If a fault occurs in the Production center, the fault can be rectified and cannot be rectified based on the original configuration. The corresponding DR switchback operations are different.
- Case 1: If a recoverable disaster (such as unexpected power failure) occurs in the Production center, rectify the fault based on the original configuration. After the fault is rectified, the storage and service hosts can still run properly. In this case, you need to perform active/standby switchover and data synchronization on the DR center, and then perform a reprotection task to switch the data protection direction from the DR center to the Production center. Then, migrate services from the DR center back to the Production center.
- Case 2: If the Production center cannot be recovered based on the original configuration, rebuild the data center. After the Production center is rebuilt, delete the original protection task from eReplication, reconfigure the DR relationship between the DR center and the Production center, and perform planned migration to switch services to the Production center. After the planned migration is complete, you need to perform reprotection to reset the replication relationship, that is, data is replicated from the HyperMetro center to the DR center. In this way, DR protection is provided for production services again. This section describes how to migrate services back in Case 1. For case 2, contact Huawei technical support to provide a detailed technical solution.
After Production centers recover from a fault, services will be migrated back to them.
Prerequisites
- Application-based protection policies and recovery plans have been configured for database applications and FusionSphere deployed on physical machines.
- LUN-based protection policies and recovery plans have been configured for other applications.
- For protected groups that use asynchronous remote replication, the DR center has at least one copy of intact service data.
- The fault recovery has been successfully executed based on the recovery plan on the original DR center.
In the Oracle application, the cluster is not stoppd by default. To stop the cluster, you need to open the configuration for stopping the cluster and then perform the next step. For details about how to stop the cluster, see 4.2 Stopping the Oracle Cluster.
Procedure
- Log in to the service system in the original production center and check the service startup conditions.
Ensure that the applications in the production center have been stopped and prepare for disk uninstallation.
In scenarios where databases are deployed, you must manually close the applications in the production center.
- Uninstall disks on application hosts.
- Perform an active/standby switchover.
Log in to the storage array in the DR center and perform an active/standby switchover on the remote replication or consistency group that is in split status.
- Enable write protection of the secondary resource.
After the function is enabled, data cannot be written into the secondary LUN, ensuring data security of the secondary LUN.
- Log in to the storage management page in the production center and select the remote replication pairs whose secondary resource you want to protect on the remote replication management page.
- Click Properties, choose More > Enable Secondary Resource Protection and change the value of Status of Secondary Resource Protection to Read-only.
- Refresh the resources that are registered in eReplication.
On the Resource page, select the device to be refreshed, and click Refresh in the Operation column to refresh the device information. Table 1 describes the mapping relationship between protected objects and devices to be refreshed.
Table 1 Mapping relationship between protected objects and devices to be refreshed Protected Object
Device to Be Refreshed
Oracle, DB2, SQL Server, or SAP HANA database
Hosts where databases reside and storage devices related to databases
VMware VM
Storage devices associated with VMs.
LUN
Storage devices where LUNs reside
- Perform reprotection.
Copy data from the DR center to the production center to complete the initial data synchronization.
- In the navigation tree, click
Utilization > Data Restore. - Select the recovery plan and choose More > Reprotection in the Operation area.
- In the dialog box that is displayed, carefully read the content and click OK to confirm the information.
To ensure the normal running of protected groups and recovery plans after reprotection, the system automatically clears protected and recovered configurations, including startup configurations of protection policies and recovery plans, self-defined execution scripts, and self-defined execution steps. The protection polci will be changed to On-demand scheduling. In addition, reconfiguration of protection and recovery policies is recommended to ensure the continuity of DR services.
- In the navigation tree, click
- Test the recovery plan and clear the test data on eReplication. During the testing, select a host or host group in the production center if necessary.
Before the services are switched back, you need to perform a DR test to verify the data availability to ensure success rate of service failback. After the test is complete, clear the test data generated from the DR test to prevent the test data from adversely affecting the service migration switchback rate. For details, see DR Test/Clearance.
- Refresh protected groups.
Because eReplication's adjustment to the storage replication relationship remains unknown, protected groups are in Invalid status. You need to manually refresh them.
- On eReplication, perform a planned migration.
This step aims to switch services back to the production center from the DR center. After migration, data must be checked and test data must be cleared. For details, see 1 to 3 in Performing Planned Service Migration in a Production Center.
- On eReplication, perform reprotection again.
Copy data from the production center to the DR center to recover the protected group status.