Passive Checks

[KWGagel]

I'd like to move to passive checks and the NCPA agent. We're still using NSClient++ on our Windows based machines with NCPA on linux. I've got one server working fully with NCPA (and NSClient++) but I'm confused on the onboarding of passive checks.

It seems that each host I switch over to NCPA passive checks will result in one new unconfigured object which I then need to configure before it gets monitored properly. During testing I noticed errors in my new config and made changes which resulted new unconfigured objects and additions to the host when the objects were applied. The only way I see on how to edit that was to delete the host from Nagios and wait for the 'new' unconfigured objects then apply them again.

Is there a better way to go about this?

[cdietsch]

Hey @KWGagel,

Unfortunately deleting the host from Nagios and re-adding the new unconfigured object is the simplest way that I know of also.

[kg2857]

The OP might try investigating the difference between the old hosts and the new ones created from the unconfigured objects.

[cdietsch]

Ahh thanks @kg2857 for help clarifying that.

@KWGagel It sounds when you are adding the passive the checks they are showing up in the old host object. This is happening because the old host object and the new unconfigured object have the same hostname in Nagios.

One thing you can do to prevent this from happening is on the client machine's ncpa.cfg, there is hostname setting in the [nrdp] section. You could set this to something like regular-hostname-passive. The client will then show up in unconfigured objects with that hostname that you set, so it will be added as a new host object and not affect the existing host object.

[KWGagel]

Ahh thanks @kg2857 for help clarifying that.

@KWGagel It sounds when you are adding the passive the checks they are showing up in the old host object. This is happening because the old host object and the new unconfigured object have the same hostname in Nagios.

One thing you can do to prevent this from happening is on the client machine's ncpa.cfg, there is hostname setting in the [nrdp] section. You could set this to something like regular-hostname-passive. The client will then show up in unconfigured objects with that hostname that you set, so it will be added as a new host object and not affect the existing host object.

No, they're not showing up in the old host object. The old host object is named using lowercase and the new host object is all upper case. What I was referring to were the services that were configured from the unconfigured wizard. Due to errors I made in the cfg file NCPA was unable to parse some lines correctly. After fixing and then changing others when the fixed (new) items showed up in unconfigured objects I simply used the wizard again and it didn't remove anything, it just appended the fixed items as new and left the defective entries alone.

As I mentioned in the original post, the only way I found to 'fix' that was to delete the host and use the wizard again on the unconfigured objects.

[cdietsch]

Hey @KWGagel,

Okay, thank you for the further clarification. An alternative to removing the entire host object, is you should be able to go into the CCM -> Services and remove the old/bad passive service check(s) for that host. Hopefully this helps!

[KWGagel]

That will be helpful, I'll run some tests and try that out.

Is there a way to pre-populate services that will be immediately associated with new unconfigured objects or is the only option to use the gear/wizard each time there is new unconfigured objects?