What could be spiking load at midnight?

[joeker]

I have a dedicated server running DA and everyday at midnight the server goes into a 5-min tailspin spiking the load to about 90 and becoming unresponsive.

I'm assuming it's something being run by a cron....mayeb the logrotate or webalizer?

Anyways, I've disabled webalizer by adding "webalizer=0" to the directadmin.conf file because I dont need it running.

I have a rather large forum on the server, it's logfiles hit a few hundred MB each day.

Any other suggestions?

Cron:

* * * * * root /usr/local/directadmin/dataskq
2 0-23/6 * * * root echo 'action=vacation&value=all' >> /usr/local/directadmin/data/task.queue;
#5 5 * * 0 root /sbin/quotaoff -a; /sbin/quotacheck -augm; /sbin/quotaon -a;
10 10 * * * root echo 'action=tally&value=all' >> /usr/local/directadmin/data/task.queue
20 11 1 * * root echo 'action=reset&value=all' >> /usr/local/directadmin/data/task.queue
0 4 * * * root echo 'action=check&value=license' >> /usr/local/directadmin/data/task.queue

 

[BestBoard]

maybe some backups? or some cron jobs
look at the log files and you see yourself what happened in midnight...

 

[joeker]

maybe some backups? or some cron jobs
look at the log files and you see yourself what happened in midnight...

Backups run at about 3am, and they do slow the server down, this is something else. Logfiles are inconclusive, they show mysql errors but thats because the server is overloaded. Top while it's happening doesn't really show anything except kswapd maxing out, but thats just the swap file.

 

[chatwizrd]

Is there anything in /var/log/cron at midnight?

 

[nobaloney]

Have you tried running top every five minutes beginning at 11:45 PM, mailed to you?

Jeff

 

[joeker]

It's gotta be the task.queue:

Sep 6 00:00:01 server crond[14337]: (root) CMD (/usr/local/directadmin/dataskq)
Sep 6 00:00:01 server crond[14339]: (root) CMD (/usr/local/nobody_check/nobody_check >/dev/null 2>&1)
Sep 6 00:01:01 server crond[15769]: (root) CMD (/usr/local/directadmin/dataskq)
Sep 6 00:01:01 server crond[15770]: (root) CMD (run-parts /etc/cron.hourly)
Sep 6 00:02:01 server crond[15798]: (root) CMD (echo 'action=vacation&value=all' >> /usr/local/directadmin/data/task.queue
Sep 6 00:02:01 server crond[15799]: (root) CMD (/usr/local/directadmin/dataskq)
Sep 6 00:03:01 server crond[15949]: (root) CMD (/usr/local/directadmin/dataskq)

I mean that's log rotating? webalizer, what all runs with that once a day?

It peaked at 225 load today

 

[joeker]

I am getting this in the cron.daily email report:

/etc/cron.daily/prelink:

/etc/cron.daily/prelink: line 47: 30535 Aborted /usr/sbin/prelink -av $PRELINK_OPTS >>/var/log/prelink.log 2>&1

any ideas?

 

[joeker]

Hey everyone, I figured it out. Hopefully this info can help someone else.

The daily spike at midnight had nothing to do with the server itself or DirectAdmin. It was Vbulletin daily tasks, about 4 of them all scheduled at midnight, and since we have a large forum with almost 2 million posts and over 11,000 members, it makes sense that it was VB spiking the server load.

Regards