Sorry about the downtime today. I upgraded to Apache 1.3.26 to close a denial of service hole (since I never, ever write anything the least bit controversial–ahem–except on days that end in -ay) and then I neglected to restart it.
Welcome back to your normal, everyday life, Dave.
Incidentally, last week’s outage appears to have been due to a power failure. Steve DeLassus recognized it and e-mailed me in vain, but seeing as I’d sworn off e-mail for the week it didn’t do much good. I’m not overly concerned about it; my Linux servers’ uptime is measured in years as long as Ameren keeps the current flowing.
I’ve already decided that I’ll never get those 1 or more year uptimes that some report, no matter how well I tune my Linux boxes. Why? Well, in Houston, or at least my part of it, we seem to go a maximum of a few months between power outages long enough to take me down.
Once you exceed the runtime of my UPSs I’m down. Where I work we have a huge diesel with a huge fuel tank, but I don’t think I’ll be installing one of those at home.
I believe that there is Linux software that will work with my UPSs (1 APC 1 Tripplite, plus an APC on my Win2k box) to auto-shutdown given a long enough power failure, rather than just letting the boxes crash. Another thing that I need to find a round tuit for.
We had that problem at my first job. Our OS/2 servers got rebooted once a year whether they needed it or not because we got right around one power failure every year, like clockwork, and we never knew the duration so we had to shut the machines down.
Other than that, we only had one OS/2 server that ever crashed. It was a pre-production IBM PS/2 Model 80 with a couple of 300-meg RLL drives in it. (We’d been a beta test site for that machine, and IBM didn’t want the machines back after the testing period so they went into production and stayed there.) The machine was built in 1987. This was 1997 or 1998. Somehow, I don’t think the problem was OS/2…