In previous versions (3.x) I used to do the following: (*_NEVER TESTED BY ME ON 4.X_*)
Let's say grop2 is your dead node, grop1 one of your running nodes. From one of the running nodes: - In /etc/pve look for config files of the dead node (es. grop2) root@grop1: $ cd /etc/pve/nodes/grop2/qemu-server/ - With ls you should see the list of vms running on the dead node - Move the vm .conf file to the node where you wish to run the vm es. to move vm 111 from grop2 to grop1: mv 111.conf ../../grop1/qemu-server/ Do the same for every vm running on the dead node, until grop2/qemu-server become empty. - Restart following services on all nodes: systemctl restart pvestatd.service systemctl restart pvedaemon.service systemctl restart pve-cluster.service - Issue "qm list" to verify that vms have been moved - Start vms: root@grop1: $ qm start 111 ecc. Bye N Il 07/11/2016 11:16, Szabolcs F. ha scritto: > Hello All, > > I've got a Proxmox VE 4.3 cluster (no subscription) of 12 Dell C6220 nodes. > > My question is: how do I move a VM from a dead node? Let's say my pve11 > dies (hardware issue), but the other 11 nodes are still up&running. In this > case I can't migrate VMs off of pve11, because I get the 'no route to host' > issue. I can only see the VM ID of the VMs that should be running on pve11. > But I want to move the VMs to the working nodes until I can fix the > hardware issue. > > All my VMs are stored on NAS servers, so a failing Proxmox node is not an > issue from this point of view, I can still access the VM files. All my 12 > PVE nodes access the storage with NFS. > > Thanks in advance! > _______________________________________________ > pve-user mailing list > pve-user@pve.proxmox.com > http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user > -- +---------------------+ | Linux User #554252 | +---------------------+ _______________________________________________ pve-user mailing list pve-user@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
