Hello,
I would like to request that thresholds have a rearm criteria option, such that if a Trouble alarm is raised at a value of 90%, it can be configured to not clear until the polled value is 80% instead of 89%.
The particular use-case for this would be avoid flapping alerts, which we encountered for guest disk usage thresholds for VMware VMs. The 90% usage threshold would be breached, and then cleared once it goes back down 89%, and then it would go back up slightly and raise another alarm. I would like to be able to configure the alert to only clear once the polled value has reached 85% to reduce the volume of alerts.
Hey Ben,
Thanks for sharing this. you requirement is well-articulated use case, and the VMware guest disk usage flapping scenario is a great example of exactly why a rearm threshold matters in practice.
The good news is we're already working on this feature. The ability to configure a separate clear/rearm threshold (so an alert raised at 90% doesn't clear until the value drops to, say, 85%) is on our roadmap and is currently in development. It is targeted for release next quarter. We'll share an update here once it's available.
Appreciate you taking the time to put this forward!
Best regards,