CRITICAL: popen timeout received, but no child process

[jbennett]

Occasionally I will get this notification on some of my hosts. These are all Linux boxes.

According to this thread, I need to go through and edit the command.

I am using the check-host-alive check command as follows:

Code: Select all

$USER1$/check_icmp -H $HOSTADDRESS$ -w 3000.0,80% -c 5000.0,100% -p 5

What would be causing this? It happens intermittently.

I seem to remember there being a reason that we switched to check_icmp over check_ping for our use, but I'm not sure I know the reason.

[slansing]

It is possible though not plausible that the ICMP check timed out and the forked nagios process for the check was closed, unfortunately that error is not very verbose.. You should be totally safe switching to check_ping for your hosts as they are essentially the exact same thing.

[jbennett]

I wanted to post back an update to this issue.

It's still happening.

I've switched all host checks to check_ping from check_icmp and I still get a handfull of hosts that will show up with this time out.

Do I have any options at all?

[lmiltchev]

I've switched all host checks to check_ping from check_icmp and I still get a handfull of hosts that will show up with this time out.

Did you try adding the "-4" flag? Did this help?

Code: Select all

$USER1$/check_ping -H $HOSTADDRESS$ -w 3000.0,80% -c 5000.0,100% -p 5 -4

[jbennett]

I figured it out. This kept becoming an issue due to the logs not being purged in /usr/local/nagios/var/archives/

We had logs back to 2012 in there. Since we don't need them that far back, I was able to clear up a ton of space by removing them.

Hosts started coming back up immediately. Hopefully this will help prevent the issue in the future (and hopefully this solution helps someone else too).

[tmcdonald]

Nice. Can we close this up or was there more?

[jbennett]

Close it out.

Thanks!