Al’s having some IT static at the House of Pain. Sounds like it’s either based in budgetary issues, or in incompetent IT staffers (which is also potentially a budgetary problem, I guess…)
Thinking about how you’d set up the enterprise for a hospital is a fun intellectual exercise. I’m thinking dual live systems (in an active/passive setup) for everything in the “core”, with an on-the-shelf (i.e., racked, cabled, but powered off) spare system for emergency use. Individual workstations being down shouldn’t be that big of a deal, but you want the core stuff to be big and beefy and 24/7/365. The active/passive setup for the live systems means you could do any upgrades or maintenance on the passive half first, test it out and make sure you didn’t have any regressions, then swap the active/passive status and work on the other half — so your downtime would be reduced to sub-minute “blips” during the cutovers, instead of the hours-long outages Al is seening.
The opposite approach would be to radically decentralize everything, push it all out to the edges, and then have a lot of logic for locating bits and pulling them to other places on request — a Google-ish sort of setup. This is an initially appealing idea, but the more I think about it, the more potential downsides I see — and that’s without even thinking about HIPPA.
This topic feeds into another thing I’ve been bumping my nose on recently: the generally poor regard most people have for their IT support. At one point, I think I would have been inclined to attribute that more to the users, but I’m more and more realizing that the problem probably lies more on the IT side. The bright side of that, of course, is that I shouldn’t have any troubles staying gainfully employed; the downside is the frequent wincing and muttering when I hear people talk about the IT situation at their workplaces…