FusionCompute VMs

Before configuring disaster recovery (DR) services, check the FusionCompute environments of the production end and the DR end, and the storage end environment. If the FusionCompute environments do not meet requirements, modify the FusionCompute configurations.

Production End and DR End

In a FusionCompute VM environment, FusionCompute must be able to create data stores using storage resources provided by FC SAN, IP SAN, advanced SAN and Huawei Distributed Block Storage, create VMs based on the data stores, provide cluster, host, and port group resources externally, and create an interconnection user on FusionCompute.
configuration of FusionCompute and storage arrays, and configuration of FusionCompute providing clusters or host resources and port group resources. For details, see the FusionCompute V100R005Cxx Product Documentation.

Check and configure the FusionCompute environments at the production end and the DR end.
After FusionCompute is installed, you need to configure the FusionCompute environment, for example, configure clusters/hosts in FusionCompute, FusionCompute and storage arrays, and network resources in FusionCompute.

All the following configurations are basic configurations of FusionCompute. For details, see the FusionCompute V100R005Cxx Product Documentation.
1. Check the clusters and hosts in FusionCompute.
  1. Log in to FusionCompute as an administrator.
  2. Check whether any cluster or host resource is available. If no such a resource exists, create a resource.
  If you want to use a cluster to manage resources, ensure that the cluster contains available hosts.
2. Check FusionCompute and storage arrays.
  1. Log in to FusionCompute as an administrator.
  2. Check whether any IP SAN, FC SAN, Advanced SAN, NAS or Huawei Distributed Block Storage storage is available. If no such storage system exists, add a storage system.
  3. In the host resources of FusionCompute, select ports to which you want to connect so that the service ports are added for service communication between the storage and FusionCompute.
3. Optional: Check the storage plane configuration of the virtualized SAN.
  
  If there is no storage plane, you need to disable the function.
  1. Log in to FusionCompute as an administrator.
  2. On the menu, choose Storage Pool > Data Store.
  3. Click Configure Virtualized SAN Storage Plane, go to the Configure Virtualized SAN Storage Plane page and check that the storage plane configuration of the virtualized SAN has been disabled.
4. Check network resources in FusionCompute.
  1. Log in to FusionCompute as an administrator.
  2. Check whether any port group is available in the network pool and ensure that the network pool is associated with the host. If no such a port group exists, create a port group.
5. Check interconnection users of FusionCompute.
  1. Log in to FusionCompute as an administrator.
  2. Check whether Interface interconnection user exists. If no such user exists, create one and ensure that roles of administrator are selected.

Production End

Check whether datastores have been created on FusionCompute at the production end.

When creating a datastore, use a virtualized method and ensure that formatting has been performed.
1. Log in to FusionCompute as an administrator.
2. Check whether datastores have been created. If no datastore has been created, perform the following steps to create one:
3. Click Add Data Store to go to the datastore addition page.
4. Select IP SAN to create datastores.
Check whether VMs have been created on FusionCompute at the production end.
- Check whether VMs have been created on FusionCompute.
  1. Log in to FusionCompute as an administrator.
  2. Choose VM and Template > VM and check whether VMs that require DR protection have been created. If no VMs are created, follow the wizard to select an appropriate customized datastore to create VMs.
Ensure that virtualized datastores are configured for disaster recovery VMs.

Storage End

Table 1 describes the storage requirements of disaster recovery solutions.

Before using eReplication, refer to the remote replication configuration section in the related storage device document.

During the initial synchronization of remote replication, pay attention to the time spent on the synchronization. The larger the amount of data is, the longer it takes to synchronize the data. Therefore, if a large amount of data is synchronized in a short time, for example, 20 seconds, the data is probably not synchronized. In this case, delete the remote synchronization and start the initial synchronization over.

**Table 1** Requirements of disaster recovery solutions on storage
Technology	Restriction and Requirement
HyperMetro (SAN) and asynchronous replication (SAN) HyperMetro (SAN) and synchronous replication (SAN)	Check whether Huawei storage devices provide datastores for VMs in disaster recovery protection. Check whether all requirements of active-active (SAN) disaster recovery are met. Check whether remote replication or consistency groups of the arrays are correctly set up. If datastores used by VMs use multiple LUNs, ensure that the remote replications of all LUNs reside in the same consistency group. In the DR Star network scenario, the remote replication or consistency group must be added to one DR Star. Check whether secondary LUNs on the arrays are correctly mapped to production/DR hosts (groups). Especially for a cluster consisting of CNAs, secondary LUNs must be mapped only to CNA hosts in the cluster, and only map to one host or host group. If T series V200R001C00 is deployed, after creating mapping view, you cannot select the Enable Inband Command to modify the properties of the mapping view. Plan VMs that need DR protection on the same datastore if possible. If a datastore cannot provide sufficient storage for the VMs that need DR protection, create the VMs on datastores of different LUNs on the same storage array. If datastores of the protected VMs are in different LUNs, you need to use the corresponding LUNs of the datastores as primary LUNs to create remote replications. Then add the remote replications to the same consistency group. During remote replication creation, select Initial synchronization when setting Pair Properties. Ensure that the LUNs planned for DR at the DR end are not used to create datastores. Otherwise, the DR environment may become unavailable. During DR, the system will automatically create datastores using the DR LUNs at the DR end.
Cascading replication (synchronous and asynchronous) Parallel replication (synchronous and asynchronous) Cascading replication (asynchronous and asynchronous) Parallel replication (asynchronous and asynchronous)	Check whether Huawei storage devices provide datastores for VMs in disaster recovery protection. Check whether remote replication or consistency groups of the arrays are correctly set up. If datastores used by VMs use multiple LUNs, ensure that the remote replications of all LUNs reside in the same consistency group. In the DR Star network scenario, the remote replication or consistency group must be added to one DR Star. Check whether secondary LUNs on the arrays are correctly mapped to production/DR hosts (groups). Especially for a cluster consisting of CNAs, secondary LUNs must be mapped only to CNA hosts in the cluster, and only map to one host or host group. If T series V200R001C00 is deployed, after creating mapping view, you cannot select the Enable Inband Command to modify the properties of the mapping view. Plan VMs that need DR protection on the same datastore if possible. If a datastore cannot provide sufficient storage for the VMs that need DR protection, create the VMs on datastores of different LUNs on the same storage array. If datastores of the protected VMs are in different LUNs, you need to use the corresponding LUNs of the datastores as primary LUNs to create remote replications. Then add the remote replications to the same consistency group. During remote replication creation, select Initial synchronization when setting Pair Properties. Ensure that the LUNs planned for DR at the DR end are not used to create datastores. Otherwise, the DR environment may become unavailable. During DR, the system will automatically create datastores using the DR LUNs at the DR end.