up vote 1 down vote favorite
share [g+] share [fb]

I have one server running Windows Server 2003 and another one running CentOS 5.5. I created a proxy to allow users to connect to the CentOS server through the Windows server. Everything worked fine until there were like 30 users online. The CentOS server automatically shut down.

Here is some of the log from file messages:

Dec 2 01:59:59 localhost avahi-daemon[4596]: Server startup complete. Host name is localhost.local. Local service cookie is 3442367051.
Dec 2 01:59:59 localhost pcscd: winscard.c:304:SCardConnect() Reader E-Gate 0 0 Not Found
Dec 2 01:59:59 localhost last message repeated 3 times
Dec 2 02:00:00 localhost kernel: mtrr: no more MTRRs available
Dec 2 02:00:01 localhost last message repeated 7 times
Dec 2 07:08:28 localhost shutdown[6703]: shutting down for system halt
Dec 2 07:08:28 localhost pcscd: winscard.c:304:SCardConnect() Reader E-Gate 0 0 Not Found
Dec 2 07:08:29 localhost smartd[4658]: smartd received signal 15: Terminated
Dec 2 07:08:29 localhost smartd[4658]: smartd is exiting (exit status 0)
Dec 2 07:08:29 localhost avahi-daemon[4596]: Got SIGTERM, quitting.
Dec 2 07:08:29 localhost avahi-daemon[4596]: Leaving mDNS multicast group on interface eth0.IPv6 with address fe80::be30:5bff:fee5:c57c.
Dec 2 07:08:29 localhost avahi-daemon[4596]: Leaving mDNS multicast group on interface eth0.IPv4 with address xxx.xx.xxx.xx.
Dec 2 07:11:56 localhost syslogd 1.4.1: restart.

Both servers have more than enough resource for 30 users. What could be causing this? Thank you.

link|improve this question
Looks to me like a clean shutdown, called by the shutdown or halt command, rather than a crash. Did somebody press the power button, perhaps? – Andy Smith Dec 2 '11 at 0:51
I don't think so because it works fine with less than 30 users online. Once the number of users reaches 30+ it will shut down in about 10 mins. – Mars Ares Dec 2 '11 at 1:18
Might be worth having a look at sar to see what the load was like before the shutdown. I've honestly not seen a Linux box shut itself down cleanly for no reason... unless it's a temperature/sensors issue. – Andy Smith Dec 2 '11 at 1:32
I think it is because of my proxy. I just checked it again and noticed that half of the users are disconnected 5 mins before all users are kicked out. Checked twice and it acted the same. Thank you anyways. – Mars Ares Dec 2 '11 at 2:10
What are you using as a proxy, out of interest? – Andy Smith Dec 2 '11 at 2:11
show 1 more comment
feedback

1 Answer

up vote 0 down vote

Twice now I've had a similar experience. I have CentOS as a workstation, with VMWare workstation running. My experience is this:

At night I "lock" my console (as a regular user, not root). In the morning, I enter my password and "unlock" the workstation.

Things run fine for about a minute, then Zingo... CentOS is going into shutdown mode. This clobbers the VMs I have running with VMWare workstation (very bad juju).

I checked my crontab, and I have no cron jobs running.

How would I track down what is triggering the spontaneous shutdown?

link|improve this answer
1  
haha, you could replace /usr/bin/reboot and /sbin/shutdown with a script like "#/!bin/bash; logger "naughty person on id whoami " – Tom H Feb 16 at 17:14
feedback

Your Answer

 
or
required, but never shown
discard

Not the answer you're looking for? Browse other questions tagged or ask your own question.