|
by Senor Barborito 06/15/2003, 11:59pm PDT |
|
 |
|
 |
|
Saw OpenBSD crash for the first time in three years - for about 24 hours the HD on the new server was making loud 'click' every 5 minutes or so. According to the kernel panic syslog, around 7 AM this morning the fastfilesystem I/O subsystem went APESHIT and the system promptly panicked and hung hard.
Unfortunately the damage was pretty nasty and I'd rather test the HD via new install than attempt to dig through mountains of crap and figure out what got destoyed in the crash (this is the real difference between a hardcore hobbyist and a sysadmin who gets paid, truth be told).
There are two possible explanations: first of all the case is a PII 400 case now holding a 1.2GHz Tbird - this case has ZERO external fans/blowholes. Just some panels in the front/back punched out. It's possible either HD or proc is overheating (in case of proc, consequently causing strange HD access patterns as it resumes I/O processes). The second is, obviously, faulty HD.
I can afford a new . . . would be 8th case lying around here and some fans. I don't think I can afford a HD though (budget is $60-$75 here). Anyways, currently formatting and spending what's left of the day doing as much of the rebuild as I can - going to try and finish it all the way up by tomorrow before letting it run a few days with all processes. At that point, testing with actual code and data begins.
Advice is really fucking welcome here - the clicks are ominous, but I really don't think I'm on the market for a HD right now, and this system could use a new case DESPERATELY.
In any case, if you were wondering where the new system is, the answer is 'cooking itself to death.' To quote God's Final Message to His Creation: "We apologize for the inconvenience."
--SB
Please stay on the line, an operator will be with you shortly . . . |
|
 |
|
 |
|
|
|