Dispatch from the geeks, 2006/08/18

Greetings!This is the first in what will be a regular series of updates from Systems to the division. The primary goal here is to keep our users informed of what’s going on in the IT infrastructure they rely on. To that end, we’re going to be publishing these updates as warranted — possibly as much as weekly, but likely once or twice a month. There may be incremental updates at the Systems Blog (http://www.mcs.anl.gov/systems/blog), but we’ll include these in any mailings. This blog will also serve as an archive of these newsletters.Current topics: Web server migration, potential networking changes, and new linux builds in testingWeb server migration: With Pedro leaving, it throws a wrench into our plans to get things migrated over to the new server, but we’re still going to press ahead. We ask that you have a little patience with us, since we’ll be running short handed for a little while, but we’re doing our best to get things back on an even keel as quickly as possible. Once we get new hard drives installed in the web server, we should be able to start migrating users and projects at a fairly reasonable pace.If you’ve already sent us a req about being moved, you’re in the queue. If you have an urgent need to be moved, let us know.Networking changes: We’re testing a new way of doing things on the network. Right now, we’re in the very early testing stages, but I wanted to keep everyone informed of where we’re thinking of going with this. But first, a little background.Right now, all the networking in MCS (except Teragrid) connects through one of two routers, which then connect to the external network, which allows us to access all the lovely internets. Our firewall is implemented as a set of router rules on these routers. In general, it works okay, but network scans, malicious attacks, and occasional malfunctions can bring our network to its knees; I’m sure many of you have felt these instances when they’d occurred in the past couple of months. Compounding this effect is the age of our equipment.We’ve been presented with an opportunity to kill two birds with one stone. In exchange for moving our production office network beind the lab’s firewall, we’ll be given new equipment to improve our routing infrastructure. What this means for you right now is nothing — we’re still testing the effects of doing this. Should we implement this plan, we should be clear that this won’t affect our HPC systems and clusters, or really anything that isn’t considered our production office network. Well, that’s a lie… we’ll get better bandwidth for those parts of the infrastructure.Currently, a number of systems members are living behind this firewall. We’ll be looking for some friendly testers in the next week or so, so if you’re game, let us know. I should state that it is our goal that users won’t even notice this change if it happens, short of improved reliability. And we are especially not going to do something that’s going to break things.New Linux Builds: We’re testing a new linux build developed by Pedro (*sniff*). It’s available in both 32 bit and 64 bit builds, and is based on the Ubuntu distribution Dapper Drake. We’re in early testing with this as well (dogfooding it, mostly), but will be looking for friendly testers soon. If you’re interested, let us know.Well, that’s it for this update! Until next time, I leave you with this. There are two types of people in this world: those who finish what they begin,