proxmox 8.3.0 - qm remote migrate issue check storage plugin support!

K

kellogs

Active Member
May 14, 2024
205
32
28
source ceph target nfs


root@compute-16-114:~# qm remote-migrate 103 2005 'apitoken=PVEAPIToken=root@pam!PVEAPIToken=xxxxxxxxxxxxxxxxxxxxxxxxx
,host=172.16.16.200,fingerprint=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
' --target-bridge vlan400 --target-storage tintri1638 --online
Establishing API connection with remote at '172.16.16.200'
VM isn't running. Doing offline migration instead.
2025-02-02 06:15:45 remote: started tunnel worker 'UPID:compute-16-200:000A12DB:00B05DFB:679E9D11:qmtunnel:2005:root@pam!PVEAPIToken:'
tunnel: -> sending command "version" to remote
tunnel: <- got reply
2025-02-02 06:15:45 local WS tunnel version: 2
2025-02-02 06:15:45 remote WS tunnel version: 2
2025-02-02 06:15:45 minimum required WS tunnel version: 2
websocket tunnel started
2025-02-02 06:15:45 starting migration of VM 103 to node 'compute-16-200' (172.16.16.200)
tunnel: -> sending command "bwlimit" to remote
tunnel: <- got reply
2025-02-02 06:15:45 found local disk 'cephNVME:vm-103-disk-0' (attached)
2025-02-02 06:15:45 copying local disk images
2025-02-02 06:15:45 ERROR: no export formats for 'cephNVME:vm-103-disk-0' - check storage plugin support!
2025-02-02 06:15:45 aborting phase 1 - cleanup resources
tunnel: -> sending command "quit" to remote
tunnel: <- got reply
2025-02-02 06:15:46 ERROR: migration aborted (duration 00:00:01): no export formats for 'cephNVME:vm-103-disk-0' - check storage plugin support!
migration aborted
 
I ran into the same problem last week and the problem was not only with Ceph, also ZFS, LVM and directory had problems. I could not get any remote migration to work, which was a big bummer. The strangest thing was ZFS on the source. It first claimed, that it can only run with a snapshot and after creating a snapshot it prohibited the work because there is a snapshot present. So this feature deserves it "experimental" flag.
 
VM isn't running. Doing offline migration instead.
Click to expand...
From log it is seen, that the vm is not running, when migration was started. As far as I understood the current limitations the vm must be running at the time of migration.
 
  • Like
Reactions: kellogs
hello @fba you are correct! not just adding the flag online but the VM needs to be online too! It works great now :) Thank you for the tips.
 
Hi,
do you have libpve-storage-perl >= 8.3.3 installed on the node with Ceph? What kind of storage is the target storage?
 
kellogs said:
Ceph cluster is using libpve-storage-perl: 8.2.9 and the target storage is NFS 4.1
Click to expand...
That version doesn't support the offline export feature for RBD/Ceph yet.
 
You must log in or register to reply here.
Top Bottom