Linux box unreachable after a few hours uptime

[valicon]

Hi all,

I have a strange issue. I am running RH 8 and the box is fine until a few hours or sometimes a day passes and then the box cannot be reached via Apache or SSH. But the console is up and running, everything is fine. When I doa reboot everything returns to normal. Anyone ever see this before?

 

[thedaver]

Are you running any unusual CRON jobs??
Are you running any home-coded daemons that might memory leak?
Is your xinetd/inetd/tcpserver crashing?


"Surfinbox Shares" - A fundraising program that builds revenue from dialup Internet users.

http://www.surfinbox.com/shares/

 

[fluid11]

A few things to check...

- run 'ifconfig' and make sure that your NICs are still up
- run 'ps -ef' and look to see if Apache and sshd are still running
- check /var/log/messages and search for anything unusual, especially at the times where the service went down

If your not sure exactly what time the service drops, it might be a good idea to run some kind of network monitoring agent on the network. I would recommend Nagios,

http://www.nagios.org.

You can set up Nagios to ping your server every 3 minutes or so and check the HTTP and ssh protocols to ensure that they are up. When they go down, Nagios will alert you and tell you what time it happened. This information could be very useful in troubleshooting this issue.

ChrisP
RHCE, LPIC-1, CCNA, CNE, MCSE, +10 others

 

[aixmurderer]

I had a similar case running Mandrake 8.1, at that time I suspected the NIC. Unplugging the UTP cable and plugging it back in did the trick to restore connectivity.

Since installing RH 8 on the same hardware the problem went away.

IBM Certified Confused - MQSeries
IBM Certified Flabbergasted - AIX 5 pSeries System Administration
MS Certified Windblows Rebooter

 

[jwenting]

Had something similar on an old box.
After a period of time the NIC burped on receiving any network traffic at all and threw itself off the network.

New NIC didn't cure it, same distro with same NIC on another machine worked fine.

That left only the mobo as the cause (as far as I could determine). As it was an old machine anyway I ditched it.

 

[valicon]

Thanks everyone for the advice. I will try all of it tonight and see what happens Thanks again I really appreciate it and I will keep you guys posted!