Dear Support Community,
At our organisation we're using Zabbix (5.2) to monitor multiple environments both frontends and backends. We are currently struggling with Zabbix and HA Proxy and using the Zabbix HA Proxy templates.
For some reason when there is a problem now on the HA Proxy backends the problem only gets created after 40+ minutes. When viewing the HA Proxy discovery and latest data we can see that the status of this backend is in fact DOWN so this part seems to work as expected.
The interesting part is that once the trigger is created after the 40+ minutes have passed it does not get long to resolve once the backend is up again, this takes about a couple of minutes, 5 at most.
This is obviously concerning as we won't know if a backend is DOWN and the Problem notification will be delayed by over 40+ minutes. We noticed this since one of our dashboards stopped reporting this problem directly with a trigger label icon.
This is how the trigger is setup:
Trigger should initiate once 5 checks have been performed where the server is down. The check is decided by the parent which is HAProxy by HTTP
The preprocessing step looks like this:
Trigger prototype below, it is set to "create enabled" and discovery is also set to yes.
HAProxy by HTTP: Get stats is set to perform an interval check every minute:
In Latest data we can also see this. and the backend on the top A_APP-11-NAME-WIS-Out is also detected every minute.
However for some reason the trigger does not seem to work or it is heavily delayed. As far as we know nothing has changed recently and we are planning an upgrade, however as it is right now it is not working as expected and we're unable to figure out why also. Other triggers seem to work fine for the HA Proxy, though unsure if they're delayed or not but we do get notified when certificates are about to expire for example.
Any help or direction is much appreciated.
Many thanks.
Maher
At our organisation we're using Zabbix (5.2) to monitor multiple environments both frontends and backends. We are currently struggling with Zabbix and HA Proxy and using the Zabbix HA Proxy templates.
For some reason when there is a problem now on the HA Proxy backends the problem only gets created after 40+ minutes. When viewing the HA Proxy discovery and latest data we can see that the status of this backend is in fact DOWN so this part seems to work as expected.
The interesting part is that once the trigger is created after the 40+ minutes have passed it does not get long to resolve once the backend is up again, this takes about a couple of minutes, 5 at most.
This is obviously concerning as we won't know if a backend is DOWN and the Problem notification will be delayed by over 40+ minutes. We noticed this since one of our dashboards stopped reporting this problem directly with a trigger label icon.
This is how the trigger is setup:
Trigger should initiate once 5 checks have been performed where the server is down. The check is decided by the parent which is HAProxy by HTTP
The preprocessing step looks like this:
Trigger prototype below, it is set to "create enabled" and discovery is also set to yes.
HAProxy by HTTP: Get stats is set to perform an interval check every minute:
In Latest data we can also see this. and the backend on the top A_APP-11-NAME-WIS-Out is also detected every minute.
However for some reason the trigger does not seem to work or it is heavily delayed. As far as we know nothing has changed recently and we are planning an upgrade, however as it is right now it is not working as expected and we're unable to figure out why also. Other triggers seem to work fine for the HA Proxy, though unsure if they're delayed or not but we do get notified when certificates are about to expire for example.
Any help or direction is much appreciated.
Many thanks.
Maher
Comment