Hello,
I'm in the process of moving VMs from one Proxmox cluster to another, using vzdump and qmrestore. The older cluster is using LVM for the VM storage and the newer cluster is using Sheepdog.
When running a qmrestore, the process does not actually load the data into sheepdog. Instead, it writes it to a local file, as an example:
Just to be clear, the name of our sheepdog storage is actually sheepdog.
As a workaround, I used the `collie vdi write` command to actually get the data into sheepdog:
Does qmrestore know nothing of sheepdog, or is this a bug in the code? Any ideas would be greatly appreciated as I have several more VMs to migrate.
I'm in the process of moving VMs from one Proxmox cluster to another, using vzdump and qmrestore. The older cluster is using LVM for the VM storage and the newer cluster is using Sheepdog.
When running a qmrestore, the process does not actually load the data into sheepdog. Instead, it writes it to a local file, as an example:
Code:
root@proxmoxe:~# qmrestore /tmp/vzdump107.tar.lzo 300 -storage sheepdog
extracting archive '/tmp/vzdump107.tar.lzo'
extracting 'qemu-server.conf' from archive
extracting 'vm-disk-virtio0.raw' from archive
new volume ID is 'sheepdog:vm-300-disk-1'
restore data to 'sheepdog:127.0.0.1:7000:vm-300-disk-1' (42949672960 bytes)
42949672960 bytes copied, 308 s, 132.99 MiB/s
root@proxmoxe:~# ls -lh
-rw-r--r-- 1 root root 40G Aug 21 15:55 sheepdog:127.0.0.1:7000:vm-300-disk-1Just to be clear, the name of our sheepdog storage is actually sheepdog.
As a workaround, I used the `collie vdi write` command to actually get the data into sheepdog:
Code:
root@proxmoxe:~# dd if=sheepdog:127.0.0.1:7000:vm-300-disk-1 | collie vdi write vm-300-disk-1
83886080+0 records in
83886080+0 records out
42949672960 bytes (43 GB) copied, 1763.21 s, 24.4 MB/sDoes qmrestore know nothing of sheepdog, or is this a bug in the code? Any ideas would be greatly appreciated as I have several more VMs to migrate.