You can contact me via e-mail at:
- parris at encompasserve dot org
- keith dot parris at hp dot com
- keithparris at yahoo dot com
My areas of specialty are in using clusters of computers (particularly those running the
VMS operating system) in high-availability and disaster-tolerant configurations, in troubleshooting difficult and hard-to-solve problems, in scaling clusters to meet the incredible growth rates of businesses in areas such as e-Commerce and cellular telecommunications, in performance analysis, and in I/O issues and storage configurations, including SANs.
I presently work for Hewlett-Packard as a Systems/Software Engineer in HP Services, Global Solutions Engineering, Multivendor Systems Engineering, doing consulting and fly-and-fix work for critical customer sites.
Summary of background and experience
DECUS session notes and other publications
DECUS stands for 'Digital Equipment Computer Users' Society', the
users' group covering the former Digital marketplace. After the purchase of Digital by Compaq in 1998, the U.S. Chapter of DECUS changed its name to
Encompass. Compaq was purchased by HP in 2002. Encompass holds a yearly HP Enterprise Technical Symposium. Other chapters of DECUS exist around the world, and hold symposia in their geographical areas as well. In Europe, the former DECUS, later Compaq Users Organization, has merged with the pre-merger HP user's group, Interex, and they hold a yearly symposium called HP Enterprise Users Week. Interex in the USA holds a yearly symposium called HP World.
Here are the session notes for some of my sessions at these conferences:
Join the OpenVMS WebRing
- To attempt to answer many of the questions I got while I was the developer maintaining the VMS Connection Manager code, I presented the session
Understanding VAXcluster State Transitions, which was aimed at helping folks minimize the duration of and impact on cluster users of state transitions (which occur when a node joins or leaves a cluster). I also have a version with speaker's notes. These were written in 1992. I also have slides providing updated information as of
Spring 1997, two occasions when I presented the material again.
- While I was in VAXcluster Systems Engineering, and continuing while I worked in VMS Engineering, I had the challenge of helping Ben Brock of Dow Chemical in Houston build what we believe is still the record-holding VMS cluster in terms of numbers of nodes: it grew to a peak of 151 nodes around 1993. Based on what we learned from that experience, I wrote an article in the VAXcluster Systems Quorum Journal on Building Large Local Area VAXcluster (LAVC) Configurations (c. 1992).
- I left Digital in 1996. To help folks considering going out on their own, as opposed to continuing to work as an employee, I presented a session on Starting Out as an Independent Computer Consultant. Here are the versions presented in
Fall 1997, and
- I have worked with quite a number of different disaster-tolerant VMS cluster sites. I presented sessions on BRS User Experiences in
based on what these folks and I had learned in working together in the design, installation, and operation of this type of configuration.
- In the Spring of 1999, I explored Lock Manager Performance and Long-Distance Cluster Performance, which is based on the experience of building at E*Trade what was for a long time was the longest-distance VMS cluster in the world, at a site separation distance of 130 miles (200 kilometers). I also spoke on Disk Partitioning on OpenVMS.
- I gave a presentation on the topic of Comparing Host-Based and Controller-Based RAID in
Spring 2000 and Fall 1999.
- In the Fall of 1999, based on my work with the
International Securities Exchange, which is a new options exchange using software from
OM Group on a disaster-tolerant VMS cluster with RTR (Reliable Transaction Router) software, I presented a session Comparing VMS Clusters and RTR.
- I presented a session on Using Load-Balancing Mechanisms in VMS Clusters in
- In the Spring of 2000, at the Compaq Engineering Update (DFWdays) sponsored by the Dallas-Ft. Worth Compaq Users Group, I reported on recent work in the area of Monitoring and Controlling the VMS Lock Manager
- In October of 2000, at the Compaq Enterprise Technical Symposium, I presented 1-day seminars on VMS Clusters: Basic and Intermediate Concepts and VMS Clusters: Advanced Concepts, 4-hour interactive workshops on VMS Cluster State Transitions in Action and the
VMS Lock Manager in Action, and a session Comparing VMS Clusters and Tru64 Unix TruClusters.
- At the European DECUS conference in Lisbon in the spring of 2001, I taught a 1-day seminar on VMS Clusters: Advanced Concepts and a session on Monitoring and Controlling the VMS Lock Manager.
- At DFW Days 2001 in the spring of 2001, I did a session on Monitoring and Controlling the VMS Lock Manager.
- At CETS 2001 in Anaheim in the fall of 2001, I did a 1-day seminar on VMS Clusters: Advanced Concepts, an interactive workshop on how to use the LAVC$FAILURE_ANALYSIS tool entitled Network Monitoring for LAVCs, and a session on my experiences Riding the Dot-Com Wave: A Case Study in Extreme VMS Cluster Scalability.
- In March 2002, I joined Compaq just 7 weeks before their acquisition by HP was finalized. At the joint DECUS/ITUG Symposium in Lyon, France in May 2002, I did a session on OpenVMS Distributed Lock Manager Performance.
- At the Australian DECUS Symposium in Sydney in July 2002, I presented a session providing an Introduction to Linux Internals for OpenVMS Experts.
- At HP World in Sept. 2002, I presented a 2-hour tutorial on Disaster-Tolerant Cluster Technology and Implementation, which covered DT cluster technology across the entire range of HP platforms, including HP-UX, Linux, Windows, and of course, OpenVMS, and describing both host-based and controller-based replication schemes.
- At the HP Enterprise Technical Symposium in Oct. 2002, I presented on the topic of OpenVMS Disaster-Tolerant Clusters a
4-hour workshop, and
75 minute session. I also presented a seminar on
OpenVMS Cluster Internals & Data Structures, and a session on
Volume Shadowing Performance concentrating on full-copy performance.
- At the New York City HP Local User Group meeting in December, 2002,
I did a presentation on the topic of Disaster-Tolerant Solutions from HP.
- In the premier issue of the HP OpenVMS Systems Technical Journal, I had an article published on the topic of Using OpenVMS Clusters for Disaster Tolerance.
I ran out of disk quota on the site here, so for all presentations after December 2002, see http://www2.openvms.org/kparris/
I live in Colorado, have a beautiful wife and 3 wonderful now-grownup children, and am learning and growing closer to God as a member of the
Seventh-Day Adventist Church.
This page hosted by
Get your own Free Home Page