<div>hi all,</div>
<div> i have a nagios 3.0.6 system monitoring 1000+ services and its load average is always more than 8.</div>
<div>because most of my service check plugin are written in perl, I tried to enable ePN. </div>
<div>everything looks ok at first, the load average was brought down to about 3 or 4.<br>but then i noticed that the average service check latency is very large, even more than 300 seconds</div>
<div>while my service check period is 1 minute. I tried to tweak the use_large_installation_tweak/check_result_reaper_frequency/</div>
<div>max_check_result_reaper_time/max_concurrent_checks, even tried to test various values of service_inter_check_delay_method</div>
<div>but the service check latency still ran very high after several hours. I also tried to turn on and off perlcache when ./configure</div>
<div>and it was also useless. host check latency is also very high, and I think it has nothing to do with the perl interpreter ... which I felt very strange about.</div>
<div>but when I accidentally turned off the embedded perl interpreter, the latency value went back to less than 5 seconds ... and the host check latency</div>
<div>was less than 1sec. everything is ok now except my server's load average ...</div>
<div> I googled the web and read the nagios document(tried almost all the methods in Tuning Nagios For Maximum Performance chapter...) but haven't</div>
<div>found the reason and solution yet. is that something wrong with my configuration file or is that a bug of 3.0.6 version? hope someone can help me.</div>
<div> best wishes to you all,</div>
<div> </div>
<div>fancyrabbit</div>
<div>Feb 27 2009</div>