This message was deleted.
# harvestera
adamant-kite-43734
03/04/2025, 3:59 PMThis message was deleted.
e
enough-australia-5601
03/04/2025, 4:12 PMThe built-in migration feature in Harvester is meant to migrate VMs between nodes within a cluster. The main purpose of this is to be able to do node maintenance or upgrades without necessarily disrupting workloads, though in reality it can prove to be quite tricky, especially when the cluster has heterogeneous hardware and the workloads are under high utilization.
The best alternative I can think of is to backup the VM into an external backup target and then restore it in the other cluster: https://docs.harvesterhci.io/v1.4/vm/backup-restore#restore-a-new-vm-on-another-harvester-cluster
b
bland-article-62755
03/04/2025, 4:14 PMYeah, when you use a third party storage driver, which are supported, like the
<http://rbd.ceph.io|rbd.ceph.io>bland-article-62755
03/04/2025, 4:15 PMSo the backup/restore can't work from the docs you provided. (I'm bummed)
bland-article-62755
03/04/2025, 4:17 PMEven so, apparently that feature is locked to longhorn backed PVs only.
e
enough-australia-5601
03/04/2025, 4:22 PMThe UI crash sounds like a bug. Would you mind reporting it as an issue on https://github.com/harvester/harvester/issues ?
The backup/restore feature is tightly integrated with Longhorn. Not much can be done about that afaik.
b
bland-article-62755
03/04/2025, 4:25 PMThere's an existing issue, I can look it up. I'm not really hopeful for something to make the integrated stuff work, but rather for some kubectl hack that exports the disks somehow and allows them to get imported into a new cluster.
e
enough-australia-5601
03/04/2025, 5:39 PMThanks.
Unfortunately, I don't think the backup-and-restore route is viable until https://github.com/harvester/harvester/issues/5816 is fixed.
b
bland-article-62755
03/04/2025, 5:46 PMyeah
bland-article-62755
03/04/2025, 5:48 PMI see that
virtctl