Nagios Support Forum

The check timeout being 180 seconds is from attempts to monitor windows servers via WMI. We found a number of servers took over 2 mins to respond back with the results. So the timeout got set to 180, since we shifted out focus to using SNMP. I can look at drawing that back to something reasonable. I doubt it is something that would help this problem.

I will look into setting up a Ramdisk for the perfdata.

The security team adjusted the logging to force everything to go into /var/log/messages, which they export to a log server. The amount in information is very little in the nagios.log. If you need it, I can get it for you.

mod_gearman server runs on the same server as the Nagios server.

chicjo01 wrote: mod_gearman server runs on the same server as the Nagios server.

I suppose I should clarify. Are the workers also on the same server? If so, that is definitely going to be a problem with that many checks.

If WMI checks take two minutes, I would try 130 for the timeout. The longer checks run, the more resources you are using on the system.

The mod gearman workers do run out of the same Nagios server currently. I have not had enough time to get the remote worker server cleared of firewalls and permissions to perform all the checks we need. For the mod gearman workers, I have 300 workers setup to run, so when a large number of monitors are being checked, it performs the task without using up tons of resources.

The checks are still occurring on the Nagios server itself and taking resources so following Doug's recommendation of moving the workers to a separate server should be something to work towards. I would also strongly push to install the ramdisk. It's simple to setup and can improve things pretty drastically in most cases.

I am working on getting the ramdisk setup and tested on one of the Nagios servers. It is a slow process, since I need to go through change control. I will provide updates when I get more information. Thank you for the information, I have two things to work towards.

chicjo01 wrote:It is a slow process, since I need to go through change control.

We totally understand. We'll be here!