<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> </head> <body bgcolor="#ffffff" text="#000000"> Alessandro Ren wrote: <blockquote cite="mid:47B659D1.7050809@opservices.com.br" type="cite"> <pre wrap="">Hendrik Bäcker wrote: </pre> <blockquote type="cite"> <pre wrap="">Hi, Alessandro Ren schrieb: </pre> <blockquote type="cite"> <pre wrap=""> I got the latency to stabilize disabling the process_performance_data in nagios.cfg. Enabling it again cause the latency do start to go up. Any know bug in this part do the code? About the performance data, does it affect the nagios latency? What I mean is, does nagios starts another thread to deal with performance data or it waits for the data to get processed by the external command before returning, thus adding to the service check latency? Thanks. </pre> </blockquote> <pre wrap="">I think you are executing the perfdata command after every service check (on the fly) ? If this should be true than I can imagine of performance problems. 1 action to start a plugin 1 action to reap the output 1 action to handle perfdata per service... ouch. What addon are you using to handle performance data? - Hendrik </pre> </blockquote> <pre wrap=""> I am using perfparse to handle performance data writing realtime to the database, but the strange thing is the latency that keeps going up. This does not happen in the nagios 2.x. []s. </pre> </blockquote> I am attaching a new PDF with the latency progression. It keeps on going up with the perfparse on, and check the drop and stability in the other nagios, opmon_pgsql. Any ideias? What intrigues me is the fact that this not happens in nagios2.x, so I dont think perfparse is to be blame for the problem. []s. </body> </html>