Data Protection for VMware and Application Data in Mission Critical Environments

The number of replication group created should be tailored to the specific user environment. For
instance, the creation of replication groups can be based on a number of decision factors, for
example:
• The disk resources for each application running in a virtual machine may require their own DR
group. This allows failing over just one or multiple applications individually to a different virtual
machine without having to failover the virtual machine they are running on.
• A virtual machine or group of virtual machines may need to have all of their disk resources in a
single DR group. Virtual machines booting off of VMDK files on the same VMFS volume would need
to all be failed over together when the LUN where the VMFS resides is failed over.
The EVA provides a limited number of DR groups so understanding your environment and its
replication granularity requirements will help you reduce the number of DR groups required for your
environment and provide improved efficiency. Consult with your HP field representative for data
replication groups’ implementation strategies.
Implementation overview
As discussed above, traditional backup approaches present many challenges to the deployment of an
effective backup strategy in a virtual environment. Surveyed data collected by VMware indicated that
80-85 percent of VMware virtual machines are created on VMFS volumes. This means that any
effective backup solution has to tightly integrate with VMware Consolidated Backup framework in
order to leverage the many benefits it provides to backup the VM operating system, OS, and
application configuration, which are encapsulated and saved on VMFS volumes. This paper
demonstrates how to deploy a complete and effective backup strategy that leverages the benefits of
VMware Consolidated Backup and complements it with proven HP backup technologies, Zero
Downtime Backup and Instant Recovery, to meet data center SLAs for RPO and RTO, while providing
a robust disaster tolerant backup solution. The backup and restore processes are described below:
Backup process
1. Virtual machine level backup (OS, OS and application configuration) provided through Data
Protector Integration with VMware Consolidated Backup
2. Application data-level backup provided through Data Protector using the Data Protector online agents
3. Data Protector Zero Downtime Backup provides continuous backup of the application data of the
application running inside a VM, while providing local and remote disaster tolerance by
leveraging EVA Business Copy and Continuous Access respectively.
Restore process
1. The full virtual machine image is restored through VMware Consolidated Backup proxy host
2. For local replication, if the data is on tape then it can be restored directly to the VM. If the backup data
still exist as a replica (EVA snapclone or XP mirror), then it can be restored through Instant Recovery
3. For remote replication, whether the data is on disk or tape, recovery is done directly into the
virtual machine
14