[Bug]: new node already at cluster `Node is already a Storage Manager for different Cluster`

[TristisOris]

Describe the bug any attempts to create cluster not succeed.

root@kadalu-p01-n01:~# kadalu node add kadalu-p01-n02
Failed to add the node
Node is already a Storage Manager for different Cluster

root@kadalu-p01-n01:~# kadalu node add kadalu-p01-n03
Failed to add the node
Node is already a Storage Manager for different Cluster

root@kadalu-p01-n01:~# kadalu node add kadalu-p01-n01
Failed to add the node
Node is already part of the cluster

kadalu pool create Test-Pool replica 3 192.168.1.1:/opt/s3 192.168.1.2:/opt/s3 192.168.1.3:/opt/s3 --auto-add-nodes
Failed to create pool
{"error":"Node is already a Storage Manager for different Cluster"}

kadalu pool create Test-Pool replica 3 192.168.1.1:/opt/s3 192.168.1.2:/opt/s3 192.168.1.3:/opt/s3
Failed to create pool
Node 192.168.1.2 is not part of the Cluster

kadalu node add 192.168.1.2
Failed to add the node
Node is already a Storage Manager for different Cluster

same if use ns names (resolve via /etc/hosts conf)

To Reproduce Steps to reproduce the behavior:

1. done installation according to https://kadalu.tech/storage/quick-start/
2. right after service starting, try to create cluster.

Actual behavior here https://docs.kadalu.tech/kadalu-storage/latest/setup/ it say No nodes are added to the cluster by default.. info config from nodes: {"pool_name":"","id":"0db99011-4883-494e-a6f5-fafc9c103530","name":"","token_hash":"","mgr_hostname":"","mgr_port":3000,"mgr_https":false,"mgr_token":"","joined":false} So nodes really not at pool.

Environment: ubuntu 22 Server Version : 1.2.0 CLI Version : 1.2.0

[aravindavk]

Are these nodes used earlier? You can remove /var/lib/kadalu in each node and try again.

[aravindavk]

@TristisOris I moved this bug to GitHub/kadalu/moana repo. Are you looking for Gluster alternative or curious about Kadalu Storage?

[TristisOris]

no, all 3 nodes are fresh created vms.

Not familiar with gluster too, but found that k8s drop it from support.

[TristisOris]

• removed /var/lib/kadalu
• restart service
• again created admin account
• login into account

  
  root@kadalu-p01-n01:~# kadalu node add kadalu-p01-n01
  Node kadalu-p01-n01 added to the cluster
  ID: 32fffd6e-abe3-4377-a63e-3ad3308645da

root@kadalu-p01-n01:~# kadalu node add kadalu-p01-n02 Failed to add the node Node is already a Storage Manager for different Cluster

[aravindavk]

Thanks for confirming. I will check it tomorrow and update here.

[aravindavk]

no, all 3 nodes are fresh created vms.

Not familiar with gluster too, but found that k8s drop it from support.

Thanks. This project is not created to use with k8s. Please use https://github.com/kadalu/kadalu project if you are looking for k8s storage solution.

Moana is for on-premise use cases without using k8s. Both these projects are based on core GlusterFS.

[TristisOris]

ah, got it. Also we need some S3 not only for k8s, so i'll try both.