Sudden cpu spikes


I have noticed in the last couple of days that our server has multiple intermittent CPU spikes as you can see in the graph attached. I ran a dstat for a while and noticed that awastats and collectinfo are using a lot of CPU.

Any ideas to help troubleshoot this?

Thanks, Dan.



Howdy -- you could try taking a look at the file "/etc/cron.hourly/awstats" -- you wouldn't actually need that to run, you may just want to comment that out if it's not already.

Also, you may want to look at root's crontab, and see if there are any processes being run near when you're seeing those spikes.

If so, you may want to change them so that they run later in the day, and perhaps outside of business hours.

Also, another thing you could do is to go into System Settings -> Virtualmin Config -> Status Collection, and in there, you may wish to increase the value for "Interval between status collection job runs". I believe it defaults to 5, you could try 30 or 60 instead, and see if that helps with the issue you're seeing.

I worked through it a bit more, and found now that it is not the awstats or the, it is rpm causing the spikes. Would you know if there is any service in Virtualmin that would run the rpm command?

Thanks, Dan.

Collectinfo may be running rpm, in order to determine if new packages are available for download.

Try setting that to run less frequently, as described above, and let us know if that helps.