Exim_tidydb

Allan

Verified User
Joined
Sep 3, 2006
Messages
16
Hi, i have an issue where tidydb is running at 100% cpu, and if left will just continue to use 100% cpu forever?

When its not running my load averages on the server are ususaly around 0.2 so its not under any heavy usage...

Is there any way to find out why the tidydb is using 100% cpu ?
 
Tidydb doesn't seem to run on our DirectAdmin based servers. If you're not using DirectAdmin then you're asking in the wrong place.

If you are using DirectAdmin please post the exact code you use to see that it's running.

Jeff
 
I am using Debian 3.1
Directadmin 1.31.5 and exim is version 4.68.
Im just using top and can see it running under the process list.

It will be using between 99% and 100%, so sits at the top of the process list...
The load averages will be 1 1 1 roughly so everything appears to work fine, just the fact that the cpu is running at max
 
Last edited:
tidydb doesnt run by default I remember adding it myself to my crontab. Other than checking the logs and maybe lsof I dont know how you could see what its doing.
 
It could be possible that it has a cron, will check this out.
at the moment i am just monitoring when it starts and then killing the process
 
ok i have found a exim file in /etc/cron.daily which seems to be running exim_tidydb, so im going to move this script out and hope this stops it from booting up :D
 
exim_tidydb: program to maintain exim mailer hint databases

apparently !
 
Ok well, removing the cron job didnt work, so i moved the actual hints DB out into a backup and let exim create a new one.

I then manualy ran exim_tidydb

/usr/sbin/exim_tidydb -t 1d /var/spool/exim retry
and
/usr/sbin/exim_tidydb -t 1d /var/spool/exim wait-remote_smtp

running these before resulted in 100% cpu usage and the high loads, after removing the old DB's it looks like its working fine

I get the message Tidying complete now instead of just locking the machine up :D


so now i have put these in as cron jobs into crontab so it will run nightly and hopfully clean all the undelivered mail out... will keep an eye on it :D
 
How would it clean all the undelivered mail out? If user Chrysalis is correct, then it should be cleaning out the list of mx IP#s to use; not the emails themselves.

Jeff
 
possibly just cleaning out old logs then, to be fair i dont have any undelivered mail in my list.

but hopfully this will solve the 100% cpu issue :D
 
Not the logs ... only the hints which I believe are defined as the IP#s to use, to avoid having to look up the IP# again for mail already in the queue.

I could be wrong; anybody want to do the research :) ?

Jeff
 
Ok well, the cpu load is running 0.1 load now and normaly by now i would be seeing exim_tidydb rincing the cpu, so it looks like its solved :D

Ta for the hints on the cronjobs, looks like the problem was just a courpt hints DB..

Got my 2 e-mails this morning to confirm that all ran as planned, so i guess thats problem solved :D

Tidying Exim hints database /var/spool/exim/db/wait-remote_smtp
deleted zadonsk.net (empty)
updated mail.zsdfg.com
Tidying complete

Tidying Exim hints database /var/spool/exim/db/retry
Tidying complete

Next thing on the list is to setup the second DA server :D

ta for the help guys
 
Last edited:
Back
Top