This message was deleted.
# harvestera
adamant-kite-43734
04/29/2026, 8:52 PMThis message was deleted.
b
bland-article-62755
04/29/2026, 10:56 PMThere's a scan interval that hits
bland-article-62755
04/29/2026, 10:57 PMThe
newClusteroldClustert
thousands-advantage-10804
04/29/2026, 10:59 PMit was 30 minutes and it never showed up.
h
happy-cat-90847
04/29/2026, 11:21 PMI've asked them to add a manual refresh button, Can you open an issue? It all cant come from me. 😉 Please share the issue number
t
thousands-advantage-10804
04/29/2026, 11:24 PMFor you, anything. I create one tonight.
thousands-advantage-10804
04/30/2026, 2:06 AMh
happy-cat-90847
04/30/2026, 2:08 AMI doubt we can go back in versions for a restore… let's see!
t
thousands-advantage-10804
04/30/2026, 2:13 AMIt worked once.. Just not a second time. so confused how to make it work.
thousands-advantage-10804
04/30/2026, 2:15 AMI have tried 3 different pcs, 3 different NVME and 2 versions and can’t get pci passthrough to work at all right now. SO BIZARRE. the only thread is that they are all AMD. IT was working for a year and 5 versions. Once upgraded to 1.8.0 the poo hit the fan.
f
few-ability-69534
04/30/2026, 5:55 AM1.8.0 has changed the naming of storageclasses
few-ability-69534
04/30/2026, 5:56 AMe.g. this is old names:
# kubectl get storageclass
NAME PROVISIONER RECLAIMPOLICY VOLUMEBINDINGMODE ALLOWVOLUMEEXPANSION AGE
harvester-longhorn (default) driver.longhorn.io Delete Immediate true 245d
longhorn driver.longhorn.io Delete Immediate true 245d
longhorn-image-9qpj6 driver.longhorn.io Delete Immediate true 135d
longhorn-image-cfftg driver.longhorn.io Delete Immediate true 218d
and new names on 1.8.0:
# kubectl get storageclasses.storage.k8s.io
NAME PROVISIONER RECLAIMPOLICY VOLUMEBINDINGMODE ALLOWVOLUMEEXPANSION AGE
harvester-longhorn (default) driver.longhorn.io Delete Immediate true 2d13h
lh-34d4c111-7c7a-4cd8-aab4-0e94119cb91a driver.longhorn.io Delete Immediate true 42h
lh-605f4b8d-ea93-4dde-8bcf-16fabd8da665 driver.longhorn.io Delete Immediate true 10h
lh-68cbed1a-4744-408e-ba92-dc611839a947 driver.longhorn.io Delete Immediate true 10h
few-ability-69534
04/30/2026, 5:56 AMwhat I did is created the storage classes matching old names pointing to same vmimages:
allowVolumeExpansion: true
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: longhorn-image-pw2m6
parameters:
backingImage: vmi-68cbed1a-4744-408e-ba92-dc611839a947
migratable: "true"
numberOfReplicas: "3"
staleReplicaTimeout: "30"
provisioner: driver.longhorn.io
reclaimPolicy: Delete
volumeBindingMode: Immediate
kubectl apply -f ...
🥂 1
few-ability-69534
04/30/2026, 5:57 AMand it worked after
few-ability-69534
04/30/2026, 5:57 AMhttps://github.com/harvester/harvester/issues/5165 - it's caused by this change
t
thousands-advantage-10804
04/30/2026, 5:15 PMthat did the trick thanks!
thousands-advantage-10804
04/30/2026, 5:45 PMAnd the restores won’t boot.
f
few-ability-69534
05/02/2026, 10:58 AMmine worked ok, all guest VMs are Ubuntu with ext4
t
thousands-advantage-10804
05/02/2026, 12:30 PMthis was user error. I did not have the exact image in place.
b
bland-article-62755
05/05/2026, 3:15 AMBeen burned by that before. I wish there was a way to flatten the image in the backup so it could restore to any storage class at the block level.
🐿️ 1
p
prehistoric-morning-49258
05/06/2026, 9:26 AMZFS does make some things easier 🫣
16 Views