Having what seems to be a DNS issue with the kubernetes/helm setup locally.
Using this helm chart: camunda-platform-helm/kind/camunda-platform-core-kind-values.yaml at 3080e08a594cbb32690cdd5e1761938b4207804c · camunda/camunda-platform-helm · GitHub
Created a cluster using kind, then ran helm install. Containers start up, but tasklist and operate are never able to connect to elasticsearch due to UnknownHostException.
I’ve tried:
- Completely removing kind, kubectl and helm, and reinstalled them
- Changed to k3d instead of kind
- Verified that I’m able to publish messages through the zeebe connector
- Restarted coredns in kube-system
- Checked that nslookup is able to resolve camunda-elasticsearch in both tasklist and operate containers
- Tried deploying a dns-utils container for checking the cluster dns and its also able to resolve the host
- Killed the tasklist/operate containers in case there is some dns caching causing issues.
After all of this, I came back the next day and it suddenly worked on the first try after leaving it for 15 minutes, but I have no idea why. I tried deleting and recreating the cluster, and it does not work again. Even if I leave it running for 30+ minutes, it does not fix itself.
I’m running Ubuntu 22.04, but I’ve also tried this on two other computers. Another 22.04 where it worked, and a vm on a windows pc which ran into the same issue as my pc.
Any ideas? 
Will gladly provide more details.