Monitoring outages on MikroTik devices in Zabbix 5

[jan.brodecky]

Hello,

I am running Zabbix 5, and some of MikroTik devices are experiencing regular data collection outages. I use the same Template for all MikroTik devices, and data is collected via SNMP. The monitoring configuration is the same across all devices.

For example, MikroTik 1 (RouterOS version 6.48.6) is monitored without issues, but MikroTik 2 (RouterOS version 6.45.8) shows short monitoring outages. During these outages, all values in the Latest data show 0%, but the system does not trigger alerts or send notifications.

Additionally, I am providing the monitoring output (screenshot) from MikroTik 3 (RouterOS 6.48.6), where monitoring frequently fails. MikroTik 3 has the same RouterOS version as MikroTik 1 (6.48.6), which is monitored correctly. The configuration is also identical (template + SNMP).

Attached are screenshots illustrating the issues.

I have not been able to find any correlation between the MikroTik devices where monitoring works correctly and those where it does not. The monitoring outages are regular but vary in duration and the intervals between them for each MikroTik.

If any additional information is required, please let me know.

Thank you in advance for your help and have a great day.

[NgRox]

Next, in this case, a more precise analysis is required, so:

1 - Check the internal metrics of your Zabbix Server/Proxy, specifically the "Utilization of Poller." If the value is low, we can rule out the issue of having too few poller workers. Let's move on to the next step:

2 - Verify if the Bulk request option is enabled in the host configuration. If it is, disable it and check how the polling behaves. If part of the polling fails, it may discard the rest. If you disable it and the problem persists, proceed to the next step:

3 - Check if there is more than 1 community defined on the host, I had collection problems because 2 communities were defined and thus caused collection errors. I didn't find out if that was the case, but after removing it, it resolved my case.

4 - Check the Zabbix Server/Proxy logs for any connection errors such as: "temporarily disabling SNMP agent checks on host 'XXXXX': interface unavailable" or "failed: first network error, wait for xx seconds" If you find these errors related to the hosts, then there is a connectivity issue between Zabbix and these hosts. This could be due to SNMP connection limits on the host (configured on the host, not in Zabbix).

Hope this helps!