Is it possible to change IP address of nodes in a cluster ?

[galphanet]

Hello,

I have 6 servers in a cluster and I need to change the subnet of these servers.

Is is possible without breaking all the config of the cluster ?
Is there a particular way to do it ?

Bonus question : is it possible to have vmbr interfaces without addresses ?

Thanks you for your help.

 

[dietmar]

Is is possible without breaking all the config of the cluster ?
Is there a particular way to do it ?

The cluster config does not store ip/subnet addresses. What exactly do you mean by 'change the subnet'? You only want to chnage the netmask?

Bonus question : is it possible to have vmbr interfaces without addresses ?

Sure

 

[galphanet]

The cluster config does not store ip/subnet addresses. What exactly do you mean by 'change the subnet'? You only want to chnage the netmask?

Thanks you for your fast answer.

I have nodes with IP addresses like 10.10.10.2XX and I have to change them to 10.20.20.1XX

 

[dietmar]

I have nodes with IP addresses like 10.10.10.2XX and I have to change them to 10.20.20.1XX

If you change the IP address you need to edit /etc/network/interfaces and /etc/hosts (and restart those nodes).

 

[galphanet]

Ok so I can change them and multicast will take care of the rest ?

If so, this is not possible to have nodes in two differents subnets, right ?

 

[dietmar]

Ok so I can change them and multicast will take care of the rest ?

AFAIK yes, but I never tested that.

If so, this is not possible to have nodes in two differents subnets, right ?

No, that is not possible.

 

[sysopsprime]

Hi,
I would like to change the IPs of all nodes in a cluster.
current: 10.0.23.xx netmask 255.0.0.0
required: 10.249.0.xx netmask 255.255.255.0

I tried changing it by changing the ip in three different places,
/etc/network/interfaces
/etc/hosts
/etc/pve/corosync.conf
but once i restarted the node it became red and i was not able to change the corosync.conf at all.
could you please help me out.

 

[wbumiller]

The corosync service uses /etc/corosync/corosync.conf, you need to make sure the same changes are applied there. This file is usually updated by the cluster service when the corosync.conf in /etc/pve gets updated by one of the nodes, but if you make manual changes, you need to make sure they're both up to date, and once a node is out of the cluster it won't get the required updates anymore.

You can stop pve-cluster.service and corosync.service, run `pmxcfs -l` as root to get /etc/pve running in local mode, then copy the file from there, then stop pmxcfs again and restart the two services.