Failing Over the Central Management Store

In the event of a pool or site failure, the first consideration must be of the Central Management Store (CMS). The CMS must be functional in order to provide failover of any other services. If a Front End Server pool has failed, or the site hosting that pool has failed and it was running the CMS, the CMS first must be failed to the secondary pool. The CMS is backed up as part of the pool pairing configured in earlier steps. Figure 15.15 provides an overview of the configuration to be used in the examples that follow. Pool A is currently running the CMS, and Pool A will be the failing pool.

Figure 15.15. Lync Server 2013 disaster recovery example.

1. To find out which pool is hosting the CMS, run the following command from the Lync Management Shell:

Invoke-CSManagementServerFailover -WhatIf

2. The following example assumes the CMS is hosted on LYNCPOOLA. Next, identify the backup pool relationship for Pool A. You can complete this by opening Topology Builder and reviewing resiliency settings, or by running the following command:

Get-CSBackupRelationship -PoolFQDN LYNCPOOLA.COMPANYABC.COM

3. Before initiating the CMS failover, ensure that the secondary pool has been configured to host the CMS in the event of a failover. To do this, check the database state for the existence of a CMS mirror. Run the following command (the secondary pool in this example is LYNCPOOLB.COMPANYABC.COM):

Get-CSDatabaseMirrorState -DatabaseType CentralMgt -PoolFQDN lyncpoolb.companyabc.com

4. This command displays the SQL Server FQDN that is currently acting as the primary for the secondary pool. If the SQL Server Store for Pool B is in a mirrored configuration, it is possible that there could be a primary or a mirror running the database, depending on the current health status. Assuming that the primary server is listed, run the following command to invoke the CMS failover. In this example, the primary will be SQL3.COMPANYABC.COM.

Invoke-CSManagementServerFailover -BackupSQLServerFQDN SQL3.COMPANYABC.COM

5. When this command has completed, you can validate that the CMS has failed over by running the following command. The ActiveMasterFQDN and ActiveFileTransferAgents attributes should point to the FQDN of Pool B.

Get-CSManagementStoreReplicationStatus -CentralManagementStoreStatus

6. After you have validated that the secondary pool is now hosting the CMS, run the following command to validate CMS replication across all servers. When all replicas have a value of True, you can continue with your disaster recovery procedures.

Get-CSManagementStoreReplicationStatus

At this point, the CMS should now be hosted on Pool B. The next section outlines how to initiate a pool failover.