"HP OpenVMS Technical Journal V2 July 2003"

This document is the HP OpenVMS Technical Journal, Volume 2, from July 2003, comprising a collection of articles related to OpenVMS technologies and solutions.

The journal features several key technical papers:

1. A Survey of Cluster Technologies: This article by Ken Moreau compares various cluster technologies across HP server platforms (HP-UX, Linux, NonStop Kernel, OpenVMS, Tru64 UNIX, and Windows 2000). It defines key terms like scalability, reliability, availability, and manageability, and categorizes clusters into single-system-view and multi-system-view (active-passive, active-active) modes. It delves into cluster file systems, quorum mechanisms, configurations, application support (single-instance vs. multi-instance), recovery methods, cluster resilience, data high availability, and disaster tolerance, offering a comprehensive comparison of how different HP offerings address these aspects.
2. Local Area Network Cluster Interconnect Monitoring: Keith Parris details the LAVC$FAILURE_ANALYSIS tool in OpenVMS for monitoring and troubleshooting LANs used as cluster interconnects. It explains how this tool detects failures via "Hello" packets, uses a mathematical graph of the network, and reports issues via OPCOM messages. The article provides a step-by-step guide on how to implement, maintain, and customize the tool, including gathering MAC address information and editing the template program for specific network configurations.
3. Internet Technologies for OpenVMS: Ken Moreau discusses integrating Internet and web services technologies (Java, XML, Enterprise Application Integration, database connectivity, graphical front-ends for legacy apps, web browsers/servers) with OpenVMS. It highlights how OpenVMS can leverage these modern technologies while retaining its core strengths in security, reliability, and scalability. The paper covers Java APIs (J2EE, JDBC, JMS, JNDI, JSP, JTA, JPDA), data and application integration using XML, Message Oriented Middleware (MOM), and object-based communication (CORBA, COM, SOAP). It also mentions HP's efforts in providing OpenVMS-compatible open-source tools like Apache and NetBeans.
4. Configuring TCP/IP for High Availability: Matt Muggeridge focuses on TCP/IP high-availability solutions for OpenVMS, including failSAFE IP, IP Cluster Alias, and Load Broker/Metric Server. It compares these technologies, outlining their protection scope, protocols, load balancing, and failure detection capabilities. The failSAFE IP service, which provides IP address redundancy and interface health monitoring, is described in detail, including its configuration, detectable failures, application benefits, and management utilities.
5. DCPI for OpenVMS a Technical Introduction to a “System Microscope”: Anders Johansson introduces the Digital Continuous Profiling Infrastructure (DCPI) for OpenVMS, a tool that uses Alpha chip hardware performance counters for fine-grained system profiling. It explains DCPI's components (device driver, daemon, analysis tools) and how it works by sampling events like CPU cycles or instruction fetches. The article outlines how to collect and analyze data using DCPI tools like dcpiprof and dcpilist to identify performance bottlenecks.
6. RMS Performance: Duplicate key chains: Hein van den Heuvel addresses performance issues caused by "duplicate key chains" in OpenVMS RMS Indexed files. It explains how these occur with secondary keys, leading to excessive I/O operations during record inserts. The article describes RMS indexed file internals, methods to detect the problem (e.g., monitoring I/O rates, ANALYZE/RMS/FDL, rms_tune_check), and solutions such as dropping unselective keys, using NULL KEY values, increasing bucket size, and deduplicating key values by adding key segments.
7. A Customer Case Study of Oracle Rdb Database Consolidation: This paper details SIAER's server and Oracle Rdb database consolidation project on an OpenVMS AlphaServer cluster. It describes the benchmark setup, hardware and software configurations, and the testing methodology involving "Osservato" and "Disturb" processes to simulate user activity and workload. The results highlight performance improvements achieved through consolidation and the use of Oracle Rdb's Row Cache technology, validating HP's proposed architecture for enterprise-level applications.
8. Best of the HP Customer Support Center and Best of Ask the Wizard: These sections by Mark Jilson and Steve Hoffman provide guidance on utilizing HP OpenVMS support resources, including collecting system profiles (configuration, hardware setup, system setup, boot, performance) to diagnose performance changes and system bugchecks. They also list various OpenVMS support resources, FAQs, ECO kits, and formal assistance channels.
9. Server-Agnostic Perl/DCL CGI Programming with WASD and OSU: Dick Munroe shares insights on creating server-agnostic CGI programs using Perl/DCL with WASD and OSU web servers on OpenVMS. He discusses challenges and solutions for portability across different CGI execution environments, emphasizing the importance of consistent execution, sane CGI interface libraries, and constant output.

Overall, the journal provides valuable technical information for OpenVMS system administrators, developers, and architects, covering topics from system performance and high availability to application integration and disaster recovery.