High Avalibility for different Lync components:

FE servers - have multiple FE's in a pool. There should be at least 3 nodes in a pool to survive if one node fails. Windows Fabric" This is due to the  

"replication architecture based on Azure EDGE server - have multiple edge servers in a pool. Pool - Do pool pairing to achieve pool resiliency BE SQL - Use SQL mirroring/clustering with or without witness File Share - use the Distributed File System (DFS).  

  

If a pool is failed over, all users of the affected pool are forced to sign out and then sign into the backup pool. For a brief period users who sign  

into the backup pool may be in resiliency mode. In Resiliency mode, users are unable to perform tasks that would cause a persistent change on  

Lync Server, such as adding a contact. After the failover is complete, all users can get all services from the backup pool. 

Any sessions a user has when the pool fails are disrupted, and the user must re-establish those sessions after failover to continue. 

Users are not rehomed during failover or failback. Users who are homed on a pool that fails will be temporarily serviced by the backup pool.  

When the home pool is restored, the administrator can fail back these users to be serviced by their original home pool.  

Note in Lync 2013, the Location Information Server database is not replicated to the backup pool. For best practice, the administrator should  

regularly back up the LIS database and use the latest backup copy to restore the LIS database in the backup pool after the failover. 

  

Services (Backup service and Centralized logging are new in Lync 2013)  

 Lync Server Backup Service – This service will appear only when pool pairing is done. It provides real-time data replication to keep the pools synchronized.  

The heart of the backup process is the file share. The File share has this folder structure: 

BackupService>BackupStore>   

CentralMgmt - Note - The Master Replica and Replica Agents are responsible for keeping the RTCLocalxds database up to date.  The backup service is responsible for keeping                                    the RTCxds database up to date.   

ConfServices - conference data 

UserServices - replication of buddy lists, conference directory ID and other user services information. 

TaskAssignment -  

 Similar to the CMS replication process, the actual process of copying data from PoolA to PoolB is via an SMB file transfer.   

What is different however, is that it appears the backup service itself is responsible for both the file copy and consumption of data. 

Get-CSBackupServiceConfiguration  This cmdlet shows  backup service will sync information every two minutes.  Additionally, it has a maximum conference size of 100Mb.