-
Notifications
You must be signed in to change notification settings - Fork 594
Decommission single broker node in Kubernetes with 3 node redpanda cluster #11743
Replies: 2 comments · 3 replies
-
hey @mdfaizsiddiqui welcome to the community. i shared this internally on our slack. we'll try to have the k8s team chime in cc: @dswang @StasiaZam |
Beta Was this translation helpful? Give feedback.
All reactions
-
This is a misleading message. It just means that the decommission command has been submitted to the controller. This was fixed recently.
Can you share the output of |
Beta Was this translation helpful? Give feedback.
All reactions
-
Output for
Logs -
|
Beta Was this translation helpful? Give feedback.
All reactions
-
Its trying to reallocate one partition from the dead/decommissioned node, can you share the output of ..
and
|
Beta Was this translation helpful? Give feedback.
All reactions
-
Output -
|
Beta Was this translation helpful? Give feedback.
-
Hey
I'm running Redpanda on kubernetes cluster (Server Version: v1.27.2-eks-c12679a) with local-storage persistent volumes.
I am following this guide Decommission Brokers in Kubernetes to decommission a broker node/pod on a 3 nodes freshly built Redpanda cluster.
First I've set broker id
2
for decommission -But when I see the cluster health, the node 2 is still part of the "All nodes" field here -
Logs from the broker pods -
Can I know why it is stuck in draining state and not proceeding with decommissioning? Is there anything which is stopping it? Let me know if you need more information from me, I'll appreciate your help on this.
Thanks
Beta Was this translation helpful? Give feedback.
All reactions