Backup-Manager: Restart of Apache & Plesk

[huschi]

Since the update to Plesk 11 on a Debian 6 there are errormessages from the backup-manager:

The Plesk Web Server service on host server.nordbyte.de is down.
The problem was discovered on Jul  7, 2012 03:53 AM.

-----

The Plesk Web Server service on host server.nordbyte.de has been started on
Jul  7, 2012 03:58 AM.

-----

The Web Server (Apache) service on host server.nordbyte.de is down.
The problem was discovered on Jul  7, 2012 04:03 AM.

-----

The Web Server (Apache) service on host server.nordbyte.de has been started
on Jul  7, 2012 04:09 AM.

-----

Backup task finished.

Task was created by  with guid (5f2f0d92-de07-4e0d-adab-6c5ceae0ae7b)


Creation date is: 2012-Jul-07 03:24:02
Task status is: error

Dump full name is: web1_1207070324.tar

The error.log of apache said at this time:

[Sat Jul 07 04:00:09 2012] [error] server reached MaxClients setting, consider raising the MaxClients setting
[Sat Jul 07 04:04:07 2012] [notice] caught SIGTERM, shutting down

Followed by the usual restart-messages.

This came every night since the last Plesk-update.
It seems to me, that the Update-Manager does a lock to any server process (maybe mysql?) till the monit-scheduled-process do a restart to them.

Have anybody else that problem?
Have anybody an idea to resolve this?

Thanks!

 

[abdi]

Don't you need to increase the MaxClients for plesk apache? This instruction I hope can help...

http://kb.mediatemple.net/questions/770/(dv)+HOWTO:+Misc.+performance+tuning#dv

 

[huschi]

Sorry, but that's not the point. I don't need more locked Apache-Childs.
The point is, that something is locked until apache restarts. And it isn't only the apache. The Plesk-Server (lighttpd?) is getting down too.
Always and only at the time the Backup is running. I can start the backupmng in the console with the same effect.

huschi.

 

[SacAutos]

I too have noticed this

I just posted a reply to another thread like this. I'm also noticing a pattern of Apache and nginx tending to yo-yo during the evenings when I'm running automated backups. This started happening when I updated from Plesk 10 to 11. Could there be some kind of resource battle going on between the backup process and the web service? A battle that Apache and nginx keep losing??