Scale-Out Backup Datasheet

 

Datrium  Benefits

Fast
Instant operations (snaps, clones, restores) without overheads.

Effortless
Dynamic VM & datastore-file policies with Protection Groups.

Efficient
VM & vDisk-level granularity, fully dedup/compress-aware.

Resilient
Calmly find protected objects in Snapstore

Simple, Granular, and Agile Data Management

In an era where data can be used for business insights, the focus for data management has shifted from just storing copies (snapshots) of data to using the snapshots for dev/test, analytics, reporting and compliance. With 1000s of virtual machines (VM) and objects to manage, finding the desired object(s) and performing operations as quickly and efficiently as possible has become a critical requirement.

Unfortunately, current approaches don’t go far enough in meeting these requirements. In traditional arrays, data protection is performed at the granularity of LUNs, which can store scores of VMs. A LUN-based snapshot of all stored VMs is often used to protect only a small subset.  Subsequently operating on a specific object for restore or sharing becomes a cumbersome and inefficient exercise. While some newer hyperconverged infrastructure (HCI) systems improve upon arrays by offering VM-level granularity, customers often desire to restore or share data without replacing and reconfiguring the entire VM.  VMs have multiple virtual disks attached to them and just like customers want to target specific objects within LUNs, customers want to target virtual disks within a VM.

So while granularity is a challenge, agility is also a problem. In traditional arrays, customers have to deal with a LUN-centric approach, which requires them to manually map VMs to specific LUNs so they can be snapped. And, this is just part of the complexity. Operating on the snapshots is similarly complex. The appropriate LUN has to be discovered, mounted and then the desired object has to be discovered by manual browsing in the LUN before any action can take place. This is time consuming and not suited for today’s dynamic virtualized environments. While some HCI systems provide the ability to group VMs and set snapshot and retention policies, the application of policy does not evolve as new VMs are added to the environment—which happens literally every day in most environments.

In addition, applications require groups of VMs to be snapped at a consistent point in time. LUNs are internally consistent but have all the usability challenges described above. With some HCI systems, not all VM groups can be snapped consistently. Even if this is possible, a particular VM can only be a part of one consistency group. These considerations and limitations can affect recovery and ultimately availability.

What customers need is a more granular, automated data management solution.

Introducing Datrium Scale-Out Backup: Effortless Data Management

 

Fast & Efficient: Datrium Scale-Out Backup provides customers the ability to snapshot (point-in-time copies) and operate on the snapshots at a finer-grained granularity. Customers can snapshot or operate not only at a VM-level but also at a datastore-file level. Any file that is on the Datrium DVX datastore (e.g. virtual disks, VM templates, ISOs, OVAs, etc.) can be individually snapped and operated upon. This eliminates the capacity-waste of retaining unneeded objects in snapshots, and speeds action with follow-on operations on snapshots.

Datastore-file granularity can be useful in several use-cases:

1. Snapshot a specific set of virtual disks versus all virtual disks, which saves space.
2. Restore a specific virtual disk from a snapshot to a live VM, which improves RTO and reduces admin time.
3. Clone a specific virtual disk from a snapshot and attach to a live VM for dev/test or analytics/reporting use-cases, improving time spend on software development versus infrastructure.
4. Mass clone an OS virtual disk for fast & efficient provisioning of VMs.

Snapshots, clones and restores leverage the deduplicated, compressed and erasure-coded core provided by DVX system for always-on efficiency. Any of the three operations do not result in any physical copying of data in the system, rather only meta-data updates are performed. This speeds up these operations significantly, making them near instantaneous. Furthermore, a redirect-on-write implementation eliminates any performance impact to the original object, ensures that snapshots, clones and restores can be performed at scale without an adverse effect on production operations.


Effortless & Predictable: When dealing with 1000s of VMs (and even more virtual disks), an easy to use policy-based construct is critical. Scale-Out Backup comes with Protection Groups (PGs), which enables users to create dynamic groupings of any VM or datastore-files across the entire DVX cluster with a few simple clicks. VM or datastore-file membership to PGs can be bulk-defined using name wildcards (eg. SQL* or *Exchange*), or can be selected individually. With wildcards, PGs perform dynamic late binding, which allows new objects (VMs or datastore-files) to automatically be added to the PG based on the name pattern and inherit related policies. After membership definition, policy definitions for snapshot frequency, retention and replication can be applied to all objects within the PG. 

PGs also provide an easy construct for performing operations to a group of VMs and/or datastore-files. Once PG snapshots are taken, clone or restore operations can be performed at the PG-level, individual VM-level or individual datastore-file level.

Scale-Out Backup supports both application- and crash- consistent snapshots to ensure the level of state consistency required for applications. In addition, each snapshot can serve as a standalone backup, since there is no dependency of between snapshots or with the snapped object. This allows users to specify different retention periods for snapshots across local or target site (if replicated). Snapshots can be retained for as long as necessary and can be accessed in an easy fashion from Snapstore, even if the original object no longer exists. Snapstore, effectively a backup-class catalog, provides a simple, search-based construct for finding and selecting the specific object that needs to be operated upon.

 

The Bottom Line

Datrium Scale-Out Backup eliminates a variety of challenges that prevent customers from taking greater advanatage of copy data management opportunities to derive business value. By combining fine granularity with dynamic policy-based management, Scale-Out Backup empowers customers to create and utilize point-in-time copies of data in the most efficient, fast and effective manner. And, with smart search and dynamic policies, customers get the full spectrum of capabilities for managing data at scale not just for protection but also for dev/test, analytics/reporting and compliance use-cases.

Download the Scale-Out Backup Datasheet