Murphy attacks, is deflected :-)

Aug 21, 2006 20:09

I never power down my machine at work; I log in on Monday morning and reboot on my way out on Friday. But I was going to be away for almost two weeks, and I know that sometimes IT departments get cranky about leaving machines on and unattended, so I did the responsible thing and shut down before leaving for Pennsic. I know, I know -- I'll never do it again.

This morning I went in, pushed the button -- and nothing happened. Hmm, I said -- is that button merely a "soft" reboot, and for this I need to do something more drastic? After a hard power cycle there was still no activity, so off I went to find an IT person. (I have no idea what happened. Minimal diagnostics implicated the motherboard.)

Our corporate overlords have a policy of not maintaining our pre-existing machines (and maybe not even their own; I can't tell). When a machine breaks, their answer is to replace it. Ok, said I, do we have a buffer -- a machine I can take now? Well no, we don't order until there's a specific need, and it takes a couple weeks. (We don't even have loaner laptops any more.)

Think about that for a moment. A 70,000-person company has a policy of idling their employees for significant periods of time. At least officially; clearly it was time to use unofficial means to solve this problem.

"Fortunately" (and I never thought I'd use that word in the context of this event), two developers just left; their last day was Friday. Clearly I should appropriate one of their machines, either temporarily (while waiting for a new machine) or permanently (forget the new box; this'll be fine). It took me a while to get everyone who might have a stake in those machines to say ok, but finally I got permission to take one.

I was not looking forward to the process of clobbering the old environment and setting up my own. Configuring a clean machine takes a couple days (at least) to get everything right; tweaking an existing machine takes longer unless you start by wiping the disk, and I no longer have access to OS disks.

But this is where Murphy blinked. Our friendly support guy (contrator, not actual overlord) -- who did me a favor he didn't have to do here -- popped my disk in as a secondary drive, but it didn't work on the first try (probably missed some jumper settings). When he opened it back up, I asked if he could swap the drives and just boot from mine. That should work, I said, modulo some device drivers, right? He said "maybe" and gave it a shot. He had the bright idea of moving the graphics and network cards at the same time (which reminded me that I had a "good" graphics card, an upgrade from the standard issue, and keeping that would be helpful). Voila -- that worked. I haven't tried to do anything with the CD drive, and I didn't notice whether it's a burner or just a reader, but that can wait. The important thing is that I don't have to re-configure everything. Instead of costing me days, this cost me hours, due largely to a contractor willing to go beyond the rules and a conveniently-timed departure.

I did send off mail to the head of our group suggesting that we really, really need to keep a spare machine or two around for such emergencies. A former co-worker limped along on a failing machine for three months before his departure because that was just easier than dealing with getting a new one. That sort of thing shouldn't be necessary. Hardware is cheap compared to developers.

A past employer got this sort of thing right. They settled on a standard hardware configuration and gave everyone removable hard drives (that is, you didn't even have to pop the case). If your machine died, you took your drive to a spare and were up and running again in 10 minutes. Smart.

viz

Previous post Next post
Up