User or Developer story:

As a user, I want metrics to be more reactive than the configured period, when possible. For example, cluster health metrics could be updated on the according datastore calls.

Note, that central exposes some health metrics on the internal endpoint, which might suit better.

Acceptance Criteria:

A list of specific needs or objectives that this story must deliver in order to be considered complete. Complete during Refinement status.

Not in Scope (Optional):

List of items that are out of scope . Complete during Refinement status.

Engineering Details (Optional):

In the 4.9 implementation, the "health" metrics are collected once an hour. This might be considered unfortunate that an alert on a status health metric will fire for at least one hour even if the cluster was unhealthy only for a couple of minutes. Instead, datastores can request a premature metric update on some events like cluster status or policy update. This would be a nice improvement, especially for the "cheap" metrics, which don't take much resources to collect.