Service Severity Weighting

[highpeak]

We have 4 load balanced web servers providing a service. I have created a service with the 4 web servers as children. If the web service fails on one server, it is a "high" severity for that server, but given the service is is still running, it should be a "warning" if one of them fails and "medium" if two fail, etc.

OpenNMS has a powerful mapping and reduction algorithm as a way to define the parent status based on various child statuses

https://docs.opennms.org/opennms/releases/24.0.0/guide-admin/guide-admin.html#_business_service_definition

Does Zabbix have anything remotely similar to this?

[tim.mooney]

Isn't it possible to do what you're asking about purely with a few carefully-constructed triggers? Something like example 13 from the trigger documentation?

This operates somewhat separately from your IT Service, but it would (I think) satisfy your goal of getting a range of severity levels based on how many of the backend systems have problem results for a particular trigger.