Currently, in RHOSO, the user can configure alarms into aodh and that will query Prometheus every minute checking if the configured metrics go over the configured threshold.

If that goes above the threshold, aodh will trigger a webhook on Heat to start an autoscaling operator, so the alarming is currently tied very specifically to autoscaling.

The user could also configure alarms on alertmanager, but we might want something more OSP integrated, and thats why this epic exists.

The first question to answer is if we want to re-enable events to get better and quicker alarms, as well as expand aodh capabilities so the user can configure multiple alarms and they will always have the tenancy incorporated.

Also, we need to build some system to show the alarms to the user.