How do I enable event handlers for workers on Nagios XI?
I couldn't find the specific option in the GUI. I can enable it by editing the worker configuration file, but clicking "sync" resets it to "no."
Nagios XI Mod-Gearman event handler
-
gwesterman
- Posts: 268
- Joined: Wed Aug 23, 2023 11:29 am
Re: Nagios XI Mod-Gearman event handler
Hi @occ,
We have disabled the eventhandlers from running on the Gearman workers because the eventhandlers that are used in XI require access to the XI server's data, files, database, etc. which the Gearman Worker cannot access so the workers cannot have that enabled.
You can always override the interface and manually edit the worker configuration files (as you've noted), but that will mean your worker is out of sync with XI and re-syncing it will override your changes.
What is your use-case for running eventhandlers via Gearman?
Thank you!
We have disabled the eventhandlers from running on the Gearman workers because the eventhandlers that are used in XI require access to the XI server's data, files, database, etc. which the Gearman Worker cannot access so the workers cannot have that enabled.
You can always override the interface and manually edit the worker configuration files (as you've noted), but that will mean your worker is out of sync with XI and re-syncing it will override your changes.
What is your use-case for running eventhandlers via Gearman?
Thank you!
-
DoubleDoubleA
- Posts: 272
- Joined: Thu Feb 09, 2017 5:07 pm
Re: Nagios XI Mod-Gearman event handler
Hi @occ,
To add just a little more context, it is entirely an XI thing. XI has default event handlers that would break if we allowed handlers to run on workers, since they are expecting to run locally with access to local info.
If your primary goal with a Nagios Mod-Gearman approach is to increase capacity by shifting compute off the XI, depending on your network architecture, you could put an https server on the worker to receive a call from your XI to run the event handler on the worker.
I gave a talk at the Nagios World Conference last month about shifting event handlers out to Event Driven Ansible, which is another decent strategy for getting compute off your XI.
Aaron
To add just a little more context, it is entirely an XI thing. XI has default event handlers that would break if we allowed handlers to run on workers, since they are expecting to run locally with access to local info.
If your primary goal with a Nagios Mod-Gearman approach is to increase capacity by shifting compute off the XI, depending on your network architecture, you could put an https server on the worker to receive a call from your XI to run the event handler on the worker.
I gave a talk at the Nagios World Conference last month about shifting event handlers out to Event Driven Ansible, which is another decent strategy for getting compute off your XI.
Aaron