Prerequisites for configuring applications for disaster recovery
To be able to perform disaster recovery (DR) operations on the applications in the data center, ensure that the following requirements are met.
If applications are installed on VMware virtual machines, you need to configure the VMware vCenter for discovery. Similarly if the applications are installed on Hyper-V virtual machines, you need to configure the Hyper-V server for discovery.
The replicated storage provided to the application hosts from the storage arrays is discovered if you have configured the discovery hosts and the enclosures.
Replicated storage can be provisioned to the application in numerous ways. For applications installed inside virtual machines, Resiliency Platform discovers storage that is provisioned to the virtual machines in raw mode (e.g. VMware RDM mode) from the hypervisors. Raw mode can also be achieved by using NPIV HBAs and provisioning storage (LUNs) directly from the arrays to the Virtual HBA Ports assigned to the virtual machines.
If the storage is provisioned to the application virtual machines using hypervisor, then you need to provision appropriate snapshot devices to the virtual machines in the same manner as the replicated devices. This is required to perform the rehearsal operation successfully.
For NetApp SnapMirror and HP 3PAR Remote Copy, the snapshot devices do not exist beforehand and are created as a part of the Rehearsal operation. These snapshot devices should be mapped to the application hosts. As part of the Rehearsal operation, Resiliency Platform maps the snapshot devices to the same hosts to which the replicated LUNs are mapped. If the replicated LUNs are mapped through the hypervisor, then Resiliency Platform cannot further provision them to the application virtual machines.
As a result, Rehearsal is not supported with these replication technologies when the application is installed in virtual machines with replicated storage mapped as raw LUN by the hypervisor.
Applications must be members of the same Veritas Replication Set.
They must use the same replication technology.
Application binaries should be stored on local storage and data files on replicated storage.
The DNS server settings should be configured for both data centers. DNS settings are required for binding a host name to different IP addresses on the DR site. This is required only if you plan to use the Resiliency Platform for performing DNS updates.
Ensure that you have disabled the 'Quick removal' policy for disks in Windows Server 2008 R2 and disabled the 'write-cache' policy for disks in Windows Server 2012 R2.
Applications should be preconfigured at the recovery data center before creating resiliency groups and proceeding with the DR operations. Discovery of these applications on the recovery data center must be complete and status should be offline. Ensure that the application instance name should be same at the production and recovery data center.
Linux LVM Volume Group used for data file storage of an application should not be in exported state.
On Linux application hosts, the data mount point of the application should be present in
/etc/fstabfile.For more information on configuring applications for remote recovery using EMC SRDF, RecoverPoint, and NetApp SnapMirror replication see the following topics:
See Configuring applications for disaster recovery using EMC RecoverPoint replication.
See Configuring applications for disaster recovery using NetApp SnapMirror replication.
More Information