Litespeed restarts multiple times a day

[skooboy]

I see this very often in the log almost every two hours.

2015-01-21 00:11:28.860 NOTICE Server Restart Request via Signal...
2015-01-21 00:11:28.860 NOTICE [PID: 24510]: start cgid: 28675, /usr/local/lsws/bin/lscgid
2015-01-21 00:11:28.860 NOTICE [Child: 24512] Start shutting down gracefully ...
2015-01-21 00:11:28.860 NOTICE [Child: 24513] Start shutting down gracefully ...

I don't see anything in the logs of what would cause it. Please advise. Thanks.

 

[NiteWave]

check WHM/cPanel settings, if rotating logs every 2 hours

 

[skooboy]

Where is that? I can only find rotate bandwidth stats every 2 hours.

There are more hat I did not paste from the error_log but it's the usual entries for a lsws restart.

 

[NiteWave]

please refer http://forums.cpanel.net/f442/why-apache-restarting-every-two-hours-254022.html#post1056361
"Why is apache restarting every two hours?"

 

[skooboy]

It's not exactly every 2 hours. Something is not right, where else can I look for clues?

 

[NiteWave]

as that cPanel thread suggested,
" enabling that logging option in WHM > Apache Configuration > Piped Log Configuration area to see if the graceful restarts every 2 hours cease to occur."

 

[skooboy]

After disabling smart keep-alive, it doesn't restart anymore.

 

[NiteWave]

surprising ... how did you come out the idea to change that setting which looks unrelated at all ?

 

[skooboy]

That was the only setting that was different on the new server.

 

[skooboy]

Spoke too soon, it just restarted 20 minutes ago. It was up for almost 24 hours.

 

[skooboy]

I found a pattern of it restarting almost every 24 hours. It looks like whenever /scripts/runweblogs is ran to process the domain logs, litespeed restarts. This never occurred on old server with piped logging disabled.

 

[NiteWave]

how about manually run /scripts/runweblogs , will lsws restart ?

 

[skooboy]

Yes, it does and same when it's not run manually. Can lsws be changed in the future to avoid having to be restarted? I do not want to enable piped logging.

 

[justint]

I'm having exactly the same problem. Any way to avoid restarting?

 

[skooboy]

I changed the Log Processing to every 96 hours and it still restarts at least once a day. I think it's not the log processing that causes it. How do I debug it?

 

[NiteWave]

I'd suggest to try piped log for 48 hours anyway to see if any difference.
this can narrow down the scope greatly.

 

[skooboy]

Still the same. Any other way to debug?

 

[NiteWave]

lsws restart log is located at
/usr/local/lsws/logs/lsrestart.log
logs of WHM/cPanel , just searched out a good link:
http://www.inmotionhosting.com/supp...s-for-access-apache-email-error-ftp-mysql-whm

can compare with
Apache restarts done through cPanel and WHM :::::: /usr/local/cpanel/logs/safeapcherestart_log

help to know the restart is from WHM/cPanel or lsws itself.

 

[skooboy]

Both restart logs for lsws and cPanel and WHM point to the same time.

 

[NiteWave]

yes. I mean comparing the time stamp of latest log entries in /usr/local/lsws/logs/lsrestart.log and /usr/local/cpanel/logs/safeapcherestart_log (or other WHM/cPanel log files), if they match, then the restart comes from WHM/cPanel, like adding an account/terminate a account, change apache's configuration in WHM which will require a restart.
if no match, then the restart may come from lsws itself.

how often does lsws restart now ?