-
Feature Request
-
Resolution: Done
-
Undefined
-
None
-
None
-
None
-
False
-
False
-
Undefined
-
As an operator author I want more control over how my operator health is reported to the administrator. Today my operator health is reported entirely based on:
1) installplan success (which just means operator resources were created)
2) operator deployment rollout success(? not sure if this is used or where)
Some operator teams want to control their operator health more specifically, for example some operators are "singletons" which means that installing the operator is equivalent to instantiating the operand, and they want their operator health to reflect whether or not that singleton operand has been created successfully and is healthy.
Rather than relying on the state of resources the operator itself cannot control (installplan, operator deployment), operator teams should be given an api they can use to report their operator status/health explicitly. If they do not use that api, we can continue to fallback to the existing mechanisms.
open questions:
1) exactly which status/health reporting are teams referring to? which information on the console?
2) where does that information derive from today (just installplan, or also operator deployment status?)
Implementation Details:
Can be done by extending Operator Conditions.
- is related to
-
OPRUN-2364 Top-level OLM metrics (and alerts) for over-all operator health and operator upgrade status
- New