Virtualization in VMware environments is supported for DPX Block Data Protection.
Note: Ensure that TLS protocol is enabled on the NetApp Storage System. See Enabling or disabling TLS on NetApp's Support site.
DPX supports three types of virtualization from two different sources:
• | From a snapshot: |
• | Full Virtualization: Creates a production machine. |
• | Instant Virtualization: Creates a temporary machine, typically for setup or testing. |
• | From a virtual machine created with Instant Virtualization: |
• | Rapid Return to Production (RRP): Quickly transforms a temporary machine to a production machine. |
• | “Virtualization from a Snapshot: Full and Instant Virtualization” |
• | “Migration from Virtual Snapshot Storage to Local Storage: RRP” |
DPX creates virtual machines from Bare Metal Recovery (BMR) snapshots stored on a storage system. In the case of RRP, the source is an existing virtual machine created with Instant Virtualization.
Creating virtual machines with DPX is covered in the following main topics:
• | “Creating Virtual Machines with Instant or Full Virtualization” |
• | “Rapid Return to Production” |
• | “Post-Virtualization Considerations and Procedures” |
• | Virtualization jobs are based on routine Block backups. These backups may be referred to in varying ways in the manuals, Help, knowledge base, and white papers: |
- Backup instance
- Source instance
- Recovery point
- Snapshot
- Image snapshot
The terms backup instance and source instance are more general and can refer to either file-level or block-level backups. On the other hand, recovery point, snapshot, and image snapshot refer to Block backups. Virtualization documentation generally uses the term snapshot.
• | You may see the term virtual controller in a virtualization wizard or documentation. This term refers to an ESX server or vCenter Server. |
• | VM, the common acronym for virtual machine, is generally used in documentation. The term virtual machine is used in headings and in virtualization wizards. |
DPX provides two distinct types of virtualization based on routine snapshots: Full Virtualization and Instant Virtualization.
Note: RRP is a different case, as RRP virtualization is based on an existing Instant Virtualization VM, not a snapshot.
Most Full Virtualization jobs from snapshots of Windows clients are eligible for RRP processing and will transparently use RRP to bring the machine back in minutes. Such jobs are called Full Virtualization with RRP (shortened form, Full with RRP). For more information about RRP, see “Migration from Virtual Snapshot Storage to Local Storage: RRP”.
Note: All the source disks will be created as VMDK files on the specified datastore, based on the snapshot image from the secondary storage system.
VMs created with Instant Virtualization are considered temporary, since they depend on snapshots that may expire or be deleted. Such temporary VMs may also run slower due to network communication constraints. Once a temporary machine is ready for production, it can be made available quickly by using RRP. See “Migration from Virtual Snapshot Storage to Local Storage: RRP”.
Note: All the source volumes will be virtualized as RDM disks, based on the snapshot image from the secondary storage system.
Rapid Return to Production (RRP) creates a full production VM from an existing VM created with Instant Virtualization. RRP makes the new production VM available for use in minutes, while DPX transfers data behind the scenes, using VMware Storage vMotion to populate the selected datastore
RRP is particularly useful when you have used Instant Virtualization to create a test environment that is ready for production. When migrating a temporary, test VM to a permanent VM, RRP can save many hours, or even days in the case of large data transfer over a slow connection.
RRP is similar to Full Virtualization, but the source for RRP jobs is an existing VM rather than a snapshot.
Tip: RRP virtualization jobs can be scheduled to run just as any other job: once at one or more specific times, daily, weekly or monthly. In most cases, an RRP job is for a single use, so you would schedule it to run once at a specific time.
Note: Most Full Virtualization jobs are eligible for RRP processing. In those cases, RRP is utilized in the background to create a full VM, reducing time to production availability to a matter of minutes. Such jobs are called Full with RRP. Note that Full Virtualization jobs on Linux always run with RRP.
For more information about RRP, see “Rapid Return to Production”.