action=tally just hangs

C

cgiscripts4u

New member
Joined
Jan 20, 2006
Messages
4
Hi

I currently have 3 servers all running DA and I am extremely happy with it, however on the latest server, which was just set up a few days ago the tally cron goes on for ever, to a point when I have to kill the process because it is pushing my server load through the roof.

My site is a high traffic site, so much so I have had to mirror it and use round robin dns so to use the two web servers, the other server serving the database. Each web server is getting upto 100,000 visitors a day.

Yesturday I was watching the latest server at around 10am and noticed dataskq was running and had been since 0:10, webalizer stats had updated and my log files had been rotated so I killed the process.

Today, I checked the server and noticed the same, dataskq was still running since 0:10 this morning, webalizer stats had updated and logs had been rotated. The bandwidth figure in DA appears not to be updated though as it shows something like

Bandwidth (gb) 0.0018 unlimited

but I am not worried about this as I MRTG showing my bandwidth.

Is there a way to disable the disk space/bandwidth tallys, so that only webalizer stats are updated and log files rotated and then nothing else or do I have to totally disable the tally cron and update webalizer and logs directly.

BTW the html pages are on a mounted nfs directory shared between the three servers, I don't know what effect this would have.

Thanks

Greg
 
Last edited:
Hello,

What is the command using the most CPU when you run:
Code: 
top
when it's in this state? Is it the actual dataskq, or is it something called by the dataskq.. like webalizer or tar. Also check the /var/log/directadmin/system.log to see what the last entry it was on.. also the errortaskq.log for any other clues.

The nfs may have an effect on it.. DA does search through user homes for domain usage.. which could cause a hangup if the nfs isn't responding.. or DA isn't able to get the data it needs for whatever reason.

John
 
I am having a similar behaviour, and the top command also shows that the load went to almost 100%
 
I found it to be the size of the logs for one of the domains, I now run the cron far more frequently so the logs do not grow so large and do not have any problems
 
I have also noticed that the dataskq process is taking almost 100% CPU load
everytime is executed.

Which batch command is the one you are running now more frequently?
 
You must log in or register to reply here.
Back
Top