Description of computers Jim managed in California

Bet you want to know what a systems engineer uses at home and work, right?

Well, at home I am running Windows 2000 on an Athlon 1.4 GHz machine with a ¼ GB of main memory. That blows my mind. When I started working with mainframe computers in the early 80s, we ran a machine with 256 KB of memory that supported 100 people! The computer field has changed vastly in my working lifetime. And adapting has always been fun.

This PC is usually 100% busy running SETI@Home analysis, using the new BOINC version of the client. I completed nearly 10,000 work units in the original SETI project. You can see my stats via the BOINC version here. Even more detailed BOINC stats are available.

You can pop up a skymap showing the locations that the Athlon processed data from in the original project. You can also pop up an older skymap showing the datapoints prior to July 16th, 2001.

In addition to the Athlon, I have an AlphaServer 800 5/400 running version 8.2 of the OpenVMS operating system. I'm a specialist in this operating system, and am currently looking for a position in Sydney that will utilize my skills. In its spare time, the Alpha used to crunch original SETI data. So far, noone has come up with a port of the new Boinc version of SETI for OpenVMS. I've looked at the code, but because of the design decisions that the Berkeley team made, unless OpenVMS Engineering comes out with a full implementation of fork(), BOINC won't port any time soon (unless some brave soul wants to convert it to pthreads).

I also have a brand new AMD 2800+ Semperon system that I just purchased to replace a machine I originally scavanged from a dumpster (!) This machine is running Ubuntu linux. While I'm a VMS guy all the way, I use this machine to develop web stuff targetted at unix boxes. For example, my professional web site is developed like this as I use a web hosting company that runs unix boxes. Again, while not busy doing anything else, this box also crunches SETI units.

All of the machines report uptime to The Uptimes Project. You can see their stats by going to my uptimes account page.

The machines at my previous employer were large. I mean LARGE. There were nine main clusters of machines running OpenVMS: Development, Staging, Production, Data warehousing, FTP, EDI, a cluster that records Syslog info, a legacy cluster of VAXen that I am not allowed to mention, and an online data store (which is just another data warehouse). The basic building blocks for these clusters were AlphaServers ranging in size from DS10s with one CPU to GS160s and GS1280s with up to sixteen CPUs and 64 GB of memory.

In addition to this, there were a bunch of standalone OpenVMS systems are used for things like console control, and performance analysis.

Oh, in case you're wondering, none of these machines ran SETI *grin*

The disk farm on these suckers was huge. When I left, I'd estimate that there was about 140 terabytes of usable disk on the OpenVMS systems. That's 1,400,000 gigabytes. We're talking around 1,500 individual disk drives. The storage was connected in a variety of ways, but mostly via 1, 2, and 4GB fibrechannel.

All this stuff supported around 4,000 production end users.

Also, I installed a couple of rx2620 Itanium-based boxes running OpenVMS 8.2-1 before I left to demonstrate the portability of the application.

Here are a couple of interesting computer links:

Jim's Home Page