37
verification, node passwords for,
209-10
Verification messages, 58, 59, 62
virtual circuits, 24
in X.25 Gateway Access
protocol, 162
in X.25 protocol, 153-55
Virtual Network Processor (VNP)
commands, 221
virtual terminals (NVTS), 15,
138-47, 189-90

w
wide area networks, 3, 16-17,
72-74
network virtual terminals in,
140
windows (packets), 159-62

290 • Index

x
X.3 recommendation (CCITT),
168-69
X.25 Frame-Level module, 163-64
X.25 Gateway Access module,
157
X.25 Gateway Access protocol,
155-65
X.25 Gateway Server module,
157-59
X.25 Packet-Level module, 159
X.25 protocol,S, 18,27, 151-55
checksums in, 64
initialization and monitoring
functions in, 62-64

PSI software and, 166-67

routing in, 56
X.3, X.28, and X.29
recommendations and, 169
X.28 recommendation (CCITT),
168-69
X.29 recommendation (CCITT),
168-69
X.29 Virtual Terminal Driver
168
'
XEP (vAx-ll X.25/X.29
Extension Package), 167
PAD emulation in, 169
Xerox Corporation, 39

Notes