False positives avoidance

[hsmith]

My last reply was almost 9 months ago. I will need to review this and see what the original question was before I can respond.

Thanks Eric!

[eloyd]

The following design should work: a simple addon that instead of checking the service, compares the JSON reply from multiple Nagios servers and results in the concluded state based on its own logic. No event handler is required.
Any comments?

Okay, this is not the same poster as the original poster, so I'm assuming you want to do something similar? Here's my advice, which is basically what I said before:

If you have three checks, and want to notify when two or more of them are in a failure state, then you need to use BPI. BPI needs to run on a single Nagios server to check those checks. I would implement this as: passive checks that send/forward results to a single Nagios server, run BPI on that Nagios server, and have it alert. Since that Nagios server may fail, I would replicate this across all three Nagios servers so that they are all potentially going to notify if something goes wrong.

This is not elegant, but cross-server monitoring and notification is not easy or straightforward, and I'm still not sure why the OP wants to do what they want to do in the first place.

Not sure where JSON comes into play here, but I'm assuming you're talking about Nagios APIs so one Nagios server could query the others, get the service check result, and then go from there? That's doable, but you are again limited to - what if that Nagios server is down? Unless you introduce a fourth Nagios server to monitor the three that are monitoring the service....

[hsmith]

The following design should work: a simple addon that instead of checking the service, compares the JSON reply from multiple Nagios servers and results in the concluded state based on its own logic. No event handler is required.
Any comments?

Okay, this is not the same poster as the original poster, so I'm assuming you want to do something similar? Here's my advice, which is basically what I said before:

If you have three checks, and want to notify when two or more of them are in a failure state, then you need to use BPI. BPI needs to run on a single Nagios server to check those checks. I would implement this as: passive checks that send/forward results to a single Nagios server, run BPI on that Nagios server, and have it alert. Since that Nagios server may fail, I would replicate this across all three Nagios servers so that they are all potentially going to notify if something goes wrong.

This is not elegant, but cross-server monitoring and notification is not easy or straightforward, and I'm still not sure why the OP wants to do what they want to do in the first place.

Not sure where JSON comes into play here, but I'm assuming you're talking about Nagios APIs so one Nagios server could query the others, get the service check result, and then go from there? That's doable, but you are again limited to - what if that Nagios server is down? Unless you introduce a fourth Nagios server to monitor the three that are monitoring the service....

Thanks Eric.

@meliezer, let us know what you come up with.

Thank you!