i have problem replace host
  

Newbie917245 Lv1Posted 01 Oct 2024 15:47

i have problem replace host

setting cluster 3 node
node1 ,i reinstall os (disk virtual storage is ok)

when i try replace host , error say disk same size and performance (yes cause disk storage is ok , i just reinstall os)
how to fix it?

9357366fba8bb12281.png
Sheikh_Shani Lv2Posted 01 Oct 2024 18:40
  
Hello Dear

Replacing a host in a 3-node cluster can be challenging.


To resolve the issue:


Method 1: Manual Configuration

1. Ensure the new node (Node1) has the same:
    - Disk configuration (same size, type, and layout)
    - Network configuration (IP address, subnet mask, gateway)
    - Storage configuration (same virtual storage setup)
2. On the existing cluster nodes (Node2 and Node3):
    - Update the cluster configuration to reflect the new node's information
    - Run the following command: cluster update node <new_node_name> <new_node_IP>
3. On the new node (Node1):
    - Run the following command: cluster join <cluster_name> <existing_node_IP>


Method 2: Force Replace

1. On the existing cluster nodes (Node2 and Node3):
    - Stop the cluster service: service cluster stop
    - Run the following command: cluster replace node <new_node_name> <new_node_IP> --force
2. On the new node (Node1):
    - Run the following command: cluster join <cluster_name> <existing_node_IP>


Method 3: Modify Cluster Configuration File

1. On the existing cluster nodes (Node2 and Node3):
    - Edit the cluster configuration file (/etc/cluster/cluster.conf or similar)
    - Update the node information for Node1
    - Restart the cluster service: service cluster restart


Additional Steps

1. Verify disk alignment and configuration on Node1.
2. Check for any firmware or driver updates for storage controllers.
3. Ensure Node1 has the same OS version and patches as Node2 and Node3.


Common Errors

1. Disk size and performance mismatch
2. Network configuration mismatch
3. Storage configuration mismatch
4. Cluster configuration file errors


Troubleshooting Resources

1. Cluster documentation for your specific distribution (e.g., Red Hat, CentOS)
2. Cluster management software documentation (e.g., Pacemaker, Corosync)
3. Online forums and communities (e.g., Reddit, Stack Overflow)

Enrico Vanzetto Lv4Posted 02 Oct 2024 00:01
  
Hi, unfortunately the alert on your ssd means that it's not supported properly on HCI. I had same issue time ago after installing HCI on a third part server (HPE on my case). I solve by open a ticket to Sangfor technical support. On a remote session, they connect on involved HCI via ssh and made some setting to make ssd fully compatible with HCI.