Nagios XI 5.4.4 Appluconfiguration taking forever

[dwhitfield]

Can you post your /usr/local/nagiosxi/var/xi-sys.cfg? We've only seen the issue twice, but as this thread has been ongoing, we might as well take a look.

UPDATE: file shared with techs

[emartine]

I PMd you the file.

[tgriep]

The next time you see the orphaned message, tale a look in the worker log file for any messages that could of caused the worker to not run the checks.

Code: Select all

/var/log/mod_gearman2/mod_gearman_worker.log

Your worker.conf file has a few options that were changed, try changing the following options from

Code: Select all

enable_embedded_perl=on
use_perl_cache=o

n
to

Code: Select all

enable_embedded_perl=off
use_perl_cache=off

Save the file and run the following as root to restart the worker.

Code: Select all

service mod-gearman2-worker restart

[emartine]

As I mentioned before this is still going on for hosts in our DR environment. Only occurring on host check. I will make the changes yousuggested and see if it changes anything.

[tgriep]

I do understand that it is only the host checks that are failing. The worker log file should have some info in why they are not running 100% of the time.
From what I read, orphaned host checks mean the worker did not run them in the time allotted and returns an orphan message.
If the Nagios process, and the Gearman processes are all running on the same system, you can disable gearman in the nagios.cfg and see if that works for you.

[emartine]

I am still experiencing the orphaned hosts issue. I am not seeing anything in the logs that are related to this.

[emartine]

What should nagios.cfg look like when gearman is disabled?

[scottwilkerson]

The line with the broker_module for mod_gearman either would not be there or would be commented out

[emartine]

In DR I have removed the broker_module line for gearman and it seems to be working ok now.

Last night our production server had this same issue again and sent many pages regarding orphaned hosts checks which angered a lot of folks. Seems to be happening every few months. I have been working on this issue with seemingly no resolve... and tomorrow I am being called into a meeting to see if they want to keep this tool. I am all ears on what you guys believe I should do. I am very inclined to get rid of gearman all together.

[scottwilkerson]

With a XI system with as many resources as yours I would also highly consider getting rid of the additional points of failure and remove the mod_gearman broker.