VMs HA problems with last update with kernel 6.14.11-2

[gianlucagiacometti]

Hello,
I'm experiencing a problem in migration of VMs after last update today (kernel-6.14.11-2).
All hosts, regardless of the hardware, hang when I reboot. HA cannot stop VMs for migration and the taskbar switches from "stop VM" to "Error migrating" continuously. The problem doesn't affect CTs. My solution was to hard reboot the hosts.
If it may help, with the same upgrade qemu-server goes from 9.0.19 to 9.0.21
Anyone had the same problem?

 

[dakralex]

Hi!

What is the config of such a VM that fails to migrate (output of qm config <vmid>)? What does the tasklog show if the VM is migrated through HA? What does it show if the VM is migrated manually (does it work that way?)?

 

[gianlucagiacometti]

Example of qm config

agent: 1,fstrim_cloned_disks=1
boot: order=scsi0;ide2;net0
cores: 8
ide2: none,media=cdrom
memory: 49152
name: jupiter
net0: virtio=16:70:1A:40:86:65,bridge=vmbr0,firewall=1
numa: 0
onboot: 1
ostype: l26
scsi0: primary-rbd:vm-102-disk-0,size=256G
scsi1: primary-rbd:vm-102-disk-1,size=4T
scsihw: virtio-scsi-pci
smbios1: uuid=862109f9-bc83-4446-ab3b-cf48b5dd6c38
sockets: 2
spice_enhancements: videostreaming=all
vga: qxl
vmgenid: 4f1b8b60-4227-4a8c-9617-15367e4aca21

tasklog either shows

no content

or

task started by HA resource agent
2025-09-23 10:27:11 conntrack state migration not supported or disabled, active connections might get dropped
2025-09-23 10:27:12 use dedicated network address for sending migration traffic (10.72.11.7)
2025-09-23 10:27:12 starting migration of VM 209 to node 'host07' (10.72.11.7)
2025-09-23 10:27:12 starting VM 209 on remote node 'host07'
2025-09-23 10:27:13 [host07] Unknown option: nets-host-mtu
2025-09-23 10:27:13 [host07] 400 unable to parse option
2025-09-23 10:27:13 [host07] qm start <vmid> [OPTIONS]
2025-09-23 10:27:13 ERROR: online migrate failure - target node host07 is too old for preserving VirtIO-net MTU, please upgrade
2025-09-23 10:27:13 aborting phase 2 - cleanup resources
2025-09-23 10:27:13 migrate_cancel
2025-09-23 10:27:14 ERROR: migration finished with problems (duration 00:00:03)
TASK ERROR: migration problems

I guess I found the problem:

"Unknown option: nets-host-mtu"
"ERROR: online migrate failure - target node host07 is too old for preserving VirtIO-net MTU, please upgrade"

Is it something that should have been done before the upgrade or there is a problem with one of the updated packets?

Thank you

 

[gianlucagiacometti]

I guess it tries to migrate into a still-not-upgraded host (I have a 17-host cluster)

 

[gianlucagiacometti]

Among other upgrades there's two of them related to network: libpve-network-api-perl libpve-network-perl

And I confirm that the problem is related only to those VMs migrating to host not yet upgraded

I ended up upgrading all hosts before rebooting them one by one. This way I still have an error related to network, but the migration does not prevent the host from rebooting:

2025-09-23 13:47:44 ERROR: Cleanup after stopping VM failed - org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.