Hello ! :)

Regarding XCache, can you tell me if it works better with the highest possible value for xcache.size ? Or should xcache.size be set to a reasonnable value in any case ?

I'm asking this because :
- I had OOms when xcache.size was set at 64M. So I decided to increase xcache.size up to 96M.
- when looking at the administration panel, I notice that there are blue and green bars :

but after some time, when the bars are near to be filled, everything is cleared :

Is this normal ?

Thank you for your help ! :)

2nd SS is wierdo. I am using 1.2.x xcache and I have never seen that happening. May be it stops working or it crashes?

Try updating to version 1.3.1? May be this guy having same issue?

http://forum.lighttpd.net/topic/101380#20523

I would check through your logs, are you sure Apache isn't being restarted?

You are right. I saw in /usr/local/apache/logs/error_log that Apache is being restarted about 14 times each day... Example :

[Fri Jan 07 00:15:52 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 02:12:59 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 04:09:51 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 06:06:35 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 08:03:20 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 10:00:09 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 12:13:41 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 13:12:35 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 14:05:53 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 16:02:50 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 18:16:25 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 20:13:28 2011] [notice] Graceful restart requested, doing restart

[Fri Jan 07 22:10:37 2011] [notice] Graceful restart requested, doing restart

The complete error message is :

[Thu Jan 13 20:05:54 2011] [notice] Graceful restart requested, doing restart

[Thu Jan 13 20:05:54 2011] [warn] WARNING: Attempt to change ServerLimit ignored during restart

[Thu Jan 13 20:05:54 2011] [notice] Apache/2.2.17 (Unix) mod_ssl/2.2.17 OpenSSL/0.9.8e-fips-rhel5 mod_auth_passthrough/2.1 mod_bwlimited/1.4 FrontPage/5.0.2.2635 PHP/5.2.9 configured -- resuming normal operations

The Apache uptime isn't changed in spite of all the restarts.

What should I do then ?...

Thank you for your help ! :).

you should investigate why apache is restarting.

The way you said things seems like a cron is controlling the apache restars due to it being gracefull, If there were any errors it would show as --force.

The only other alternate is the timeout for apache, If this is to low and many things are running during that time then it would restart.

Try running:

/scripts/fixeverything

And see how that goes, Another thing is if your online and it does a restart quickiy check the messages and post the output:

tail -f /var/log/messages

This would show if anything other than apache is causing it.

Are you running cPanel, If so run the following:

/scripts/upcp --force