Hi all, quick question. I believe that the official method of changing IP address, if you have a ProxVE2.X cluster of 3 (or more) physical nodes:
-- backup VMs
-- reinstall nodes, use new IPs
-- restore VMs, then adjust IPs of VMs as desired:required.
(please correct me if wrong! :-)
However I also have found in the forums an 'unsupported' method:
http://forum.proxmox.com/threads/959...4508#post54508
which suggests it is possible to remove a cluster member from a cluster (ie, then change its config - for example IP address - and then re-join as member of a new cluster in this new IP range ..?). However I'm curious (a) if there is any experience with this (b) if it generally works tolerably for those doing it this non-standard way (c) if there are other unsupported ways that are known to work better, (d) icing on cake, is there a more supported way to do this kind of thing?
I ask because: I've got a client who wants a 3-node proxVE cluster deployed into pre-production now. But plans to have network topology change in ~3-5 months (new LAN - DMZ topology being setup - with a New firewall deployment etc) which would result in the IP range used for pre-deploy becoming obsolete; and requiring a change to the IPs of the ProxVE cluster nodes. From what I can tell the official fix for this is, backup your VMs, redeploy from scratch, and restore VMs from backup. But I'm hoping there might be a cleaner way to get this done.
Many thanks for any/all replies:comments.
Tim Chipman
-- backup VMs
-- reinstall nodes, use new IPs
-- restore VMs, then adjust IPs of VMs as desired:required.
(please correct me if wrong! :-)
However I also have found in the forums an 'unsupported' method:
http://forum.proxmox.com/threads/959...4508#post54508
which suggests it is possible to remove a cluster member from a cluster (ie, then change its config - for example IP address - and then re-join as member of a new cluster in this new IP range ..?). However I'm curious (a) if there is any experience with this (b) if it generally works tolerably for those doing it this non-standard way (c) if there are other unsupported ways that are known to work better, (d) icing on cake, is there a more supported way to do this kind of thing?
I ask because: I've got a client who wants a 3-node proxVE cluster deployed into pre-production now. But plans to have network topology change in ~3-5 months (new LAN - DMZ topology being setup - with a New firewall deployment etc) which would result in the IP range used for pre-deploy becoming obsolete; and requiring a change to the IPs of the ProxVE cluster nodes. From what I can tell the official fix for this is, backup your VMs, redeploy from scratch, and restore VMs from backup. But I'm hoping there might be a cleaner way to get this done.
Many thanks for any/all replies:comments.
Tim Chipman