lately every couple of weeks, the server stops running.
Did you check what config actions, upgrades, whatever else was done around the last date the box was stable?
Do you have any long range SAR data to check for irregularities?
Where/are there any anomalies in the system/daemon logs except the oops itself?
The culprit seems to be a series of nasty kernel errors. They conclude with the following statement: Kernel Panic - Not Syncing: Aiee, killing interupt handler.
The "Aiee" line in itself is not important, the surrounding lines are. Have a look at /usr/src/linux/Documentation/oops-tracing.txt. Basically sez feed oops log to ksymoops. Posting those lines could help.
Is there a cause for something like this or has the server been hacked?
Can't evaluate: not enough info.
|