I'm using a Linode since march 2010 and it works fine but since a week i've got issues with the CPU an Disk IO.

It's a CentOS 5.5 (32Bit) Distribution which runs LAMP and Virtualmin.
I host just 3 simple websites, none of them have that much pageviews to cause this behaviour

I've checked and disabled some of the cronjobs in Webmin(there's none that runs every hour), optimized apache and mysql but nothing seems to work.





I'm not a very advanced Linux User and it's quite frustrating that I can't solve this on my own.
I hoping anyone has a clue to help me solve this issue.

Those spikes are rather wide. Whatever's causing it probably runs for at least 10 minutes before it quits. Wait until 5 minutes after the hour, whip out top or htop and see exactly which process is gobbling up the CPU at that time.

I'm monitoring with htop and the following command uses more than 100% CPU:

/usr/bin/python /usr/bin/yum info

I've had the same problem, but it seems to be only happening with webmin...

I think I got the culprit.

There is a process that is ran in webmin, but the way linode interprets it (5 minute average) is a bit misleading when trying to find what is causing these spikes.

Look for a cron job that is running /etc/webmin/virtual-server/collectinfo.pl. It runs this every 5 minutes and what the script does (from what I can understand), is poll every virtual server service and the server itself for updated information that is shown in webmin.

This can cause a huge, but short, load on a low-resource machine. I disabled it and will post the results in an hour.

Edit: After an hour and a few minutes; no more spikes, and average CPU load has decreased to 1-2%.