Disk Backup Appliances with Data Deduplication Need to Keep Pace

ExaGrid Briefing Note

The performance and capacity capabilities of production storage are increasing to meet the demands of the virtualized data center. These systems can now scale to support highly dense virtual machine populations and the never-ending demand for more storage capacity. Data protection software is keeping pace with features like changed block backups and in-place recovery, but for these new features to be fully realized requires that the data protection storage also keeps pace. Many are woefully falling behind.

ExaGrid’s unique landing zone stores the most recent backups in their full un-deduplicated form for restore, recovery, and VM boot performance. This landing zone technique is an example of a backup storage solution that is keeping pace. Recently ExaGrid announced the EX40000E that enables it to take these capabilities from its traditional mid-market customer into the enterprise.

The Disk Backup Appliance Problem

Most disk backup appliances seem to solve the same problem: storing lots of backups with the same technology; deduplication. The problem is that most of these devices use inline deduplication and do not scale beyond the original chassis. The problem with inline deduplication is that it impacts ingest performance, and due to storing only deduplicated data impacts large recovery performance, and VM performance if the device is used to boot virtual machines through an instant recovery feature.

ExaGrid’s landing zone feature makes it immune to these problems as all three of these interactions occur within a non-deduplicated area of the appliance. The landing zone area works under the premise that most restores, at least those where performance matters, come from the most recent backup so it makes sense to keep recent backups at-the-ready for optimal restore performance. All backups are deduplicated for optimal long-term storage.

The problem with the lack of scale is that as these disk appliances reach their performance or capacity limitations, the data center is forced to purchase additional appliances that need to be managed individually, reducing administrator efficiency. A scale-out design overcomes these limitations by clustering the backup appliances into a single logical system. When the data center needs more performance or capacity, it just adds another node to the cluster.

The IT planner needs to be careful to make sure the scale-out strategy makes sense. Backup storage typically has a 2X longer service life than production storage. The scale-out design needs to be flexible enough so it can support a variety of different node types as technology and organizational needs change. ExaGrid uses a loosely coupled cluster that allows new nodes to be inter-mixed with nodes of any age, providing the ultimate in investment protection.

Enterprise Meet ExaGrid

ExaGrid is dominant in the mid-tier data center. But its landing zone’s ability to empower faster backups and recoveries, as well as better enable instant VM recovery, certainly appeals to the larger enterprise. With the EX40000E, ExaGrid clearly has its eyes set on going upmarket. The EX40000E is sized to effectively store a 40TB backup data set (fulls plus reasonable incrementals). Twenty-five of them can co-exist in a cluster or “GRID”, enabling protection of a 1PB data set. Current ExaGrid customers can seamlessly add the EX40000E to their existing GRIDs.

Capacity is not the only enterprise aspect of the EX40000E; its ingest rate (how quickly it can receive backup data) is 8TB per hour, which means a GRID of 25 should achieve a 200TB per hour transfer rate. Restore speeds should be equally impressive since there is no overhead, in most cases, of rehydrating the deduplicated data.

StorageSwiss Take

ExaGrid is one of those companies with the right technology at the right time. The landing zone feature seems tailor made to solve both backup and recovery performance issues. Their choice in architecture is ideal for the use case. Backup data is hard and expensive to migrate so the ability to have a scale-out architecture that allows for different node types is ideal for a storage infrastructure that may last 3X or longer than the production storage it supports. The EX40000E positions ExaGrid to not only take care of its current customers’ capacity growth, but it is also the strongest case made to the enterprise yet.

Twelve years ago George Crump founded Storage Switzerland with one simple goal; to educate IT professionals about all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought after public speaker. With over 25 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS and SAN, Virtualization, Cloud and Enterprise Flash. Prior to founding Storage Switzerland he was CTO at one of the nation's largest storage integrators where he was in charge of technology testing, integration and product selection.

Tagged with: , , , , , ,
Posted in Briefing Note

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 22,221 other followers

Blog Stats
%d bloggers like this: