Digital PDFs

AN-013

November 2000

2 pages

Original

78.0kB

Document:	Serial line & LAT terminal performance in CHARON-VAX
Order Number:	AN-013
Revision:	0
Pages:	2
Original Filename:	AN-013.pdf

OCR Text

CHARON-VAX application note
AN-013 Serial line & LAT terminal performance in CHARON-VAX
Author:

Software Resources International

Date: 8-Nov-2000

The performance of serial line emulation in CHARON-VAX using LAT terminal servers was tested on a
Windows 2000 system. The performance was measured with different numbers of lines and transfer
speeds, with both single and dual CPU hosts.

Configuration used
Component
Base system

Disks
Network
Peripherals
LAT devices

Description
Intel P2 system 500 Mhz, 1CPU/2CPU’s
64Mb RAM
S3 video board
SCSI adapter NCR
FDD 3.5”
IBM 9 GB IDE ATA
4 port card on Digital 143 chipset
VT100 Terminal
DECserver 90L
DECserver 90L+

Software
W2000 installed
CHARON-VAX build 11778

With loop-back connector

Test procedure
Two types of tests were each performed on two hardware configurations (A dual CPU system and a
one CPU system) for 3 different speeds (38400bps, 19200bps and 9600bps) and a different number of
lines (8,16 and 32) in loop-back mode. The first type sends files with a fixed size to the LAT terminals.
The second type copies a number of 32 bytes blocks to the LAT terminals.
Test №
1.1
1.2

Test Title
Testing LAT terminals by copying a test file to each of the terminals with COPY command
Testing LAT terminals with TSTLAT utility (32 byte blocks)

The time for 1.1.test passes was found manually using the system clock; several measurements
(usually 7-10) were done. The speed & time for them indicated in the following tables are averages.
The 1.2 tests reported the time and performance automatically.

Test conclusion
1. LAT works well on both single and dual CPU host systems. Up to 32 terminal lines are handled
without problems.

2. The LAT throughput is approximately 1.5-2 times larger on a dual CPU system compared to a
single CPU host.

©2000 Software Resources International. This document is provided for information only and is not a legally binding offer.
Software Resources International reserves the right to change the product specifications without prior notice or retire the
product. The CHARON name and its logo are a registered trademark of Software Resources International. For further
information: www.charon-vax.com, Email: vaxinfo@vaxemulator.com

CHARON-VAX application note AN-013
Test Results for a dual CPU system
LAT Terminal server; every port set at 38400 bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
49
67
72

Speed
(Bytes/s)
3249.63
2376.60
2211.56

Bytes
159232
159232
159232

Test 1.2
Time (sec)
50
72
150

Speed
(Bytes/s)
3184.64
2211.56
1047.58

LAT Terminal server; every port set at 19200bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
92
103
103

Speed
(Bytes/s)
1730.78
1545.94
1545.94

Bytes
159232
159232
159232

Test 1.2
Time (sec)
85
85
154

Speed
(Bytes/s)
1873.32
1873.32
1033.97

LAT Terminal server; every port set at 9600bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
172
172
191

Speed
(Bytes/s)
925.77
925.77
833.68

Bytes
159232
159232
159232

Test 1.2
Time (sec)
170
170
170

Speed
(Bytes/s)
936.66
936.66
936.66

Test Results for a single CPU system
LAT Terminal server; every port set at 38400 bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
82
127
155

Speed
(Bytes/s)
1941.85
1253.80
1027.90

Bytes
159232
159232
159232

Test 1.2
Time (sec)
86
145
250

Speed
(Bytes/s)
1851.53
1098.15
632.92

LAT Terminal server; every port set at 19200bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
93
122
155

Speed
(Bytes/s)
1712.17
1305.18
1027.30

Bytes
159232
159232
159232

Test 1.2
Time (sec)
88
149
240

Speed
(Bytes/s)
1809.45
1068.67
663.47

LAT Terminal server; every port set at 9600bps
№ of
lines
8
16
32

Bytes
159232
159232
159232

Test 1.1
Time (sec)
174
173
225

Speed
(Bytes/s)
915.13
920.42
707.70

Bytes
159232
159232
159232

Test 1.2
Time (sec)
176
174
248

Speed
(Bytes/s)
904.73
915.13
642.06

Note: The difference between the Tests 1.1 and 1.2 is a result of the VMS processing of “submit”
requests, impact of VMS internal queues, etc. Because of that, the test 1.2 gives a more realistic
result. At 32 ports @ 9600 bps, the port speed is still the limiting factor for a dual CPU host.

Page 2 of 2