Creating a Protected Group for FusionCompute VMs

This section describes how to create a protected group at the production site for FusionCompute VMs that need DR protection. After the protected group is created, the system automatically matches suitable protection policies for the protected group.

Prerequisites

At least one production site has been created.
Resource mappings have been created.

Precautions

Only remote replication pairs can exist between the primary and secondary LUNs.
Only HyperMetro pairs can exist between HyperMetro LUNs.
For VMs that are protected, the boot mode cannot be set to Specify Device Boot Sequence, and do not select such VMs when you create a protected group.

Context

In the HyperMetro (SAN) and asynchronous replication (SAN) scenario, storage LUNs used by protected objects in a protected group must reside in a same HyperMetro consistency group, and the HyperMetro consistency group only have this LUNs.
When using active-active (SAN) and asynchronous replication (SAN) DR technology, ensure that active-active and asynchronous remote replication pairs must be in the same consistency group, respectively.
If the HyperMetro (SAN) + asynchronous replication (SAN) DR technology is used in the DR Star network, select HyperMetro and Array-Based Replication(DR Star): Disaster Recovery Data Center Solution (Geo-Redundant Mode).
After a protected group is created, it may change due to increase or decrease of VMs or VM capacity. After the change, previous DR requirements cannot be met and adjustments must be made.

After adding or removing disks for a protected VM, refresh the information about the VM and manually enable DR for the protected group where the VM resides in time.

Table 1 describes the protection status of a VM after its protected group changes.

**Table 1** Protection status description
DR Technology	Protection Status Description
Storage replication	Protected after the change	Capacity of VM disks increases. Some disks are deleted from the VM. New disks are mounted to the VM and reside in the current datastore group. Disks mounted to the newly-created VM must reside in the current datastore group and in the same FusionCompute and FusionManager as the VM. When FusionCompute and FusionManager of the VM are refreshed, the VM is added to the protected group and protected automatically.
Storage replication	Unprotected after the change	New disks are mounted to the VM but do not reside in the current datastore group. If you want to protect the VM, migrate the newly-mounted disks to the datastore group to which the protected group belongs by referring to the Event Reference of the User Guide. Not all disks mounted to a newly-created VM reside in the current datastore group. Therefore, the VM is not totally residing the current datastore group. The system reports an alarm. If you want to protect the VM, migrate the newly-mounted disks to the current datastore group by referring to the Event Reference of the User Guide. Alternatively, add LUNs of disks that are not residing in the current datastore group to the consistency group of remote replication. A VM is deleted from the protected group or some or all disks of a VM are moved out of the current datastore group. The remote replication pair or consistency group corresponding to the protected group is deleted or not active. Configuration information about a protected VM changes. As a result, resource mappings no longer exist and the system reports an alarm. If resource mappings configured for a newly-created VM do not exist, an alarm is reported.

Procedure

On the menu bar, choose Protection.
Click Create.
Select a protected object.
1. Select Production Site, FusionSphere Component and Protection Type.
2. Select the datastore group where the VMs to be protected reside.
  The system automatically protects all VMs in the selected datastore group.
  
  A datastore group is the minimum independent unit for failure switchover or failure test. It corresponds to a remote replication pair or consistency group on the storage array. Every datastore in a datastore group must contain at least one VM.
Click Next.

Set a protection policy.

On the Protection Policy page, the system automatically matches supported protection policies. Select a policy based on your data protection solution.
- When the HyperMetro (SAN) and asynchronous replication (SAN) DR technology is used, select HyperMetro and Array-Based Replication (Asynchronous):Geo-Redundant DR solution.
- If the datastore where VMs reside managed by FusionCompute uses the HyperMetro (SAN) + asynchronous replication (SAN) DR technology of DR Star, select HyperMetro and Array-Based Replication(DR Star): Disaster Recovery Data Center Solution (Geo-Redundant Mode).
  - If there are matched protection policies, you can select a protection policy based on your data protection solution and modify the time policy of the protection policy.
  - If there are no matched protection policies, check whether the storage devices used by protected objects have been configured as planned. If the storage devices have not been configured, configure them as planned. After configuration, perform the following operations: On the menu bar, click Resources, select a node under the management server. In the site node, click Storage, in the line where the storage device resides, click Refresh to refresh the storage device information under the site. Refresh the FusionCompute or FusionManager information too and recreate a protected group.

Click Set to set the protection policy for the protected group.

For the HyperMetro and Array-Based Replication (Asynchronous):Geo-Redundant DR solution protection policy:

Click the Scheduling Policy tab. In the Time Policy area, set the time policy for the protected group. Table 2 lists related parameters.

**Table 2** Time policies
Time Policy	Description
On-demand scheduling	Manually starts protection tasks.
Period-based scheduling	Automatically starts protection tasks based on a period set by the system. NOTE: To ensure data consistency, the execution timetable for a protected group policy template must be set. You are advised to set the execution time to off-peak hours. Tasks can be executed at multiple time points of a day, a week, or a month. Alternatively, at a specific time segment of a day, a week, or a month with a fixed frequency. It is recommended that the starting time of the validate period is set to a time later than the completion of the initial synchronization. If daylight saving time (DST) is enabled on the UltraVR management server, a DR protection task whose trigger time point falls within the DST shift period cannot be executed at the scheduled time. Details are as follows:(UTC-08:00) Pacific Time (United States and Canada) is used as an example. The DST starts at 2016-03-12 02:00:00, and ends at 2016-11-06 02:00:00. If the start time of the protection policy falls within the first hour (2016-03-12 02:00:00 to 03:00:00) after the DST starts, for example, 2016-03-12 02:30:00, the protection group will be automatically executed with a delay of one hour, that is 2016-03-12 03:30:00. If the start time of the protection policy falls with the last hour (2016-11-06 01:00:00 to 02:00:00) before the DST ends, for example, 2016-11-06 01:30:00, the protection group will be executed after the DST ends for half an hour.

When the Scheduling Policy is set to Period-based scheduling, in the Expected RPO area, if you select Enable PRO Requirement Satisfaction Check, the actual RPO is check against the configured RPO. If the actual RPO exceeds the expected RPO, the system generates an alarm, indicating that the RPO requirement is not satisfied.

Expected RPO can be set to 0-31 days, 0-23 hours, and 5-59 minutes.

If days and hours are not set, the Expected RPO is 15 minutes by default.

Click Replication Policy tab to set the remote replicate rate between storage arrays. Table 3 lists related parameters.

**Table 3** Replication policies
Replication Policy	Description
Default configuration	The remote replication rate is the default one set on the storage array.
Manual configuration	Users can set the remote replication rate for each time segment. Configure the automatic remote replication rate of the storage based on the pre-set replication rate of time period. Select Manual configuration. Click a time color block to modify the replication rate during a time segment.

Click OK.

For the HyperMetro and Array-Based Replication(DR Star): Disaster Recovery Data Center Solution (Geo-Redundant Mode) protection policy:

Click the Scheduling Policy tab. In the Time Policy area, set the time policy for the protected group. Table 4 lists related parameters.

**Table 4** Time policies
Time Policy	Description
On-demand scheduling	Manually starts protection tasks.
Period-based scheduling	Automatically starts protection tasks based on a period set by the system. NOTE: To ensure data consistency, the execution timetable for a protected group policy template must be set. You are advised to set the execution time to off-peak hours. Tasks can be executed at multiple time points of a day, a week, or a month. Alternatively, at a specific time segment of a day, a week, or a month with a fixed frequency. It is recommended that the starting time of the validate period is set to a time later than the completion of the initial synchronization. If daylight saving time (DST) is enabled on the UltraVR management server, a DR protection task whose trigger time point falls within the DST shift period cannot be executed at the scheduled time. Details are as follows:(UTC-08:00) Pacific Time (United States and Canada) is used as an example. The DST starts at 2016-03-12 02:00:00, and ends at 2016-11-06 02:00:00. If the start time of the protection policy falls within the first hour (2016-03-12 02:00:00 to 03:00:00) after the DST starts, for example, 2016-03-12 02:30:00, the protection group will be automatically executed with a delay of one hour, that is 2016-03-12 03:30:00. If the start time of the protection policy falls with the last hour (2016-11-06 01:00:00 to 02:00:00) before the DST ends, for example, 2016-11-06 01:30:00, the protection group will be executed after the DST ends for half an hour.

When the Scheduling Policy is set to Period-based scheduling, in the Expected RPO area, if you select Enable PRO Requirement Satisfaction Check, the actual RPO is check against the configured RPO. If the actual RPO exceeds the expected RPO, the system generates an alarm, indicating that the RPO requirement is not satisfied.

Expected RPO can be set to 0-31 days, 0-23 hours, and 5-59 minutes.

If days and hours are not set, the Expected RPO is 15 minutes by default.

Click Replication Policy tab to set the remote replicate rate between storage arrays. Table 5 lists related parameters.

**Table 5** Replication policies
Replication Policy	Description
Default configuration	The remote replication rate is the default one set on the storage array.
Manual configuration	Users can set the remote replication rate for each time segment. Configure the automatic remote replication rate of the storage based on the pre-set replication rate of time period. Select Manual configuration. Click a time color block to modify the replication rate during a time segment.

Click OK.

Click Next.
Set Name and Description of the protected group.
- Name: contains 4 to 32 characters, including letters, digits, underscores (_), and hyphens (-), but cannot start with a hyphen (-).
- Description: contains 0 to 255 characters.

Optional: If you do not want to manually create a recovery plan for the protected group, select Automatically create a recovery plan after creating protected groups in the lower left corner.
Click Finish.
Click OK.

Related Operations

If the protected objects of a protected group or their used storage devices change. On the menu bar, select Resources. And refresh resources or storage devices where the protected objects reside. Then go to the protected group page and refresh protected objects of the protected group by clicking More > Refresh. After the command for refreshing information is sent, you can view details about the command execution in Background Task.

If a recovery plan has been created for a protected group, you can refresh protected objects only when the recovery plan status is Ready, Clear completed, Rollback completed, or Reprotection completed.
To ensure that information about protected groups is refreshed successfully, you must refresh information about storage devices and hosts at both the production site and DR site.