centos dedicated server is restarting every 30 minutes

Question

I'm not sure what's going on but its definitely not ddos attacks or hacking attempt but do you have any recommended solutions for our problem?

httpd, spamd, ftpd has been continuesly failing the past 2 days and we have a lot of websites in that server.

Is there anything in /var/log/messages at the time of the reboots? You haven't really given us a lot to go on here...

Kronick · Answer 1 · 2009-12-09 11:22:25Z

It looks like an hardware issue. The first thing to do is :

check logs : system.log and kernel.log.

Mostly it's an hardware failure as : - Broken fans, so the cpu become hotter (try $acpi -V) - Memory corruption (see above)

The things you can do are also :

you will have the total of memory under "used" column : used free shared buffers cached Mem: 3369312 3242568

now do :

$ dd if=/dev/urandom bs=768304 of=/tmp/memtest count=1050

$ md5sum /tmp/memtest; md5sum /tmp/memtest; md5sum /tmp/memtest

if the checksum do not match, that mean you have a memory failure

Check the network also. Do it work ? With these steps, you should have at list a piece of info

OK, couple of things. fsck will check the filesystem, not the disk. for that you want badblocks. Also, your dd test will only work if /tmp is a ramdisk/tmpfs. The better option would be to boot the machine into Memtest86 and leave it running for a couple of days.

ysimard · Answer 2 · 2009-12-09 14:04:11Z

Recently, I unplugged from a ups a server being rebooted by itself after 7 days. The ups is still connected to the apcupsd daemon.

I got this message yesterday.

Emergency! Batteries have failed on UPS ups1000_cabinet1. Change them NOW

The machine is still up and running, so the message never had the needed time to be written in the log file.

2 Answers