I have a question about how storage works on Proxmox, and I'm not sure if I've set up my storage incorrectly. I am coming from an ESXi environment where there are compute-only migrations and am looking to emulate that behavior in Proxmox.
I have 8 hosts, each with a fiber nic connected to a switch with 2 paths to the SAN. I set them up in proxmox and connected a volume from my SAN to each host; it's the same volume presented to all the hosts.
I then clustered the hosts; when I click my datacenter/cluster in proxmox and go to the storage tab, I see the iscsi targets listed. I created an LVM volume and it shows up perfectly.
I then created a VM. I attempted a migration, and it worked, however it copied the disk. This is what the migration task log shows:
So it's copying the disk to the same location. Did I set this up wrong? Or is this how Proxmox migrations should work, and there is no equivalent concept of compute-only migration?
I have 8 hosts, each with a fiber nic connected to a switch with 2 paths to the SAN. I set them up in proxmox and connected a volume from my SAN to each host; it's the same volume presented to all the hosts.
I then clustered the hosts; when I click my datacenter/cluster in proxmox and go to the storage tab, I see the iscsi targets listed. I created an LVM volume and it shows up perfectly.
I then created a VM. I attempted a migration, and it worked, however it copied the disk. This is what the migration task log shows:
Code:
2021-09-28 09:40:05 use dedicated network address for sending migration traffic (10.1.###.##5)
2021-09-28 09:40:05 starting migration of VM 100 to node 'PROX-05' (10.1.###.##5)
2021-09-28 09:40:06 found local disk 'DS-PROX-Servers:vm-100-disk-1' (in current VM config)
2021-09-28 09:40:06 starting VM 100 on remote node 'PROX-05'
2021-09-28 09:40:09 volume 'DS-PROX-Servers:vm-100-disk-1' is 'DS-PROX-Servers:vm-100-disk-0' on the target
2021-09-28 09:40:09 start remote tunnel
2021-09-28 09:40:10 ssh tunnel ver 1
2021-09-28 09:40:10 starting storage migration
2021-09-28 09:40:10 scsi0: start migration to nbd:unix:/run/qemu-server/100_nbd.migrate:exportname=drive-scsi0
drive mirror is starting for drive-scsi0
drive-scsi0: transferred 0.0 B of 80.0 GiB (0.00%) in 0s
drive-scsi0: transferred 104.0 MiB of 80.0 GiB (0.13%) in 1s
.....
drive-scsi0: transferred 80.0 GiB of 80.0 GiB (100.00%) in 12m 23s, ready
all 'mirror' jobs are ready
2021-09-28 09:52:33 starting online/live migration on unix:/run/qemu-server/100.migrate
2021-09-28 09:52:33 set migration capabilities
2021-09-28 09:52:33 migration downtime limit: 100 ms
2021-09-28 09:52:33 migration cachesize: 512.0 MiB
2021-09-28 09:52:33 set migration parameters
2021-09-28 09:52:33 start migrate command to unix:/run/qemu-server/100.migrate
2021-09-28 09:52:34 migration active, transferred 107.3 MiB of 4.0 GiB VM-state, 114.8 MiB/s
...
2021-09-28 09:52:46 migration active, transferred 1.3 GiB of 4.0 GiB VM-state, 119.1 MiB/s
2021-09-28 09:52:46 average migration speed: 316.4 MiB/s - downtime 173 ms
2021-09-28 09:52:46 migration status: completed
all 'mirror' jobs are ready
drive-scsi0: Completing block job_id...
drive-scsi0: Completed successfully.
drive-scsi0: mirror-job finished
2021-09-28 09:52:47 stopping NBD storage migration server on target.
Logical volume "vm-100-disk-1" successfully removed
2021-09-28 09:52:56 migration finished successfully (duration 00:12:51)
TASK OKSo it's copying the disk to the same location. Did I set this up wrong? Or is this how Proxmox migrations should work, and there is no equivalent concept of compute-only migration?
