Big trouble with cron - high systemload and not working jobs

Lord of Gore lordofgore at logsoftgrup.ro
Thu Jan 18 09:50:58 UTC 2007


Götz Reinicke wrote:
> Hi,
>
> we do have some big problems with one of our RHEL3 (running the latest 
> updates) Servers.
>
> Since last week the systemload increases within two or three days up 
> to 60! and no cronjobs are done anymore, they hang.
>
> The jobs are the usual systemjobs (make-whatis, updatedb for slocate) 
> and our backupscript which worked fine for the last two years.
>
> Furthormore some servises hang, e.g. NetAtalk (appletalk 
> fileservises), ssh login and local login. For example Samba still 
> works fine.
>
> If I reboot the server, all jobs will be killed and for one or two 
> days the load is not above 1 and all servises are fine, crond ass well.
>
> Restarting or killing the crond and the cronjobs fail too.
>
> Any troubleshooting hints? googling didn't helped so far and there are 
> no usable logmessages :-(
>
>
> Thanks a lot and best regards
>
> Götz Reinicke
What do you mean by "there are no usable log messages" ?
It could be anything, from a software to a hardware failure. Maybe even 
you got hacked. *Only* the logs can get you out of this mess.
Check if by plugging out the LAN cable things get back to normal. 
Restart (hard) the server and pay very close attention to log messages. 
I don't know about the rest but I always add *.* /dev/tty12 to my 
syslog.conf.
As a starting point I would see how named is doing and check my ECC 
memory modules. I'd bet 1 box of Denmark's best beer that it's either 
one of those.




More information about the redhat-list mailing list