Trigger didn't fire off when service went down.

[Aaron88]

Hello,

I created an item and a trigger for the service MYSQL80. The service went down this morning, but it did not alert us. Does anyone have any ideas?

Template: MYSQL by Zabbix agent 2
Item:


Trigger:

[SanMonitor]

Hey Aaron88,

could you show something like 'last 500 values', this trigger would only become active with state '6', maybe it was a different value this morning?

[Aaron88]

Would this be in the expression? I tried using ChatGPT and CoPilot and neither were not helpful with creating expressions. I tried searching what the values are and the meaning but could not find anything. When I set it to zero, I got alerts for the service running. So, I am assuming I have the wrong value to fire off the alert.

[cyber]

Can you show the gathered data from that period, when trigger should have fired?

[Aaron88]

Cyber, are you referring to this graph? I have attached a screenshot.

[cyber]

Cyber, are you referring to this graph? I have attached a screenshot.

basically yes... you can change (from top right corner) also to values ...


But I think this graph already shows, that there are really values as expected, not some other values... And your trigger should work...:/ your service changed status already in the evening.. 6PM-ish...
you can try, what happens if you set trigger to ">0" instead of "=6" But it should really not matter...

[Aaron88]

Okay, thank you for your reply. When I changed the value to "0," it sent alerts for "Running." I then changed it to "1" and waited until the service stopped. This happened last Friday, and we were not alerted. So, I changed the value again to "6." (I can't stop the service because it's a production server.)

[Aaron88]

When I search by values and the last 500 values, I see this on the day the service went down. I see there is an item and a trigger created by the Windows service discovery. Is Zabbix case sensitive? The actual name of the service is MySQL80. The item and trigger I created I used "MYSQL80".



I see there's an item and a trigger created by the Windows service discovery. I have attached it.

Questions:

1. Would it be better to edit this item and the trigger?
2. How do I edit it? It does not allow me to do it from that Window
3. Would I modify the expression like in the trigger I created?

[cyber]

yes, zabbix is case sensitive. At least in Linux side... For windows... ugh... You would not have had any values, if it would be..:P

Items/triggers from discovery are created based on prototypes in that discovery rule. Changes will affect all items/triggers based on them. You cannot edit already created items/triggers directly. If you change prototype, discovery needs to run again to apply changes, it is not instant

That trigger on the pic does not do whats written in its description "Service has a state other than "Running" for a last 10 times"
"min value during last 10 checks is 0 and there is no data for 10 minutes" will fire only if it has been working (opposed to "not in running state") and suddenly there is no data any more, making first part of expressio useless...
It should be "<>0" and that nodata delays it even more, I guess to avoid something during restarts or something...

And it does not seem to be from default windows template, which looks like this one

[Aaron88]

Thank you for your input and help. What would be the best practice to get alerts when MySQL80 service stops or not in a "Running" state? I am assuming the item and trigger I created won't be useful.

[cyber]

You have everything here already ... your item is pretty much ok, follow the case. Trigger example is also present.. Just experiment.