Changing IP Addresses of Nodes in a Proxmox Cluster

Introduction

Changing the IP addresses of a Proxmox cluster can seem daunting, especially when the cluster is already live. Recently, I needed to migrate my cluster from the 192.168.1.0/24 subnet to 10.0.10.0/24. This post documents the steps I followed to update my three-node cluster without breaking it.

Planning the Migration

Before diving in, it’s essential to prepare:

1. Backup Configurations: Always back up configuration files in case you need to roll back changes.
2. Check Cluster Health: Use pvecm status to verify that the cluster is healthy and operating normally.
3. Define New IPs:
   

   10.0.10.1        - Gateway
   255.255.255.0    - Netmask
   10.0.10.51       - TATOOINE (The main node)
   10.0.10.52       - VERUNA (second node)
   10.0.10.53       - NABOO (third node)
   

4. Order of Updates: Plan which node to update first to minimize disruptions.

Making the Changes

Here are the key files to edit, along with the commands to directly open these:

nano /etc/network/interfaces - On every node!

Update the network configuration for each node. For example, on TATOOINE:

auto eth0
iface eth0 inet static
    address 10.0.10.51/24
    gateway 10.0.10.1

Note that if you append the CIDR /24 to the address, you don’t need to supply the Subnet mask in your config

nano /etc/hosts - On every node!

Update the host entries to reflect the new IP addresses:

10.0.10.51 TATOOINE.mydomain TATOOINE
10.0.10.52 VERUNA.mydomain VERUNA
10.0.10.53 NABOO.mydomain NABOO

nano /etc/pve/corosync.conf

Modify the ring0_addr values in Corosync configuration:

nodelist {
    node {
        name: TATOOINE
        nodeid: 1
        quorum_votes: 1
        ring0_addr: 10.0.10.51
    }
    node {
        name: VERUNA
        nodeid: 2
        quorum_votes: 1
        ring0_addr: 10.0.10.52
    }
    node {
        name: NABOO
        nodeid: 3
        quorum_votes: 1
        ring0_addr: 10.0.10.53
    }
}

nano /etc/resolv.conf - On every node!

Update /etc/resolv.conf for DNS settings if necessary.

search mydomain
nameserver 10.0.10.1

nano /etc/issue - On every node!

Edit /etc/issue to reflect the new network details for banner consistency.

------------------------------------------------------------------------------

Welcome to the Proxmox Virtual Environment. Please use your web browser to
configure this server - connect to:

  https://10.0.10.51:8006/

------------------------------------------------------------------------------

Applying the Changes

After making the changes, restart relevant services:

systemctl restart networking
systemctl restart corosync
systemctl restart pve-cluster
systemctl restart pvebanner

Reboot the nodes if needed.

Testing and Verification

Once the changes are applied, verify the setup:

1. Network Connectivity: Use ping to ensure the nodes can reach each other.

    ping TATOOINE
    ping VERUNA
    ping NABOO
   

2. Cluster Health: Run:

    pvecm status
   

   Check that all nodes are in the cluster and communicating.

3. Logs: Review logs to catch any potential issues:

    journalctl -xe
   

Final Thoughts

Migrating a Proxmox cluster to a new subnet can be done without significant downtime if approached methodically. The new network has improved scalability and organization in my setup.

What challenges have you faced with Proxmox networking? Let me know in the comments!