High Zabbix Queue.

[drucej31]

Guys can someone please help with this problem. I've got a ridiculously high queue on my main server. I'm running approximately 10 proxies, and around 500 hosts / 30K items.

Attached is a picture of my Zabbix queue i have over 10K items and counting waiting to be updated....

The Zabbix performance parameters are all fine. This is my zabbix_server.conf main parameters...

root@lon-aai-deb-zab:/etc/zabbix# more zabbix_server.conf
StartPollers=40
StartPingers=10
StartPollersUnreachable=40
StartIPMIPollers=10
StartTrappers=20
StartDBSyncers=8

The server performance is fine ...

root@lon-aai-deb-zab:/etc/zabbix# uptime
21:26:48 up 21 days, 6:21, 3 users, load average: 0.73, 0.45, 0.45

Please can someone help it is literally driving me round the bend why these figures are so bloody high...

I've also done all of the MySQL partitioning steps so the database although is very big, it is significantly smaller than it was.

[drucej31]

Thanks for the responses

Solved the issue myself in the end... The issue was with clocks on the proxy nodes being out... Installed NTP on each server and sync'd time... Now queue is a lot happier.

[aib]

Also, from time to time, check the Graph "Zabbix Data Gathering Process Busy".
Pay attention for any processes which is constantly take >10-20% of CPU.

If any poller/pinger processes going higher - increase the amount of straring process and restart server.

I'm not sure if that numbers is high enough for your environment

StartPollers=40
StartPingers=10
StartPollersUnreachable=40
StartIPMIPollers=10
StartTrappers=20
StartDBSyncers=8

[tchjts1]

StartDBSyncers=8

I would recommend that you put this setting back to the default of 4.