Consider the following scenario:
You have multiple Microsoft Lync Server 2013 Front End servers in a pool.
You have multiple Lync Server 2013 Front End pools that have pool pairingenabled.
You move users back and forth between primary and backup pools by using the Move-CsUser PowerShell cmdlet or through the Lync control panel. (This includes a scenario where the user is moved to a pool that is not a backup pool and then moved back to the original pool.)
Either of the following issues occurs during the scheduled maintenance window:
Lync Server 2013 Front End servers in a pool is restarted.
Lync Server 2013 Front End pool is failed over and then failed back.
In this scenario, certain users may experience data loss on their client.
Note please call into Support for guidance if data loss is experienced even after you apply this update.
This issue occurs because the RG data (Routing Group) in the rtclocal database may be overwritten by an incorrect PSD (PersistedServiceData, which is a log of all the transactions performed by a user that are mapped to an RG). If the users are moved back to a pool they are previously homed on, their RG assignment may change, and the pool may recognize the pre-existing version of the user data instead the new one.
To greatly reduce the chance of data loss in the described scenarios, install the April 2016 cumulative update 5.0.8308.956 for the Lync Server 2013, Front End server and Edge server and April 2016 cumulative update 5.0.8308.956 for the Lync Server 2013, core components. This update guarantees:
When a user is moved, the source pool and destination pool will try to clean up any stale user state.
If any Front End Servers are down during this operation, a warning will be flagged and the administrator will have to follow Clean up user data after a user move by using the Skype for Business Server Cleanup Tool.
This update makes sure that the user state is coordinated across the source pool and destination pool.