Host-alive check AND ping service?

[oneida]

Hi there,

I am currently in the process of merging two Nagios core servers into a single NagiosXI server. Currently we're monitoring about 540 hosts and each is set up with host alive checks and a service ping check. Is it just me, or is this rather redundant and a waste of resources?

If the PING service fails, the host-alive check will also fail. If the host-alive check fails, the PING service check is irrelevant because the host is considered down. So it just seems that with the way we currently have this setup, we're doubling the number of checks that need to be performed for no reason.

Is this a typical setup or would you recommend removing the PING service checks from every host? Thanks in advance!

[slansing]

If you have host alive configured it is running check_icmp which is a ping service, you could go without running one or the other as they will return the same information if your host/service goes down.

[oneida]

I disabled all service PING checks and left host-alive checking to running "$USER1$/check_icmp -H $HOSTADDRESS$ -w 3000.0,80% -c 5000.0,100% -p 5".

Much better. Cleared up 540 unnecessary service checks. The reason I was apprehensive about doing this is that in our older versions of Nagios Core (currently in production), warning messages are generated when a host does not have a service associated with it. That does not seem to be the case now.

I also noticed that if a service is created and then removed, the configuration file it creates does not seem to go away. In my case, the service (and config file) referenced a service group called "PING-Group". Even though I removed the service through the WebUI and applied the configuration changes, I was not able to deactivate or remove the service group "Ping-Group." This was because that config file still existed and was referencing it. To fix this, I had to delete this config file from the server (from /usr/local/nagios/etc/services/). After that, no problems.

Not sure if this is a known issue...

Anyway, thank you much for the help.

-Matt

[scottwilkerson]

Not sure if this is a known issue...

Actually it is and I believe this is fixed in 2012R1.4 due to be released today or tomorrow.