As you should have noticed, we were unavailable for much, much longer than anticipated. Everything seemed okay when we got the new motherboard and CPU running Saturday afternoon, but after a short time (of course after we'd left the computer room), we lost network connectivity.
Willfon spent some time during Saturday and Sunday trying to figure out what went wrong, but was unable to get to the bottom of it, since we seemed to have a problem with the router, and the router people at the University were simply not available until today.
After much debugging of configuration options, which after all turned out to be quite correct, I brought my laptop, a crosslink TP cable and my own 100 Mbit network card, just in case that might help.
In short, yes, it was the network card that was probably fried in one of the recent unannounced (or too poorly announced) power outages, and replacing the network card did the trick.
We are now going to look for a UPS of our own (if we're allowed to install it), and see if there is one that works with Debian that we can rely on to take the MUD AND computer down nicely if there is another power failure, to avoid these eventually costly hardware replacements.
On behalf of myself, I'd like to apologize for the long service outage. I hope the new CPU and extra RAM proves to be worth it, though!
Jan (Dios)
The hardware upgrade will be today, between 15:00 and 16:30 UTC.
This weekend (2001-12-15 - 2001-12-16), we are going to perform a hardware upgrade. Frogo has donated a motherboard (Asus A7V) and a CPU (AMD Athlon 1200). Dios has bought a new case (Chieftec Dragon DG-01W), a new graphics adapter (because the old one was ISA based, and we still need a VGA console) and a 512 MB DIMM (the old memory can't be reused). The "old" SCSI controller, harddisks, diskette drive and network adapter will still be used.
Well, that seemed to go well ...
We expect another power outage from 03:00 UTC tonight. Alleged downtime is supposed to be 20 seconds, expect two hours or so.
Viking lost electrical power at 04:05-ish tonight due to a planned outage that the administration didn't see any warning about. It turns out that an alert was sent to a local newsgroup 18 hours in advance.
The administration apologizes for the inconvenience, and thanks to Sigurd Mytting at USIT for turning the power back on and running the file system integrity check for us, getting the MUD back up at 12:08.
The DGD test MUD was back up at 12:12.
Nothing particularly exciting happened.
Nothing particularly exciting happened.
The kernel upgrade was successful.
We plan for some downtime Friday 2001-09-28 21:00 UTC - 21:30 UTC for a kernel upgrade, which should help on some performance problems. (It may not be noticeable for players.)
Our thoughts go to the citizens of the United States of America in the horrible terrorist acts against them. Please refrain from witty comments, jibes and jokes on this subject, and rather show your respect for the survivors. Dios, for the Administration.
The upgrade went well.
The MUD stays down between 07:55 and 08:15, approximately, since we are installing a new Linux kernel.
Nothing particularly exciting happened.
Viking rebooted today at 07:06 UTC, but then the driver decided the world was a mean place to boot in because of a typo in some domain code, and dumped core before really starting up.
Dios elegantly took advantage of the downtime to upgrade the operating system kernel from 2.4.2-jipatch to 2.4.5.
After the system reboot, the thing the driver went tits up over was fixed, and it came back up at 12:43.
Someone turned off our computer sometime between 2001-06-11 21:07 and 21:08 UTC! This is just plain incredible, and bloody annoying. Some files may have been corrupted, but we don't know at the moment. (It is possible that it was a temporary failure in the power supply unit, though.)
We have been informed that the location that hosts Viking will lose power at 1:30 local time, which is about 7:30pm EST if I converted that correctly. We do not know how long the interruption will last.
Nothing particularly exciting happened.
Around 2001-03-08 16:05 UTC, the computer went down for unknown reasons. It failed to come back up because of file system corruption, and it wasn't until around 11:30 UTC today that we managed to get it back up. The web server was disabled because it (and other software) had to be reinstalled to work correctly, and didn't come up until 13:18 UTC.
The system will probably be taken down for a complete system reconfiguration during the weekend, as things aren't working properly yet.
The MUD was down for a hardware upgrade between 15:00 and 16:30 UTC. The upgrade was supposed to be a new motherboard, new CPU (500 MHz PIII), and an additional 64 MB RAM. Unfortunately, the new motherboard was defective, and the current motherboard doesn't support the new CPU, so we're back to the original setup. But the new RAM works! :)
The MUD is back up as of 10:17 UTC.
Between 01:35:01 and 06:39:47 UTC, Viking lost electrical power in a planned blackout for the computer room we're lucky enough to be in. Unfortunately, nobody thought to give us any notice of the blackout, so it was a very hard shutdown for us. Currently, we're undergoing work to fix some minor issues and a kernel upgrade (since we're already down), and expect to be back up around 10:00 UTC.
The operating system crashed around 04:07 (UTC) today, for unknown reasons. The problem was reported around 11:15 (UTC), and the system booted successfully with a new kernel at 11:38 (UTC).
Nothing particularly exciting happened.