Nagios Support Forum

Hi. This is an odd question. We are using ping for host checks and use multiple checks that access the remote server using specific TCP ports.

I want to know if I can rule out the Nagios server as a likely cause of a ping failure or a check timeout. I've checked the server load and iowait stats for the relevant time periods. In each case, they are within normal limits (iowait around 0.27% and 1-minute load less than 3.)

Is there another place to look? Is there a way to examine a "check processing queue" length relative to the times that Nagios declares a specific check has timed out/failed? Or some other way to give weight to the idea that some checks are marked as failed because Nagios did not process it in time?

My initial thought is that a timeout has to be caused by the network or the remote server. That would include the agent on the server as "the remote server."

Thanks!

Generally speaking I would say the likely hood it is the Nagios server causing a ping check to fail is very low. Especially multiple times before notification.

You can specify in your ping check how many times you want the ping to be attempted, and that number is used in the calculation for % packet loss.

awilson wrote:Or some other way to give weight to the idea that some checks are marked as failed because Nagios did not process it in time?

Nagios doesn't mark checks failed because it couldn't process them in time, plugins can and do timeout, and if no response is returned before the timeout it will error, but that isn't because the plugin wasn't run, it is because it didn't receive a response.

Thanks!

Let us know if we can be of further assistance