Created attachment 1680557 [details]
engine log

Description of problem:

When invoking full backup for a disk which is currently being migrated to another SD, the process is attempting to be started although the disk is locked.
This is causing the engine to throw Exception:

2020-04-21 16:03:05,859+03 ERROR [org.ovirt.engine.core.bll.StartVmBackupCommand] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-4) [e9340d04-5937-4dc0-a6c9-e9ed9ab2d09c] Failed to execute VM backup operation 'StartVmBackup': {}: org.ovirt.engine.core.common.errors.EngineException: EngineException: org.ovirt.engine.core.vdsbroker.vdsbroker.VDSErrorException: VDSGenericException: VDSErrorException: Failed to StartVmBackupVDS, error = Backup Error: {'vm_id': 'fefd90c8-94a5-4c68-83ae-9c3462655ca9', 'backup': <vdsm.virt.backup.BackupConfig object at 0x7fb9b420d550>, 'reason': "Failed to find one of the backup disks: No such drive: '{'domainID': 'f3f88292-1287-4635-83ac-f8c2cf482a9e', 'imageID': 'c3d127e4-b6cf-4dda-a5fb-3064953e67a3', 'volumeID': '4cab4cf5-0cfc-4eac-8e15-42bbdb9a4e7c'}'"}, code = 1600 (Failed with error unexpected and code 16)


And causing vdsm to throw ERROR:

LookupError: No such drive: '{'domainID': 'f3f88292-1287-4635-83ac-f8c2cf482a9e', 'imageID': 'c3d127e4-b6cf-4dda-a5fb-3064953e67a3', 'volumeID': '4cab4cf5-0cf
c-4eac-8e15-42bbdb9a4e7c'}'


Also, the API response returning phase "starting" instead of error message telling me that the disk is locked and can not be backed up:

POST {{engine}}vms/{{myvm_id}}/backups

Body:

<backup>
    <disks>
        <disk id="{{qcow_disk_id}}" />
    </disks>
</backup>

Response:

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<backup href="/ovirt-engine/api/vms/38e0898b-b6a8-4238-b400-e17928f6a926/backups/8d8079fa-2d25-4faf-853e-856ba22f3889" id="8d8079fa-2d25-4faf-853e-856ba22f3889">
    <actions>
        <link href="/ovirt-engine/api/vms/38e0898b-b6a8-4238-b400-e17928f6a926/backups/8d8079fa-2d25-4faf-853e-856ba22f3889/finalize" rel="finalize"/>
    </actions>
    <link href="/ovirt-engine/api/vms/38e0898b-b6a8-4238-b400-e17928f6a926/backups/8d8079fa-2d25-4faf-853e-856ba22f3889/disks" rel="disks"/>
    <creation_date>2020-04-21T11:31:36.755+03:00</creation_date>
    <phase>starting</phase>
    <vm href="/ovirt-engine/api/vms/38e0898b-b6a8-4238-b400-e17928f6a926" id="38e0898b-b6a8-4238-b400-e17928f6a926"/>
</backup>


Version-Release number of selected component (if applicable):
vdsm-4.40.13-1.el8ev.x86_64
ovirt-engine-4.4.0-0.33.master.el8ev.noarch

How reproducible:
100%

Steps to Reproduce:
1. Create blank vm
2. Create Qcow disk with incremental backup enabled + attach it to the vm as os disk on ISCSI
3. Start the vm -> wait till it starts
4. Migrate the disk to another ISCSI domain
5. As soon as the migration starts, and the disk getting locked, invoke with API full backup for the subject disk


Actual results:

The full backup request starts to process, causing Exceptions on VDSM and ENGINE.
API response should be something different than regular response which indicates that the full backup process has started.
For example: "the disk is locked and can not be backed up currently"

Expected results:
The backup operations should be blocked during disk/VM migration (live or cold) and vice-versa.

Additional info:
Attaching Engine log and relevant vdsm log.