I'm exploring methods to provide redundancy/failover for VMs running on a PVE cluster. I am wondering if, using the Proxmox Replication functionality I can:
1. replicate a VM on Node1 to another node (say Node2) in the cluster at a certain time point T,
2. automatically or manually start the replicated VM (from time point T) on node 2 if node 1 becomes unresponsive/goes down
3. continue running a job on the VM (now on node2) that was running when the VM was replicated at time point
To elaborate on 3., say I have a simple bash script running on the VM on node1 that simply echos an interger to the terminal: 1, 2, 3, 4, 5....
I replicate that VM to node 2 at time point T.
Node 1 goes down at time point T+3 when the interger echoed is "8"
The VM starts on node2 and the bash script echos to the terminal: 6, 7, 8, 9, ......etc i.e. effectively rolling back in time (to time point T) but thats ok.
1. replicate a VM on Node1 to another node (say Node2) in the cluster at a certain time point T,
2. automatically or manually start the replicated VM (from time point T) on node 2 if node 1 becomes unresponsive/goes down
3. continue running a job on the VM (now on node2) that was running when the VM was replicated at time point
To elaborate on 3., say I have a simple bash script running on the VM on node1 that simply echos an interger to the terminal: 1, 2, 3, 4, 5....
I replicate that VM to node 2 at time point T.
Node 1 goes down at time point T+3 when the interger echoed is "8"
The VM starts on node2 and the bash script echos to the terminal: 6, 7, 8, 9, ......etc i.e. effectively rolling back in time (to time point T) but thats ok.
Last edited:
