Services crashes every sunday at 5:48am

[AFRC]

Hi,

I have this new server (3 weeks old). Already came with Plesk 12.5
We are running on mediatemple:

CPU: Intel(R) Xeon(R) CPU E5-2630 0 @ 2.30GHz (24 core(s))
Version: Plesk v12.5.30_build1205150826.19 os_CentOS 7
OS: CentOS Linux 7.1.1503 (Core)
Memory: 8GB

This is the 3rd Sunday that websites stopped responding at exactly the same day and same hour: Sunday, 5:48am

Plesk does not crash since i can login and reboot and everything starts working again. I guess that's some services (apache, mysql, etc) that stop working.

We dont have any backups or crons running this hour or close to this hour. But it seems this is something that is scheduled.
I cant find anything on the logs, at least the ones that are available on plesk.

Any what idea what's causing this and what to do to resolve it?

Let me know if you want me to give some more info.

Thanks.

 

[abdi]

Automatic Updates, Are they disabled?

 

[AFRC]

Thanks for the quick reply.

Automatic updates are on and i just noticed this:

Plesk version: 12.5.30 Update #9 , last updated at Dec 10, 2015 05:48 AM

Although it was not updated today (Dec 13 Sunday, 5:48am), but it crashed.
Curiously it didn't crash on the last update date (Dec 10, 2015 05:48 AM)
It happens only in Sunday's (Dec 10 was thursday)

Any thoughts on how to troubleshoot this?
I know i can disable automatic updates, but it wont solve the problem and i bet you guys want to investigate and know what's going on.

Let me know.
Thanks

 

[Chris1]

Hi @AFRC

Check this out, you may be experiencing a similar issue.

http://talk.plesk.com/threads/apach...rt-causes-apache-segfault.335534/#post-792072

 

[AFRC]

@Chris1 Thanks for pointing me that thread. I tried to search, but honestly i didn't know what was the problem either. I read the whole thread, but im not sure i can make all those tests because it's a production server with high traffic.

I see your last post is from 1 December.
Did you solved the problem? Any solution you want to share? Any word from the plesk team?

Thanks.

 

[Chris1]

Hi @AFRC

Yeah Apache stopped crashing on me after changing the "reload" command to "restart" in the following files:

/etc/logrotate.d/httpd
/etc/logrotate.d/mod_security

I'd recommend giving it a try.

 

[AFRC]

Thanks. I'll try that, however, since this only occurs on Sunday, i'll have to wait.

 

[AFRC]

So i found more info regarding this specific hour.
As far as i understand this has to do with logrotation:

This is the only entries i have on the /var/log/httpd/error_log file. Every day at the same hour:

(replaced [--REMOVED-IP] with the server ip)

[Fri Dec 18 05:48:36.331504 2015] [mpm_event:notice] [pid 23980:tid 140354401622080] AH00492: caught SIGWINCH, shutting down gracefully
[Fri Dec 18 05:48:39.868220 2015] [suexec:notice] [pid 24740:tid 140031994030144] AH01232: suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Fri Dec 18 05:48:39.870261 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH01909: RSA certificate configured for lists:443 does NOT include an ID which matches the server name
[Fri Dec 18 05:48:39.870673 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH01909: RSA certificate configured for default-[--REMOVED-IP]]:443 does NOT include an ID which matches the server name
[Fri Dec 18 05:48:39.871690 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH02292: Init: Name-based SSL virtual hosts only work for clients with TLS server name indication support (RFC 4366)
[Fri Dec 18 05:48:39.924151 2015] [auth_digest:notice] [pid 24740:tid 140031994030144] AH01757: generating secret for digest authentication ...
[Fri Dec 18 05:48:39.925046 2015] [lbmethod_heartbeat:notice] [pid 24740:tid 140031994030144] AH02282: No slotmem from mod_heartmonitor
[Fri Dec 18 05:48:39.926625 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH01909: RSA certificate configured for lists:443 does NOT include an ID which matches the server name
[Fri Dec 18 05:48:39.927032 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH01909: RSA certificate configured for default-[--REMOVED-IP]]:443 does NOT include an ID which matches the server name
[Fri Dec 18 05:48:39.928145 2015] [ssl:warn] [pid 24740:tid 140031994030144] AH02292: Init: Name-based SSL virtual hosts only work for clients with TLS server name indication support (RFC 4366)
[Fri Dec 18 05:48:39.929098 2015] [:notice] [pid 24740:tid 140031994030144] mod_bw : Memory Allocated 0 bytes (each conf takes 48 bytes)
[Fri Dec 18 05:48:39.929116 2015] [:notice] [pid 24740:tid 140031994030144] mod_bw : Version 0.92 - Initialized [0 Confs]
[Fri Dec 18 05:48:39.936623 2015] [mpm_event:notice] [pid 24740:tid 140031994030144] AH00489: Apache/2.4.6 (CentOS) OpenSSL/1.0.1e-fips mod_fcgid/2.3.9 mod_perl/2.0.9-dev Perl/v5.16.3 configured -- resuming normal operations
[Fri Dec 18 05:48:39.936667 2015] [core:notice] [pid 24740:tid 140031994030144] AH00094: Command line: '/usr/sbin/httpd -D FOREGROUND'

What can i do to get rid of this errors?