[Feature]: Provide the ability to override recommendations, fatality of health events #462
Description
Prerequisites
- I searched existing issues
Feature Summary
Currently the health monitors publish health events with all the required information populated which works great for majority of use cases. In some cases, operators may need to override some properties of health event such as isFatal, isHealthy, recommendedAction and so on. Currently it is possible to disable quarantine using CEL, but the health events show up in the Kubernetes node object and events, and this gets confusing to the end user.
For example, in some environments operators may have disabled remediation for an XID, but the XID will show up on the node object confusing end users who think that something is wrong with the node.
Problem/Use Case
As an operator I want to override properties of some health events based on my environments without having to fork the health monitoring components.
Proposed Solution
We have 2 options:
- Make every health monitor expose some sort of override options
- Override the health events in the platform connector
The second option is a bit more scalable as it will work for any health monitor both existing and new.
Component
Health Monitor