Skip to content

[Bug] Kubernetes Discovery does not "lose" targets when scaled down #1048

@andrewazores

Description

@andrewazores

Current Behavior

With Cryostat deployed in Kubernetes, and Kubernetes Endpoints discovery enabled,

$ oc scale -n apps1 deployment/quarkus-test --replicas=1
$ oc scale -n apps1 deployment/quarkus-test --replicas=2
$ oc scale -n apps1 deployment/quarkus-test --replicas=1

with some wait time in between scaling results in Cryostat retaining two discovered targets, both in the Topology UI view as well as ex. GET /api/v4/targets.

Expected Behavior

The Kubernetes discovery informer should catch an event that the target was lost and successfully remove the target and its discovery node chain.

Steps To Reproduce

No response

Environment

- OS:
- Environment:
- Version:

Anything else?

No response

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    Projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions