Related issue
https://github.com/longhorn/longhorn/issues/5218
Test step
Given Deploy Longhorn v1.3.2
And Created and attached a volume.
And Upgrade Longhorn to the latest.
And Do live upgrade for the volume. (The 1st volume using the latest engine image but running in the old instance manager.)
And Created and attached a volume with the v1.3.2 engine image. (The 2nd volume using the old engine image but running in the new instance manager.)
When Try to update volume.spec.unmapMarkSnapChainRemoved for both volumes via kubectl or GUI
Then Verify engine.status.unmapMarkSnapChainRemovedEnabled for both volumes is always false
And No error messages in the longhorn manager pod logs.
And Both volumes function correctly.