Repeated firing of alerts until resolved or acknowledged

[jediwombat]

Hi there,

I'm quite new to Zabbix and have inherited a half-setup system. We recently had a disk space incident, and I've proven via the action log that the alert did send when it first breached the 90% threshold, but was then silent until the space was cleared.

What I'd like is for the alert to continue to send every X hours while it is over the critical threshold. I've done some googling and found a few suggestions, but some are quite old and don't match modern Zabbix. The best information I can find suggests we should change the steps in the operation, but every post I can find on it assumes some knowledge I don't have yet. I'm a bit fuzzy on how all the components link together, i.e. the relationship between items, triggers, actions, operations, and so on.

The operation has a default step duration of 1h, and is configured thus:



If I change the step duration to, say, 6 hours, and change the steps be 1-0, will that only fire the alert every 6 hours? I'm hesitant to change this in case it causes it to fire every time the item is checked on the host, i.e. every 5 minutes.

Any tips or advice would be great!

Thanks,
Benjamin.

[ISiroshtan]

Hey

Yes, your suggestion is correct. Changing step duration to 6 hours and setting Steps 1 - 0 would cause this operation to be repeated every 6 hours for each trigger that matches this action condition until problem is resolved