For some reason my server has been having trouble processing all of the stats on my server...

The server is having trouble keeping up with your statistics processing schedule. You should increase the time between statistic generation, or upgrade the server. If you have recently decreased the time between statistic generation, you may wish to wait that amount of time to see if the server will catch up before changing back.


So far the logs are '118575' minutes behind, and I have no idea how to fix this or how to even reset the stat generator.
Any ideas?

Sometimes people set the stats to be generated too frequently.

Let's take a hypothetical scenario. If it took 5 minutes to process a log, but you set the stats process to be performed once per minute then the stats process would never complete (as it starts over every minute). That would cause statistics to look like they're not being generated.

Of course, in a production environment these times are much larger. As your traffic increases and/or number of clients increase, these processes can take substantially longer.

You can edit such settings in WHM if under the "Server Configuration" block on the left side of the screen you click "Statistics Software Configuration."

Yeah... I know, the stats generate every 48 hours.
But that doesn't really help me 'unfreeze' the stat generation, so it can try to generate the stats again.

Yeah... I know, the stats generate every 48 hours.
But that doesn't really help me 'unfreeze' the stat generation, so it can try to generate the stats again.

You haven't set any blackout times where stats generation cannot run... correct?

By that do you mean "Check the hours that you would like log analysis NOT to be performed."?

So if I set the times here, for 8am through 10pm, it'll stop generating the stats now?

By that do you mean "Check the hours that you would like log analysis NOT to be performed."?

So if I set the times here, for 8am through 10pm, it'll stop generating the stats now?


It is better that you don't restrict the times for the stats to generate.

Not trying to limit the time it runs... I'm trying to turn it off so it will try and generate the stats again, but I can not for the life of me, find out how to turn it off.

I have seen it many times recently...

If you have commandline access make sure cpanel logs are running.

ps aux |grep cpanellogd
(probably not)

Restart cPanel if its not running.
service cpanel restart

If it dies again in a day or so, which it probably will, see if chkservd is running.
ps aux |grep chkservd
(probably not)

Generally its due to outdated perl when you see this because for whatever reason chkservd dies or cpanellogd does.

perl -V

You should be running 5.8.7 most likely, if your cPanel is running a latest version.

Summary of my perl5 (revision 5 version 8 subversion 7)

If its not the fix, not sure but that is what has worked for lately.

ps aux | grep cpanellogd

root 26746 0.0 0.0 4772 544 pts/1 R 03:14 0:00 grep cpanellogd



ps aux |grep chkservd

root 15326 0.0 0.2 12916 2480 ? S Sep16 3:13 chkservd
root 26768 0.0 0.0 4760 676 pts/1 S 03:14 0:00 grep chkservd



perl -V

Summary of my perl5 (revision 5 version 8 subversion 7) configuration:



cPanel was running, but I tried to restart anyways... Couldn't find the command services or something similar

Well that says that chkservd is running, and you are good with Perl. Great things!

It does say that cpanel's logger daemon was down, and that's the reason your stats are so far behind.

You need to restart cpanel (easiest way to do it) to get the logger running again, and over the next 24-36 hours the system should get caught up on the stats.

it will look something like this when its running right:

root@server[~]# ps aux |grep cpanellogd
root 23413 0.0 0.6 13508 10332 ? SN Nov25 9:34 cpanellogd - sleeping for logs
root 12066 0.0 0.0 3684 572 pts/2 S 04:27 0:00 grep cpanellogd

You will have your grep and the actual process running in your output.

If:

root@server[~]# service cpanel restart

did not work, what was the error returned? What is the OS?

service cpanel restart
bash: service: command not found

OS: RH...3? 98% sure

(And thank you for the help :))

Try this then:

/etc/init.d/cpanel restart

/etc/init.d/cpanel restart

Stopping cPanel services: [ OK ]
Stopping pop3 services: [ OK ]
Stopping cPanel log services: [FAILED]
Stopping cPanel Chat services: [FAILED]
Stopping Melange Chat services: [FAILED]
Stopping InterChange services: [FAILED]
Stopping cPanel ssl services: [ OK ]

Stopping mailman services: Shutting down Mailman's master qrunner

Starting eximstats: [ OK ]
Starting cPanel services: [ OK ]
Starting cPanel Log services: /usr/bin/perl: relocation error: /usr/local/cpanel/3rdparty/lib/perl/auto/RRDs/RRDs.so: undefined symbol: Perl_Gthr_key_ptr
/usr/bin/perl: relocation error: /usr/local/cpanel/3rdparty/lib/perl/auto/RRDs/RRDs.so: undefined symbol: Perl_Gthr_key_ptr
[FAILED]
Starting pop3 services: [ OK ]
Starting cPanel Chat services:
Starting Melange Chat services:
Starting cPanel ssl services: [FAILED]
Starting mailman services: Starting Mailman's master qrunner.



According to WHM, the logs are still behind by 3 months or so...

Try the following commands and see if you encounter any errors. If you do, post them so I can see whats going on.


/usr/local/cpanel/bin/rrdtoolinstall --force
/scripts/runweblogs <Username>
/scripts/fixwebalizer <Username>

Check the logs below to see if you have any errors


/usr/local/cpanel/logs/error_log
/usr/local/cpanel/logs/stats_log

Double check to make sure you have no domlogs that over 2GB in size.

/etc/init.d/cpanel restart


According to WHM, the logs are still behind by 3 months or so...

They are going to take a day or so to get up to date, but the rrdtool fix above should get rid of the problem.

After you do /usr/local/cpanel/bin/rrdtoolinstall --force and restart cpanel again, if you get no errors next to your cpanel log service, then you are probably good to let them do their thing.

If you have users that need immediate updates you can do these sort of things on a per user bases:

/scripts/runweblogs <Username>
Runs the logs for a user

/scripts/fixwebalizer <Username>
fixes a corrupt user's stats if you come across one that wont run, but I think this one deletes all log history so use it with caution.

Can't believe it took me this long to get around to doing this....
Anyways, I think rrdtoolinstall fixed it, however I need to find a way to update all user's logs rather than user by user with /scripts/runweblogs

You can use the command below to manually run the stats.


/scripts/runstatsonce