Rapid updates of CoreDNS when running multiple vCluster in the same host namespace

[Fabian-K]

What happened?

When running multiple vclusters in the same host namespace, pods for coredns are rapidly started & terminated. This does not seem to happen when running only a single cluster in the host namespace.

What did you expect to happen?

Multiple vcluster in the same host namespace don´t interfere with each other.

How can we reproduce it (as minimally and precisely as possible)?

create 2 vclusters in the same host namespace

Anything else we need to know?

syncer logs:

syncer-logs.txt

Host cluster Kubernetes version

$ kubectl version
Client Version: v1.30.3
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.30.2

vcluster version

$ vcluster --version
vcluster version 0.22.0

VCluster Config

# My vcluster.yaml / values.yaml here
# -> no custom values.yaml