Server crashes/restarts 00:07 every night

patrik

Verified User
Joined
Sep 6, 2006
Messages
128
I need some help trouble-shooting the server. It crashes every night at 00:07 and then it restarts. Also some minutes later MySQLd fails and shuts down (without restarting).
This is the output from /usr/bin/last at the time of crash:
patrikjans ttyp0 213.134.110.241 Wed Sep 6 00:07 - 00:22 (00:14)
reboot ~ Wed Sep 6 00:07
admin@tigerlily ftp 213.101.52.81 Tue Sep 5 23:55 - crash (00:12)
admin@tigerlily ftp 213.101.52.81 Tue Sep 5 23:51 - 23:54 (00:03)
patrikjans ttyp1 213.134.110.241 Tue Sep 5 23:37 - crash (00:29)

And this is the output from /var/log/messages at the time:
Sep 5 23:54:50 obelix last message repeated 5 times
Sep 5 23:57:00 obelix last message repeated 6 times
Sep 5 23:59:54 obelix last message repeated 18 times
Sep 6 00:00:01 obelix proftpd[97363]: obelix.kontrollpanelen.se - received SIGHUP -- master server rehashing configuration
file
Sep 6 00:00:08 obelix httpd: gethostby*.getanswer: asked for "bilder.phoenix.a.se IN AAAA", got type "A"
Sep 6 00:00:39 obelix last message repeated 4 times
Sep 6 00:02:49 obelix last message repeated 18 times
Sep 6 00:04:00 obelix last message repeated 2 times
Sep 6 00:04:25 obelix httpd: gethostby*.getanswer: asked for "mail.stockholms.se IN AAAA", got type "A"
Sep 6 00:04:26 obelix httpd: gethostby*.getanswer: asked for "bilder.phoenix.a.se IN AAAA", got type "A"
Sep 6 00:07:19 obelix syslogd: kernel boot file is /boot/kernel/kernel
Sep 6 00:07:19 obelix kernel: Copyright (c) 1992-2006 The FreeBSD Project.
Sep 6 00:07:19 obelix kernel: Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
I need some help figuring out why the system crashes. Which logs are interesting, commands to run to get some info (like last)?, etc.
It's a FreeBSD 6.1 machine.

Thanks,
 
I figured out the system crashes when I run /usr/sbin/quotaoff -a. Any clue why?
 
I couldn't say why shutting off the quotas crashes the box.
I can say that there is a nightly quotacheck on freebsd in /etc/crontab (quotaoff, quotacheck, quoton)... so comment out that line and restart crond.

John
 
(This post does also answer the mail I got from DA).
You said in the mail that if quota is supported by kernel there is no need to run quotacheck?
I have googled a bit about quotacheck and it seems like people doesn't always run quotaoff and quotaon together with the quotacheck command. Maybe I only have to run quotacheck? (that doesn't crash the system).
But again, if quotacheck is not needed for kernels where quota is enabled then maybe I could skip the whole cronjob for quotas (i have enabled quota in the kernel by adding options QUOTA in the kernel config.
 
I have this error on two of my CentOS systems, one got eth0 problem with it right after the IN AAAA", got type "A" errors : eth0: PCI Bus error 8290
it's not happened on 00:07.
no idea what's wrong...
 
This seems like the cron job that running just after midnight too me.

Why dont you show us the output of crontab -l and /etc/crontab
 
Problem is already solved. Well I didn't do much more research because running quotaoff/quotaon isn't needed.
 
Back
Top