Testing a Recovery Plan

You can test a recovery plan to verify the availability of data replicated to the DR site or snapshots. Test data is generated as snapshots at the DR site without affecting the production site. After the test, you must clear the test environment. Before performing fault recovery or planned migration, you are recommended to perform at least one successful DR test.

Prerequisites

You have logged in to the eReplication as a user with DR management permission.
The production site and DR site communicate with each other correctly. The management system and the DR environment at the DR site are working correctly.
At least one recovery plan has been created in the system.
If verify the availability of data replicated to the DR site, a normal remote replication relationship has been established between the storage device in the production site and that at the DR site.
The recovery plan's status is Ready, Reprotection completed, or Clear completed.
If storage devices, hosts or VMs on production site or DR site change, storage devices, hosts or VMs at the site where a protected group resides must be refreshed, for details, refer to Refreshing Resource Information.
If application data is automatically replicated by the storage systems instead of being replicated based on the timetable specified upon the creation of protected groups, you need to suspend the data replication when performing a disaster recovery test in case of a test failure. You can use either of the following methods to suspend data replication in the device management software:
- If the status of the remote replication pair for the protected applications is synchronized and data is consistent, split the remote replication to stop data replication.
- Configure the replication policy of the remote replication for the protected applications to manual synchronization.

Demands in protection objects are described as following table.

Protection Object	Description
Oracle	For a database, the same database has been created on the DR host and the production host, with the same configuration for items such as instance name, database name, and storage path. After creating the database at the DR site, you have copied the Oracle pfile in the production site and modified the cluster configurations in the pfile copy based on the Oracle deployment at the DR site. After that, an spfile has been generated using the pfile copy and saved at the DR site. This prerequisite is applicable when an Oracle RAC cluster is used in the production site and a standalone Oracle application is used at the DR site. For a Linux operating system, the configuration file is saved in $ORACLE_HOME/dbs. For a Windows operating system, the file is saved in $ORACLE_HOME/database.
IBM DB2	For a database, the same database has been created on the DR host and the production host, with the same configuration for items such as instance name, database name, and storage path.
Microsoft SQL Server	For a database, the same database has been created on the DR host and the production host, with the same configuration for items such as instance name, database name, and storage path. In SQL Server cluster, the disk resources allocated to database files have been set to maintenance mode in the cluster manager before the database is started. This prevents disk resource mount failure during database startup.
Microsoft Exchange Server	For a database, the same database has been created on the DR host and the production host, with the same configuration for items such as instance name, database name, and storage path.
FusionCompute VM	If the network between production and DR end is not insulated. After the recovery plan is created, you need to configure different IP addresses for production and DR end recovery on the Protected Object tab page, ensure production service is normal for IP address conflict. If add or delete the disk in the protected VM, refresh the VM and execute this protected group of protected VMs reside manually.
VMware VM	If the ESXi cluster (hosts) network between production and DR end is not insulated. After the recovery plan is created, you need to configure different IP addresses for production and DR end recovery on the Protected Object tab page, ensure production service is normal for IP address conflict.
SAP HANA	Databases on DR hosts must be the same as those on production hosts, including the instance names, database names, and storage paths in use.

Context

In the DR test, snapshot mappings can be created only in starter mode and port mapping is not supported.

Since disaster tolerance testing is a major operation, there are the following considerations when conducting disaster tolerance testing:

Before testing a protected group whose protection type is database, ensure that all database connections at the DR site are disconnected.
Prevent all the system and service administrators from performing other maintenance operations.
Clear the test data after a test; otherwise, you cannot perform the next test.
Clear the test data after the test is completed. If the network is disconnected or the eReplication environment is closed on purpose during the test, some test data may fail to be deleted after the environment is restored and an automatic clearance command is delivered. When this happens, manually clear the data before you run the automatic clearance command.

Suppose the VM where the production site resides is configured with IP addresses but without NIC. After a DR test, the IP addresses of the VM where the DR site resides are:

Windows:
- VM specifications already configured: IP addresses that are dynamically obtained.
- VM specifications not configured: Consistent with those of the VM where the production site resides.
Linux:
- IP addresses that are dynamically obtained

Procedure

On the menu bar, select Utilization > Data Restore.
Select the recovery plan that you want to test and click Test on the Operation list.
The Test dialog box is displayed.
Perform either of the following operations based on the protected object type.

If Huawei multipath software has been installed on the Linux-based DR host, ensure the configured I/O hanging time is not 0 and all virtual devices generated by the software have corresponding physical devices. For more details, see the OceanStor UltraPath for Linux xxx User Guide.
- If the recovery plan is based on the snapshot or clone protection policy template, select the local copy that needs to be installed on the test host (group) and click OK.
- If the recovery plan is based on the HyperVault or snapshot protection policy template, select Test Host and then select the tested copy, and click Test.
In the Protected Object area, click View to view the currently protected objects, and used storage resources are changed in the Associate protected objects dialog.
- If the recovery plan is based on the remote replication protection policy template and the protected object type is LUN, Local File System, Oracle, IBM DB2, Microsoft SQL Server, or Microsoft Exchange Server, perform the following procedure:
  1. Select DR Site.
  2. Select Available DR Hosts or Host Groups (This operation is optional when the protected object type is LUN).
    If the storage array used at the DR site is T series V2 or later, the to-be-recovered host selected by a user can belong to only one host group on the storage array, and the host group can belong to only one mapping view. Moreover, the storage LUN used by protected applications and its corresponding secondary remote replication LUNs must belong to one LUN group, and the LUN group must reside in the same mapping view as the host group. If the storage array version is T series V2R2, deselect Enable Inband Command to change the mapping view attribute after the mapping view is created.
    If the storage is T series V2R2 or later or 18000 series, the storage support automatical host adding and storage mapping function. If the connection between the storage device and the host initiator is normal, the hosts, host groups, LUN mappings, and mapping view will be created in the storage device automatically. The creation principles are as follows:
  3. Click OK.
  4. In the Warning dialog box that is displayed, read the content of the dialog box carefully and select I have read and understood the consequences associated with performing this operation.
  5. Click OK.
- If the recovery plan is based on the remote replication protection policy template and the protected object type is VMware VM, perform the following procedure:
  1. Select the test cluster.
    The VMs are restored in a test cluster. You need to select Test Site, Test vCenter and Test Cluster.
    
    You need to configure a test cluster information before selecting the test network first time.
  2. Select the test network.
    The network used for resource mapping is used for the test by default. You may choose another network based on the onsite conditions.
    
    If the Production Resource and DR Resource are not paired, select Production Resource and DR Resource then click Add to the mapping view to pair.
  3. Select the non-critical VMs.
    Select the non-critical VMs that you want to shut down to release computing resources.
  4. Click Test.
  5. In the Warning dialog box that is displayed, read the content of the dialog box carefully and select I have read and understood the consequences associated with performing this operation.
  6. Click OK.
- If the recovery plan is based on the remote replication protection policy template and the protected object type is FusionCompute VM, perform the following procedure:
  1. Select a cluster to be tested.
    VMs will be recovered in the test cluster. Set Test Site.
    
    Upon the first test network selection, you need to set the test cluster information.
  2. Select a testing network.
    The default test network is the network for resource mapping. If you want to change the default network, plan or select another network based on site requirements.
  3. Select an available powered-on host.
    The available powered-on host can provide resources for VMs.
  4. Select non-critical VMs.
    In the Available VMs list, select non-critical VMs you want to stop to release computing resources.
  5. Click Test.
  6. In the Warning dialog box that is displayed, read the content of the dialog box carefully and select I have read and understood the consequences associated with performing this operation.
  7. Click OK.
- If a recovery plan of the protection policy template is created based on HyperMetro (SAN) and the protected object type is FusionComputeVM, VMware VM, Oracle, IBM DB2, Microsoft SQL Server, or InterSystems Caché, perform the following operations:
  1. Select a storage device to stop receiving host service requests.
    
    This operation can only be performed to storage devices of T series V3R3C10 or a later version.
  2. Click Test.
  3. In the Warning dialog box that is displayed, read the content of the dialog box carefully and select I have read and understood the consequences associated with performing this operation.
  4. Click OK.

Result

After the test starts, you can view the execution process and result. Clear failed recovery plans. If the test is failed, you can solve the problem and execute the test again after clearing the test data.