NSClient check timeouts fixed, but tmp Check files continue

[vAJ]

Yes. check_nt hitting NSClient.

[slansing]

Apologies if this has been re-verified, but have the timeouts for these services been adjusted as well? As well as on the NSClient server's side in it's NSC.ini file? These checks may be just barely on the cusp of the timeout as the max execution time was at "10.10s" so they may not alert as a timeout.

[vAJ]

I've bump up the check command in Nagios:

Code: Select all

check_xi_service_nsclient = $USER1$/check_nt -H $HOSTADDRESS$ -s "$ARG1$" -p 12489 -v $ARG2$ $ARG3$ $ARG4$ -t 15

I haven't changed anything on the clients.

[abrist]

Check your clients' nsclient/nsc config file for the:

Code: Select all

socket_timeout=

Increase it to 30.
You will find the timeout setting under the header:

Code: Select all

[NSClient]

Increase your check command in nagios to 30 seconds as well.

[vAJ]

I've increased the service timeouts. I'll have to wait for the next maintenance window to script out the new INI and restart of NSClient on the ~800 hosts in production.

Great news though! (I just saved a ton of money...)

No - really, the check files are non-existent today. I see them come in & out within a seconds. None have stayed since this change was made today. Also checked the event log, no timeouts for NSClient checks at all today.

I think I may not even make the change at the client side.

[slansing]

Great, this should be able to be controlled in the extent you require through the Nagios side, it is normal to see some tmp check files as the checks go out for their respective object.