I tried disabling child_processes_fork_twice since I'm monitoring quite a bit of stuff and the docs say that this can help you realize some CPU savings. I'm running nagios in a VM with 4 cores and the load runs between 1 and 3. With child_processes_fork_twice disabled, the load ends up between 10 and 40. I even had a problem today where some zombie processes wedged and would not return. Nagios was still waiting for them and the checks were still showing at the top of the scheduling queue with scheduled times in the past. Other checks kept on running, but the load was in the 30's. Kill -9 would not get rid of the zombies so I had to restart nagios.
I'm running Nagios 3.5.0 on RHEL 6.4.
child_processes_fork_twice=0 drives load up
Re: child_processes_fork_twice=0 drives load up
Could you test it one more time to make sure it is reproducible, and then please submit a bug report to http://tracker.nagios.org so our core devs can take a look at this issue.
Former Nagios employee
"It is turtles. All. The. Way. Down. . . .and maybe an elephant or two."
VI VI VI - The editor of the Beast!
Come to the Dark Side.
"It is turtles. All. The. Way. Down. . . .and maybe an elephant or two."
VI VI VI - The editor of the Beast!
Come to the Dark Side.