Nagios Mod_gearman worker overloaded

[mbellerue]

Okay, very first thing, let's go through a delete/write/verify. In XI, go to Configure -> Core Configuration Manager -> Config File Management. Click on the Delete Files button, then Write Configs button, then Verify Files button. Assuming the verify completes successfully, click Restart Nagios Core.

It also looks like you're trying to access the server via HTTPS, but it doesn't look like XI is responding on HTTPS. Can you go to Admin -> System Settings -> General tab, and take a screenshot of that page and send it in?

Along with that, if you could also send in your /etc/httpd/ configuration files. We'll see what we can do to get https up and working properly.

Another thing I noticed, you sent in a gearman config for worker_weeus01plnagi01, but that worker does not show up in gearman top. Looking at that config file, it looks like it's configured to talk to localhost for the gearman job server, where worker_weeus01plnagi03 is set to talk to nagios.ah.nl. So worker_weeus01plnagi01 isn't pulling its weight, assuming it is supposed to be a gearman worker.

[rajsshah]

I have sent you a PM with required details

[mbellerue]

The purpose of the Delete/Write/Verify process is to make sure that only 1 Nagios instance is running, and it is running with the proper, production configuration. We suspect a lot of those defunct PHP processes are the result of a ghost Nagios instance. This process should take care of that.

[rajsshah]

I can do the steps that you said , however as I mentioned earlier , during issue we do a complete restart of nagios processes but after a day or 2 those defunct processes comes again . SO there must be some root cause for this . correct ?? the steps that you mentioned is only to make sure that you are restarting nagios in a clear way . But root cause is still unkown.

[mbellerue]

I don't know that a restart of the Nagios services kills off any ghost instances of Nagios. You can check this, next time this happens you can shutdown the Nagios services, and then do ps -aux | grep nagios and see what comes back. But all of the defunct PHP processes aren't representative of ghost instances, but rather the checks that a ghost instance tries to kick off. The Delete/Write/Verify process should kill off any ghost instances, and it will make sure that the correct configuration files are written to disk.