Ceph » RADOS

The current implementation will not behave correctly for a rados operation consisting of a mutation followed by a rollback to a snapshot since the last mutation of the object.
The correct behavior would be for the rollback to undo the mutation earlier in the operation by resolving to the clone created at the start of the operation
(see PrimaryLogPG::find_object_context).
Instead, it will select HEAD leaving that mutation intact since the SnapSet won't yet contain that clone. This behavior exists in Crimson as well.

A test scenario is added to cover this case, see LibRadosSnapshotsSelfManagedPP.WriteRollback
The test will remain disabled until this case is supported.
AFAIK, no clients make use of this kind of transaction. Therefore, this tracker is marked as low priority and is created for future reference.