Page 2 of 3
Re: Slow Nagios server
Posted: Thu Apr 08, 2021 3:00 pm
by dchurch
What does the following command produce?
Code: Select all
mysql -uroot -pnagiosxi -e 'show full processlist'
Re: Slow Nagios server
Posted: Mon Apr 12, 2021 5:30 am
by centricrm
File is sent in PM
wkr, Humphrey
Re: Slow Nagios server
Posted: Mon Apr 12, 2021 4:00 pm
by benjaminsmith
Hi
@centricrm,
Apologies for the inconvenience, can you PM me a copy of the profile? Dan is out of the office today.
Thanks!, Benjamin
Re: Slow Nagios server
Posted: Tue Apr 13, 2021 3:15 am
by centricrm
profile is sent in PM.
wkr, Humphrey
Re: Slow Nagios server
Posted: Tue Apr 13, 2021 2:08 pm
by dchurch
One thing that causes slowness is if there is a bunch of users sitting on the service status pages in a bunch of tabs showing 1000 or all services, the auto-refresh keeps occurring and some of those queries are not setup to be as efficient as they could be on large systems like yours. Outstanding queries cause MySQL CPU usage to rise, which causes more important queries to wait, and become backlogged.
Try adjusting Admin > Performance Settings > Dashlets and change the Dashlet Refresh Multiplier from 1000 (the default) to like 5000 or 10000 to see if that helps.
Re: Slow Nagios server
Posted: Tue Apr 13, 2021 2:10 pm
by dchurch
If that doesn't work, you could PM the Admin > System Profile > View System Info to me, or you could check on this yourself:
If you have the old active_directory/ldap_auth components, having those on XI 5+ can conflict and cause slowness with the new one (ldap_ad_integration)
Re: Slow Nagios server
Posted: Wed Apr 14, 2021 3:45 pm
by centricrm
I am evaluating the last change (Dashlet Refresh Multiplier changed from 1000 to 5000).
The profile is sent in a PM.
wkr Humphrey
Re: Slow Nagios server
Posted: Thu Apr 15, 2021 11:00 am
by dchurch
Just checking in. Did increasing the Dashlet Refresh Multiplier value have any effect so far?
Re: Slow Nagios server
Posted: Thu Apr 15, 2021 11:45 am
by centricrm
The response of the server varies from good to very bad.
I'm not sure if the Dashlet Refresh Multiplier change improved the performance.
At this very moment I see a lot of processes using 100% and one 181% CPU which makes the Host Status screen slow to appear.
See output of top below.
top - 18:44:19 up 14 days, 19:25, 2 users, load average: 5.64, 6.05, 5.90
Tasks: 383 total, 6 running, 285 sleeping, 0 stopped, 92 zombie
%Cpu(s): 39.9 us, 35.5 sy, 0.0 ni, 24.6 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
KiB Mem : 16265904 total, 2671364 free, 3709132 used, 9885408 buff/cache
KiB Swap: 8380412 total, 8364796 free, 15616 used. 11304444 avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
32711 mysql 20 0 4903800 1.2g 9804 S 183.1 7.6 1800:25 mysqld
683 nagios 20 0 1135416 124400 3712 R 99.7 0.8 763:36.27 nagios
690 nagios 20 0 11116 1428 760 R 99.7 0.0 811:03.19 nagios
692 nagios 20 0 11244 1432 760 R 99.7 0.0 811:17.46 nagios
703 nagios 20 0 11244 1440 760 R 99.7 0.0 810:57.56 nagios
1245 apache 20 0 704756 127388 6716 S 7.0 0.8 4:55.29 httpd
1244 apache 20 0 721140 144036 6784 S 6.0 0.9 4:49.75 httpd
There are also some zombie processes.
Re: Slow Nagios server
Posted: Fri Apr 16, 2021 10:33 am
by dchurch
Now what does the following command produce?
Code: Select all
mysql -uroot -pnagiosxi -e 'show full processlist'