Hi,

I'm migrating some users from one Exchange 2013 SP1 CU6 database to another one. Some users fail with warnings:

Error: MigrationPermanentException: Warning: Target mailbox unlock operation was not replicated. If a lossy failover occurs after the target mailbox was unlocked, the target mailbox could remain inaccessible. Error details: Mailbox changes failed to replicate. Database doesn't satisfy the constraint SecondCopy because the commit time 7/12/2014 4:22:54 PM isn't guaranteed by replication time 7/12/2014 4:14:25 PM.

Part of the problem is likely that on the passive DAG member, the Microsoft Exchange Migration Workflow service keeps crashing and restarting.

I migrated probably 300 users no problem prior to CU6 (we were on CU4/SP1), this issue now happens on 1/10 migrated users. As far as I've seen, this causes the users no issues what so ever, however I haven't had a fail over to the passive DAG member since these migrations.

Simply re-running the migration jobs don't help. Thinking of CU7 but can't see anything in the release notes mentioning this at all.

Hi,

From your description, I recommend you run the Resume-ClusterNode cmdlet to resume the node from the paused state or brings back drained workloads to the node and check the result.

For your reference:

Resume-ClusterNode

https://technet.microsoft.com/en-us/library/hh847243.aspx

Hope this can be helpful to you.

Best regards,

If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.c

Hi Amy,

Thanks for your response. But the Failover Cluster Manager reports the passive DAG member as status Up.

get mailboxserver | get-mailboxdatabasecopystatus shows all DBs with healthy status and content index status and no CopyQueue or ReplayQueue over 1.

So it would seem the cluster node is perfectly fine and up?

Hi Trana010,

I have seen some people have the issue with yours.  Running the Resume-ClusterNode cmdlet works for them. I recommend you try to run it and check the result.

Hope my clarification can be helpful to you.

Best regards,

If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com

Hi Trana010,

Is there any update with your issue?

Best regards,

If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com

Any other ideas?

Hi,

If the issue persists, you can try to run the Get-ClusterNode | Resume-ClusterNode -Failback Immediate command and check the result.

Hope this can be helpful to you.

Best regards,

If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com

Can we please move on, my nodes are not paused.

Hi,

Well, the Microsoft Exchange Migration Workflow service is just a placeholder for a service will be implemented in the future. You can leave the service startup type set to Manual or change it to Disabled to prevent the system from trying to automatically start the service and check the result.

Hope this can be helpful to you.

Best regards,

If you have feedback for TechNet Subscriber Support, contact  tnmff@microsoft.com

That was true in Exchange 2013 CU3:

https://support.microsoft.com/kb/2901754

I haven't found anything indicating whether or not that still is true in Exchange 2013 CU6. For all I know its now been implemented in CU4, 5 or 6. Do you have a link that details if what you say is still true in CU6?

Also, I'm not actually all that concerned about the service, my main issues is these mailbox migration having failed to migrate with "If a lossy failover occurs after the target mailbox was unlocked, the target mailbox could remain inaccessible."

My concern with the service is that it could be the case of those failed mailbox migrations.