Hello,

I have an issue that is starting to repeat itself on several items.

Here's my situation :

I have triggers that used to work fine but sometimes some of them stay stuck in problem state even after the issue is resolved and what is weird is that the data of the item associated to the trigger is correct and normally collected. It seems to always be a templated/discovered trigger of a windows host (to confirm).

Example:

This templated/discovered trigger expression :

{hostname:service_state[_*Service Name*_].last()}<>0

the related item is of type "Zabbix agent (active)"

_"_OK event generation" is set on "Expression"

"PROBLEM event generation mode" is set on "single"

When the service is running the item returns 0, today the service has stopped for a while then went back up, which is confirmed by the graph of the item, yet the trigger is stuck on problem state... restarting agent and server didn't help, clearing history/trends, disabling/re-enabling the item or the trigger didn't help either, only fix i have found so far is deleting the discovered item (or trigger i don't remember and I didn't fix today's issue to understand how to fix this behavior permanently), then when it gets created again by the discovery, it solves the issue.

The server is running a quite old version (3.4.1 rev 71734) and I've seen similar issues on this JIRA but on server version 3.2, i've read those issues should be fixed in the verison I run.

But still, I have tried this query :

update triggers set value=0 where value=1 and triggerid not in (select distinct objectid from problem);

And it did not solved the issue as the said triggers didn't appear in the list of affected rows anyway.

So I don't know what to look for here... any help would be greatly appreciated, thanks.