Closed ranjithkodumbu closed 7 years ago
result_workers
is deprecated and always set to 1. The nagios core processes results serially, so more result_worker wouldn't increase the performance.
Hi Sni,
Thanks a lot for your response, please refer the below screenshot where in you can find check_results is "219681". We are using 2,00,000 monitoring services with 10 worker servers, workers are processing as expected, however check_results queue job waiting is high. Is there anyway to make check_results to work fast, I understand from your comment that nagios core process results serially, but still I am awaiting for your advice to make it work.
Thanks a lot for your support.
Please refer to the nagios xi support on how to make nagios faster. I am unable to support commercial products.
Hi sni,
Thanks a lot for the response. so you meant to say, check_results queue is building because of nagios is not processing? currently server load is less than 1% . I suspect, by increasing check_result worker might bring down queue build. Is there any way I can try that?
Thanks
Well, there are older releases which still have this option available. But i'd say there is a different issue somewhere. There must be a reason why these results pile up while the box is idling. There are only 3 blocking events in the core which are eventhandler, notifications and on-demand-hostchecks. Mod-Gearman fixes 2 of them, but notifications are still serial and blocking. Might be worth to look into this. Some future Mod-Gearman release will btw have notifications support to address that as well.
Hi All,
I am using gearman with Nagios xi. I have tried adding "result_workers=2" parameter in "/etc/mod_gearman/module.conf" file and restarted nagios and gearmand service. when I ran gearman_top command, I could see only 1 worker for check_result queue. Please help me on this issue.
broker_module=/usr/lib64/mod_gearman/mod_gearman_nagios4.o config=/etc/mod_gearman/module.conf
Thanks, Ranjith Kumar R