Situation: I had a hard drive fail on a single node (in a 10+ node cluster). The VMs on the node have been backed up to PBS. I can access the VM configs of the failed node, from other working nodes,= (e.g. /etc/pve/nodes/HOSTNAME_OF_FAILED_NODE/qemu-server/149.conf.
After digging through other forum posts, I'm hoping someone can help fill in some gaps.
0. Backup PVE config under (/etc/pve/nodes/HOSTNAME_OF_FAILED_NODE/) from a functional node.
1. Install new harddrive and install proxmox.
2. Delete old failed node from proxmox cluster. Then re-add the freshly built node to the cluster.
3. Restore PVE config (everything backed up in first step). Is it necessary to use `qm rescan` at this point (as mentioned here: https://forum.proxmox.com/threads/boot-disk-failure-recovery.127071/)?
4. Restore VM volumes from PBS.
To be honest, I don't even mind manually recreating the VMs and restoring the VM data from PBS.
When searching the forums, I also saw references to others manually backing up everything in /etc/pve or using the proxmox backup client. Is that really necessary in cases like this, where I just want to recover a handful of VMs on a single failed node? It seems like that would only be necessary if I had a full cluster failure / cluster-wide configuration issue...?
After digging through other forum posts, I'm hoping someone can help fill in some gaps.
0. Backup PVE config under (/etc/pve/nodes/HOSTNAME_OF_FAILED_NODE/) from a functional node.
1. Install new harddrive and install proxmox.
2. Delete old failed node from proxmox cluster. Then re-add the freshly built node to the cluster.
3. Restore PVE config (everything backed up in first step). Is it necessary to use `qm rescan` at this point (as mentioned here: https://forum.proxmox.com/threads/boot-disk-failure-recovery.127071/)?
4. Restore VM volumes from PBS.
To be honest, I don't even mind manually recreating the VMs and restoring the VM data from PBS.
When searching the forums, I also saw references to others manually backing up everything in /etc/pve or using the proxmox backup client. Is that really necessary in cases like this, where I just want to recover a handful of VMs on a single failed node? It seems like that would only be necessary if I had a full cluster failure / cluster-wide configuration issue...?
