Description of problem: Currently, in diskReplicateFinish when a replication is not progress or when the VM has an attribute of a replication in progress (diskReplicate attribute) the same "Drive replication error" will be returned to the engine. This makes it impossible to perform a cleanup in case of a failed live storage migration attempt. To resolve this it is required to have a separate error returned to the engine in case the replication is not in progress to allow a cleanup. Otherwise, the user will have to perform the target cleanup manually in order to make another attempt at a live storage migration. Version-Release number of selected component (if applicable): How reproducible: 100% Steps to Reproduce: 1. Start LSM 2. Once we reach the disk replication starts we abort the blockjob manually, can be done with virsh blockjob <dom> <path> --abort Actual results: The engine will attempt to finish the replication if it succeeds, it will clean up the target SD, if it will not, a clean up will not be attempted. The latter happens in described scenario Expected results: Engine should attempt a clean up if the replication was not in progress. Additional info:
Verified with the following code: ---------------------------------------- ovirt-engine-4.2.2.6-0.1.el7.noarch vdsm-4.20.23-1.el7ev.x86_64 Verified with the following scenario: ------------------------------------------- Steps to Reproduce: 1. Start LSM 2. Once we reach the disk replication starts we abort the blockjob manually, can be done with virsh blockjob <dom> <path> --abort >>>>> Engine cleans up the replication that was not in progress. 3. Moved the disk again after the previous failed move >>>>> works fine Moving to VERIFIED
This bugzilla is included in oVirt 4.2.2 release, published on March 28th 2018. Since the problem described in this bug report should be resolved in oVirt 4.2.2 release, it has been closed with a resolution of CURRENT RELEASE. If the solution does not work for you, please open a new bug report.