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

I'm currently in the process of building up a web application. However, I occasionally run into an issue where "usr/sbin/exim_tidydb /var/spool/exim wait-remote_smtp" suddenly jumps up to 99% cpu and prevents other processes of my application from running.

From what I understand Exim is a webmail service, however there are a very limited amount of emails being sent out from this application (certainly not enough to cause this).

What's going on and how can I prevent this from happening?

BTW: I'm running Centos v.5 on dedicated hardware with standard apache at the moment.

link|improve this question
feedback

1 Answer

up vote 0 down vote accepted

Exim is a MTA, much like sendmail. The tidydb service is probably scheduled via cron.

My suggestion would be to modify the cron service (check /etc/cron[tab,.*] to run exim_tidydb when your server is typically not under high load, say 3am.

link|improve this answer
Sorry, still somewhat confused as to what the actual root of the issue is. – Thenayr Nov 15 '10 at 21:54
Your system is configured to run a very high cpu service during a time when your server is already under high utilization. In order to fix this, you'll need to reschedule the service to run at a different time, or move the mail server to its own machine. – uidzer0 Nov 15 '10 at 21:58
So the "usr/sbin/exim_tidydb" requires 99% cpu to go about it's business? Unfortunately their really isn't a downtime for my server as it has a very specific set of operations every hour. – Thenayr Nov 15 '10 at 22:01
yes, its a common problem (google: exim_tidydb cpu). You could remove it from cron, but that may have adverse effects on your mail db. – uidzer0 Nov 15 '10 at 22:07
Gotcha. Thanks ;] – Thenayr Nov 15 '10 at 22:15
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.