Notification endpoints (media) based on server group

[mklatsky]

I would like to send alerts based on server category. For example, critical production servers should send alerts to our global alerting email, 24x7. Our test servers should only send alerts during business hours, and to a different email address. I have done this with Nagios, but I am not seeing an obvious way to do so with Zabbix. Any help would be most appreciated.

[mklatsky]

Can you elaborate a little bit? I have been perusing documentation on this- I have tagged my hosts, but I am still unclear on how to use those tags to add triggers (or events?) based on those tags.

For example, we have a webscenario which checks for the existence of a status word in a web page.

For servers tagged Category: Production, I'd like alerts to occur after 2 failures.

For servers tagged Category: Testing, I'd like alerts to occur after 10 failures.

Thank you- I appreciate the assistance.

[tim.mooney]

For servers tagged Category: Production, I'd like alerts to occur after 2 failures.

For servers tagged Category: Testing, I'd like alerts to occur after 10 failures.

My site started with Zabbix before "tags" were a thing, so we implemented different alert regimes based on which host group the host belongs to.

In our case, our triggers are the same between production and devel (modulo any difference in host macros, which we use heavily in our triggers to make it easy to tailor things like %used disk space for /var, etc), so even for a devel host the problem is detected right away. The difference is in alerting, which is configured via actions.

You may want to use one action for "tag value" contains "Production" and a different action for "tag value" contains "Testing", with different escalation settings.

Keep in mind that with this method, the problem event still happens when the threshold is crossed and the trigger fires, the difference only happens in your actions, and it relates to when alerts start to be generated.

[mklatsky]

Thanks for this clarity- I *just* started looking into actions. That seems to be the consensus among folks I have chatted with.

[cyber]

Tag the hosts and make separate actions based on those tags.

[cyber]

That is completely different topic, compared to what you asked in first post...
For test and prod you probably should just create 2 different sets of templates with different values for triggers.