Migrating volumes using GAD

Global Active Device for VSP One Block

Version

10.5.x

File Size

25537 KB

Audience

anonymous

Part Number

MK-23VSP1B011-11

ft:lastEdition

2026-06-23

The global-active device (GAD) feature enables you to perform data migration without interrupting business operations.

Note: Online migration is assumed, but I/O operations from the host need to be temporarily stopped if switching to the redundant path is not available. For example, depending on the specifications of the high-level configuration of the storage system (including the operating system and the multi-path software), the path to a different WWN might not be recognized as the redundant path.

Create GAD pairs between the primary and secondary storage systems.
See information about using CCI for configuration and pair management. The data on the volumes is duplicated, and the server issues I/O operations to volumes in both storage systems of the GAD pairs. For details about failure recovery and maintenance operations, see information about disaster recovery of global-active device and about maintenance of global-active device.
Note: When migrating data using GAD, you do not have to set a volume for the quorum disk.
Monitor the status of the GAD pairs, and make sure that the pair status of all pairs is PAIR before continuing.
Perform the following operations on the server:
1. Disable or delete the redundant path from the server to the primary site.
2. Stop I/O operations from the server to the primary volumes at the primary site.
At this time, do not stop I/O operations from the server to the secondary volumes at the secondary site. Confirm that there is no I/O from the server to the primary volumes at the primary site, and then move to the next step.
At the secondary site, suspend the GAD pairs by specifying the S-VOLs.
When you suspend a GAD pair by specifying the S-VOL, the pair status and I/O mode of the P-VOL and S-VOL change as follows:
- The pair status of the S-VOL changes to SSWS, and the I/O mode of the S-VOL changes to Local.
- The pair status of the P-VOL changes to PSUS, and the I/O mode of the P-VOL changes to Block.
  Command example:
```
pairsplit -g oraHA -RS -IH1
```
At the secondary site, delete the GAD pairs by specifying the S-VOLs.
When you delete a GAD pair by specifying the S-VOL, the GAD reserve attribute is applied to the volume that was the P-VOL, which prevents host access to the volume. The volume that was the S-VOL keeps the virtual LDEV ID and continues to receive I/O from the server.
Command example:
```
pairsplit -g oraHA -R -IH1
```
After you have deleted the GAD pairs, at the primary site delete the LU paths to the volumes that were the P-VOLs.
If desired, you can now delete the volumes at the primary site, as they have been nondisruptively migrated to the secondary site.
On the primary and secondary storage systems, release the quorum disk settings.
Command example:
```
raidcom delete quorum -quorum_id 0 -request_id auto -IH0
REQID: 100
```
Note: A maximum of 65,280 request IDs can be assigned to one storage system. If there is no available request ID, the command fails with EX_IDEXHA. If that occurs, see the description of the function of request IDs in the Command Control Interface Command Reference and release request IDs.
Specify the request ID that was displayed in step 7, and check the error information about the CCI configuration command that is performed asynchronously. Verify that the value in the Description column is a hyphen (-).
Command example:
```
raidcom get command_status -request_id 100 -IH0
REQID R SSB1 SSB2 Serial# ID Description
00000100 - - - 822222 - -
```
Confirm that asynchronous command processing has completed, and then release a request ID.
Command example:
```
raidcom reset command_status -request_id 100 -IH0
```
If you migrated data in a configuration with a volume set for the quorum disk, on the primary and secondary storage systems, delete, if necessary, the external volume used as the quorum disk.
If you migrated data in a configuration with no volume set for the quorum disk, you can skip this step.
On the primary and secondary storage systems, delete the remote connections between the storage systems.
Command example:
```
raidcom delete rcu -cu_free 822222 M800 0 -IH0
```
If necessary, uninstall the GAD license.
Remove the physical paths between the primary and secondary storage systems.
Stop and remove the storage system at the primary site.