I had to use my ninja-like server management skills this morning. They consist largely of knowing where the power button is, holding it down until the server shuts off, and then tapping it again to light it back up. Solves 99% of problems.

The XServe that drives weblogs.ucalgary.ca, wiki.ucalgary.ca, and pachyderm.ucalgary.ca just went south. Refuses to boot now. It’s a cluster node, so we’re trying to find a video card to see wtf is going on…

Update: OK. We found a trusty old ATI Rage 128 card, slapped it into the XServe, and booted that sucker up. It was spending a looooooong time on the boot spinning pinwheel screen – assuming it’s checking the disk here. It had to chew through ~50GB of data, so that took awhile. Then, it wanted to update firmware, so we did. It’s now fully patched, and apparently running OK. It’s seeming awfully slow, though. It’s not a RAM thing, with 5GB of the good stuff in there. It just took over a minute to connect via SSH. Definitely need to look into that. Perhaps AFTER the big afternoon presentation that relies on the software running on this server…

He’s apparently got a much more complicated setup, since there are several steps that we didn’t have/want to do, and his time estimates for others are longer than I’d have guessed. Regardless, it’s good to see another report of just how easy it is to fully configure and deploy these bad boys.

Also of note are his recommendations for backup strategies. I currently use rsync to backup important stuff to another drive on our commons webserver – but the data has gotten so big there isn’t a safe place for it off that machine. This will get better once we have the XRAID hooked up, but still…