regency.estarcion.com dead pool
Regency is a 7 year old UMax Celeron-366 machine which provides NAT/firewall and shell service here at the Wabe.
Last night while trying to diagnose a problem with our phones (diagnosis: SBC sux) I moved regency and one of its internal fans, which occasionally groan unpleasantly, began moaning horribly.
Slapping the side of the machine repeatedly, which has sometimes shut it up in the past, merely resulted in a kernel panic. Guess I hit it too hard. (It was fully functional upon reboot, but still wailing.)
I finally opened up the machine, determined it was the CPU fan making the noise, and somewhat impulsively... removed the fan from the machine entirely. We left it off all night, but now it's back up and running.
If the server were being used for anything serious, this would be nothing short of reckless. But we have an Airport base station which could take over the NAT duty in a pinch, and another machine, off-site, for shell service. The machine was basically the second-cheapest one we could find in 1998. It's performed remarkably well, considering.
The machine is usually 98% CPU-idle, anyway. It might not even overheat.
I ordered a new heatsink/fan and a replacement CPU (a slot-compatible Pentium III 533) on the assumption that this isn't really a sustainable state of affairs.
Please post a comment with:
A) Estimated date/time of death of the current CPU w/o fan
B) A description of the expected failure mode (e.g. "kernel panic, doesn't reboot", "bursts into flames", "starts global thermonuclear war due to data corruption")
C) Estimated date/time and circumstances for regency's final retirement
Last night while trying to diagnose a problem with our phones (diagnosis: SBC sux) I moved regency and one of its internal fans, which occasionally groan unpleasantly, began moaning horribly.
Slapping the side of the machine repeatedly, which has sometimes shut it up in the past, merely resulted in a kernel panic. Guess I hit it too hard. (It was fully functional upon reboot, but still wailing.)
I finally opened up the machine, determined it was the CPU fan making the noise, and somewhat impulsively... removed the fan from the machine entirely. We left it off all night, but now it's back up and running.
If the server were being used for anything serious, this would be nothing short of reckless. But we have an Airport base station which could take over the NAT duty in a pinch, and another machine, off-site, for shell service. The machine was basically the second-cheapest one we could find in 1998. It's performed remarkably well, considering.
The machine is usually 98% CPU-idle, anyway. It might not even overheat.
I ordered a new heatsink/fan and a replacement CPU (a slot-compatible Pentium III 533) on the assumption that this isn't really a sustainable state of affairs.
Please post a comment with:
A) Estimated date/time of death of the current CPU w/o fan
B) A description of the expected failure mode (e.g. "kernel panic, doesn't reboot", "bursts into flames", "starts global thermonuclear war due to data corruption")
C) Estimated date/time and circumstances for regency's final retirement