Topic: Unexpected shutdowns on Sme 7 Server

[gbentley]

In the last week I have a client who has reported that they have come in in the morning to find the SME 7 (+ all current updates) box off. The older (IBM Netfinity) 6 series is right next to it in the same power outlet (twin UK 13Amp socket) and is happliy still buzzing away !!!

Where do I go to start diagnosing this kind of thing ?

I am going to start trawling logs but would they help as if its a hardware error they might not have chance to be written ?

Any help / tips / advice really appreciated !!!

[frond]

gbentley

server manager - view log files - messages. Look for the time of shutdown & anything just before that.
Then look in other log files to see what else happened at that time. Look for cron jobs to see if something triggered the shutdown. Happy hunting.

[byte]

Have a look in the bios and see if there is any hardware monitoring of the fans and cpu temp also the volts of the PSU if any.

[mmccarn]

How reliable is the AC power?  I had an original Compaq Deskpro computer years ago that would happily survive a power hiccup that killed everything else in the office.  

Perhaps you're having power problems that the Netfinity can handle, but that the SME cannot, or perhaps the Netfinity is set to restart after a power outage, but the SME box is not.

Or, I've seen Windows servers that shut themselves down because of UPS software ("What!? I can't talk to the UPS?  I'll shutdown immediately!")

Or, I've seen lots of systems with poor CPU cooling that reboot spontaneously.

[arnie25]

Check squid logs.

squid[3268]: WARNING: Disk space over limit: 152372 KB > 102400 KB


This might be swap.state has been corrupted. Such corruption can
occur on unexpected system shutdowns (power failure, kernel panic etc).

You can try with the following

1. Shut down squid.

2. Remove the swap.state files from your cache directories.

3. Start Squid again. It will rebuild swap.state from the cache
files.