[11.25 Current] apache restarting every hour

[sharmaine001]

We have made a change under WHM -> Statistics Software

At the bottom, clicked on "Statistics Schedule" and set the stats not to run during busy hours. You can also elect to make it not run at all if you want.

At least this resolves it temporarily until cPanel gets back.

This doesnt work for me. It has been disabled since I started with the cpanel server but I still see these restart from my logs

[racakg]

This doesnt work for me. It has been disabled since I started with the cpanel server but I still see these restart from my logs

Did you try to disable apache and cpanel log rotation?

[hicom]

This doesnt work for me. It has been disabled since I started with the cpanel server but I still see these restart from my logs

Once the changes made, we restarted Apache through WHM and that took care of it. it is working now according to schedule

[cpanelkenneth]

We are currently examining a change in log handling that will remove the need for restarting Apache when processing bandwidth and domain log data.

[tweakNB]

Just my two cents

Personally Apache restarting is not too much of an issue, although of course if cPanel provides an update which removes the need it would be nice.

But in my case the major issue is with the Apache bug which results in a process not ending and becoming stuck.

Although no-one wants a restart, if it did end up being the only way to process the logs correctly it would be livable with, but when each graceful restart ends up consuming CPU, RAM and swap due to the processes not ending correctly, that is where the main issue is.

[ChrisRHS]

I know that I posted earlier in this thread that an apache rebuild cured my issues, however, I wanted to update and say that I am still having major problems with this. On some of the low traffic servers, it does not seem to cause any concerns, however, on my high traffic servers, this is causing a lot of issues. My load can shoot up from anywhere to 150+ to 400+, and brings the server to a crawl.

[LDHosting]

But in my case the major issue is with the Apache bug which results in a process not ending and becoming stuck.

Although no-one wants a restart, if it did end up being the only way to process the logs correctly it would be livable with, but when each graceful restart ends up consuming CPU, RAM and swap due to the processes not ending correctly, that is where the main issue is.

This is exactly the problem that we are facing. If it wasn't for the child processes getting stuck we wouldn't mind the graceful restarts, but as it is each graceful restart causes 1 slot to get 'lost' meaning that (unless we manually do a hard restart) after a few days the servers hit the MaxClients limit and sites go down.

[racakg]

Confirming the same problems on my high traffic server. Afer each graceful restart I have to manually do a hard restart.

[MMarko]

This is really a huge issue and it should never make it to RELEASE.

[ChrisRHS]

Going back to Stable did not fix the issue for me either.

[jdlightsey]

If you can reliably reproduce this problem with Apache processes becoming stuck in the "G" state after graceful restarts, please open a ticket and request that it is escalated to J.D.

[racakg]

If you can reliably reproduce this problem with Apache processes becoming stuck in the "G" state after graceful restarts, please open a ticket and request that it is escalated to J.D.

I did it immediately after I see this problem. cPanel support team was two days on my servers, but they failed to do anything about that.

The problem was temporarily gone because I disbale all rotation logs and statistics on my servers.

[mario-cPanel]

I did it immediately after I see this problem.

racakg I have sent you a PM or you may also email me a response to mario @ cpanel.net (please remove spaces)

[ikillbill]

any fix on this issue?

Our apache goes down 2-4 hours TOO

We have change bandwidth updates from 2 to 4 hours, no use..

server load is ususally under 1

[racakg]

How can I safely downgrade my cPanel to 11.24 stable? I have managed servers and my support team told me that downgrading is not safe...?!

Thanks