Initiating failover processing for B volumes to A volumes

Complete this task to run failover processing for B volumes to A volumes so that the B volumes become the primary volumes and A volumes become the secondary volumes. The Global Copy state of the B volumes changes from secondary to primary and suspended. This is the next step after you obtain a status of your Global Mirror transactions in the Global Mirror recovery failover process.

Before you begin

Failover processing cannot be run unless following tasks are complete:
  • Global Mirror processing complete
  • Reports that help you determine the status of your Global Mirror transactions before the disaster occurred are available

About this task

Recovery failover processing on the Global Copy volume pair ends the A-to-B volumes extended distance relationship and instead creates a B-to-A volume Global Copy relationship. Therefore, failover processing to the Global Copy secondary volume turns the secondary volumes into primary volumes and immediately suspends these volumes. Because the B volumes are now the primary volumes and the A volumes are now the secondary volumes, the Global Copy state of the secondary volumes changed from Target Copy Pending to Suspended.

To start failover processing of the B volumes to the A volumes, complete the following step.

Procedure

Enter the failoverpprc command to change the Global Copy state on the B volumes from secondary, target pending to primary, suspended.

dscli> failoverpprc -dev storage_image_ID -remotedev storage_image_ID SourceVolumeID:TargetVolumeID

Example

dscli> failoverpprc -dev IBM.2107-75FA150
-remotedev IBM.2107-75FA120 0100:0100 0101:0101 0102:0102 0103:0103
Note: Unlike most other commands, the storage_image_ID and SourceVolumeID:TargetVolumeID values are reversed for the failoverpprc command because the command is issued directly to the remote server. You must place the values for the remote server first and the values for the original primary server second. This reversal (where the remote server is no longer the primary server) can have the following results:
  • The source volumes become suspended
  • The target volumes become suspended and they are available for mounting
An information message like the following one is generated for each Global Copy pair that is changed and moved to a suspended state.
Remote Mirror and Copy pair IBM.2107-75FA150/0100:
IBM.2107-75FA120/0100 successfully suspended.

All B volumes must successfully process the failoverpprc command before you can proceed to the next step to analyze the FlashCopy® relationships and enter the appropriate commands.