In place upgrade to 5.0
- Thread starter Drax
- Start date
So can I migrate all my VMs from Node one to Node TWO, upgrade Prox on node 1 in two steps, then move the VMs back to node one from node two
Or is it move VMs from Node one to Node two. Upgrade Node one to 4.1 Move the VMs back to node one. Allow what ever conversion is required, move Node 2 VMs to Node 1 then Upgrade Node 2 to 4.1 then move all VMs back to Node 2, then upgrade Node 1 to Prox 5, then move all VMs back to Node 1 then upgrade Node 2 to Prox 5 and divide the VMs to the nodes as originally configured.
Or is it move VMs from Node one to Node two. Upgrade Node one to 4.1 Move the VMs back to node one. Allow what ever conversion is required, move Node 2 VMs to Node 1 then Upgrade Node 2 to 4.1 then move all VMs back to Node 2, then upgrade Node 1 to Prox 5, then move all VMs back to Node 1 then upgrade Node 2 to Prox 5 and divide the VMs to the nodes as originally configured.
The instructions state I need to be atleast 3.4 but my version is 3.1-21 So my upgrade page is now
3.1 - 3.4
3.4 - 4.4
4.4 - 5.0
is that correct?
Also If I backup all the VMs does that mean I have backed up all the cluster data? I am becoming worried about my abilities to do this without error or issue. Is there any experienced Proxmox user you could suggest as a hired gun? Could I hire Proxmox to handle this in place upgrade?
3.1 - 3.4
3.4 - 4.4
4.4 - 5.0
is that correct?
Also If I backup all the VMs does that mean I have backed up all the cluster data? I am becoming worried about my abilities to do this without error or issue. Is there any experienced Proxmox user you could suggest as a hired gun? Could I hire Proxmox to handle this in place upgrade?
Last edited:
Can somebody please reply to the questions I have. I am not wanting to create more problems with this upgrade than I solve. Do the VM's have to be converted to each upgrade or can I simply back them all up to a drive then copy them back to the node once it is upgraded to 5.0? Are the VM's just copied to the data folder of the node?
The upgrade wiki pages describe the steps in detail - what do you miss exactly? And yes, we you can buy commercial support for this from the Proxmox VE team, see:
https://www.proxmox.com/en/proxmox-ve/pricing
this means you have to create a backup for each container (you should also test that those backups are valid and working)
but you also get support for all other issues you (might) have, and access to the enterprise repository
I guess I don't understand what is meant by each container. That is the part that is confusing to me. What is meant by container? Does that mean each VM? or does that mean the storage system for the Node? My hope was to back up the VMs to and external storage, then move the VMs from node one to Node two. Upgrade node one, then move the VMs back to node one from node two. Can I not do that? Do I have to restore the VMs to the upgraded node one from a backup versus moving them from Node two?
I agree for a bigger business the subscription would be good. BUT for us it would be several thousand a year. For us in the 8 years we have ran Proxmox, we have not had any real issues other than the upgrade cycle. It would be great if the in place upgrade could be such that one could do it without risk to the data and VMs.
Maybe somebody here that is experienced might come forward and contact us about being a mentor or consultant in this matter. I am confident I can do the upgrade, I would just want somebody available just in case.
I agree for a bigger business the subscription would be good. BUT for us it would be several thousand a year. For us in the 8 years we have ran Proxmox, we have not had any real issues other than the upgrade cycle. It would be great if the in place upgrade could be such that one could do it without risk to the data and VMs.
Maybe somebody here that is experienced might come forward and contact us about being a mentor or consultant in this matter. I am confident I can do the upgrade, I would just want somebody available just in case.
in proxmox there are two types of guests: vms (with qemu/kvm; full-virtualization) and containers (up to 3.4 with openvz, since 4.0 with lxc)
since we changed with 4.0 from openvz to lxc, and the cluster stack changed, the upgrade is a bit more involved