Cisco 2960s causing Problem: Zabbix unreachable poller processes more than 75% busy

[Zabbix_To_Go]

Hi all. We are running one Production Zabbix 5.0.27 server on RHEL 7.9 monitoring servers and equipment in our organization and we have over 1000 enabled hosts/230,000 items/63,431 triggers. We are experiencing the "Problem: Zabbix unreachable poller processes more than 75% busy" due to the amount of older Cisco 2960A switches that respond slowly. When we experienced this problem in the past we increased the number of unreachable pollers and increased the SNMP timeout from 3 to 4 seconds. Increasing the number of pollers is what specifically helped the most in the past and I would like to ask if anyone can recommend troubleshooting steps or another solution to possibly reduce the amount and frequency of 2960 switches that are becoming unreachable? The 2960 switches are probably becoming unreachable due to high latency on the switch/slow response since they're old and no longer supported by Cisco. The Cisco 2960 switches will eventually be replaced, but not for years. Thanks in advance for your help.

[cyber]

I guess you cannot make those boxes answer quicker...

Whats the polling frequency in item config? if possible, increase it... (increase the number, thus making it poll less frequent..)
Add more pollers as you already have done.
You did not mention proxies .. Add proxies and split hosts up to be monitored by proxies...

[datacare]

Has anyone found a solution to this?
Zabbix unreachable poller processes more than 75% busy

This problem has persisted in my Zabbix monitor for over a week. Nothing I do seems to resolve it. Item checks and triggers all seem to work. But it seems I cannot get rid of this problem.

[LenR]

Try putting one on a dedicated proxy. Set the timeout way up. If it works, move all of them to the slum lord proxy.

[datacare]

LenR,
thank you for your response. I have this message:
“Zabbix unreachable poler process more than 76% busy”
man’s this was the post I found for it. I’m not monitoring any Cisco switches and I don’t know what host or hosts are causing this problem. Any suggestions on how I could narrow this down would be very welcome!

[cyber]

Increasing debug level of unreachable pollers should give you hints, what hosts are the ones those pollers deal with.
We dont know anything about your setup, it may aswell be, that your default amount of unreachable pollers just cannot keep up with the demand and you should just increase the amount... Default is 1... increase to 5, see what happens..