Issue Watchdog kills Apache

[H9k]

Since a couple of weeks, I sometimes have the following problem happening: Apache web server is down.
Even though I have the Watchdog extension enabled, it is unable to successfully start Apache again.

Checking in syslog I see that some hours before Apache actually goes down, monit detects something is wrong and wants to restart, and fails at it because of PID issues. Then tries over and over for hours, until at some point Apache is really down.

Jul 6 02:20:04 h3 monit[6233]: 'apache' process PID changed to 19533
Jul 6 02:22:06 h3 monit[6233]: 'apache' PID has not changed
Jul 6 15:50:36 h3 apache2[30934]: Stopping web server: apache2.
Jul 6 15:50:39 h3 apache2[31057]: Starting web server: apache2.
Jul 6 15:52:20 h3 monit[6233]: 'apache' process PID changed to 31111
Jul 6 15:54:22 h3 monit[6233]: 'apache' PID has not changed
Jul 6 19:16:07 h3 monit[6233]: 'apache' process is not running
Jul 6 19:16:08 h3 monit[6233]: 'apache' trying to restart
Jul 6 19:16:08 h3 monit[6233]: 'apache' start: /opt/psa/admin/bin/websrvmng
Jul 6 19:17:09 h3 apache2[3344]: Stopping web server: apache2 failed!
Jul 6 19:17:09 h3 apache2[3344]: There are processes named 'apache2' running which do not match your pid file which are left untouched in the name of safety, Please review the situation by hand. ... (warning).
Jul 6 19:17:09 h3 systemd[1]: apache2.service: control process exited, code=exited status=1
Jul 6 19:17:09 h3 systemd[1]: Unit apache2.service entered failed state.
Jul 6 19:17:12 h3 apache2[3466]: Starting web server: apache2(98)Address already in use: AH00072: make_sock: could not bind to address 0.0.0.0:7080
Jul 6 19:17:12 h3 apache2[3466]: no listening sockets available, shutting down
Jul 6 19:17:12 h3 apache2[3466]: AH00015: Unable to open logs
Jul 6 19:17:12 h3 apache2[3466]: Action 'start' failed.
Jul 6 19:17:12 h3 apache2[3466]: The Apache error log may have more information.
Jul 6 19:17:12 h3 apache2[3466]: .
Jul 6 19:18:09 h3 monit[6233]: 'apache' failed to start
Jul 6 19:18:16 h3 monit[6233]: 'apache' process is not running
Jul 6 19:18:16 h3 monit[6233]: 'apache' trying to restart
and so on...

About 3 hours later I actually got an email from monitoring telling me nginx is down (lol).
Anyway, I had to manually kill all the apache2 processes, after which Apache was able to start normally again.

Any ideas on how I can solve this issue?
Running Plesk 12.5 on Debian 8.

 

[UFHH01]

Hi H9k,

did you consider to remove watchdog and re-install it afterwards with the Plesk autoinstaller?

Example for Debian:

apt-get --purge remove psa-watchdog
/usr/local/psa/admin/bin/autoinstaller --select-product-id plesk --select-release-current --install-component watchdog​

 

[H9k]

I did as you proposed, let's see how it goes from now on...
Alternatively I could modify the shutdown command to

/usr/bin/killall apache2 && /usr/bin/killall apache_control_adapter && /opt/psa/admin/bin/websrvmng --stop