Apache logging stops intermittently on nearly all cPanel servers.

[jols]

After receiving several complaints that AWStats is recording NO traffic so far this month, we discovered that several of our cPanel servers had no recent updated log entries in /usr/local/apache/logs/access_log

I managed to start up the logging process again with the following set of commands at shell:

/etc/rc.d/init.d/httpd stop
killall httpd
/etc/rc.d/init.d/httpd startssl
/scripts/restartsrv_syslogd

But on one of our servers, logging has already stopped again.

Does anyone have a clue as to how we could resolve this one?

Thanks.

 

[jols]

By the way, the access_log file is not too large, and there is plenty of room in /usr. Also none of the servers are significantly loaded, etc. We can determine no cause for this issue.

 

[jols]

Just to offer a slight modification to the above: It appears that the following is not necessary to get Apache/traffic logging going again:

/scripts/restartsrv_syslogd

Just restarting Apache seems to work, but we can't restart Apache every 10 minutes just to make sure logging stays up. Anyone have a solution for this one?

 

[jols]

ANYONE?

Here's a survey of our cPanel servers, and the results.

How this test was conducted, by running these commands at shell:

tail /usr/local/apache/logs/access_log
date

Then comparing the time stamp of most recent entry with the time in the "date" command. We should at least see the following kind of entry being placed there every few seconds:
127.0.0.1 - - [05/Mar/2010:16:31:44 -0600] "OPTIONS * HTTP/1.0" 200 -

Here are the results:

1 - two hours delay
2 - normal
3 - six hours delay
4 - normal
5 - normal
6 - six hours delay
7 - two hours delay
8 - two hours delay
9 - six hours delay