– Wayne Salpietro, director of marketing at Permabit, says:

In today’s information technology world, it is time to adopt deduplication across the entire IT spectrum from applications to operating environments across all of the storage media (flash, SSD, HDD) and implementations (primary, 2nd tiers, archiving and the cloud) and, last but not least, in all virtual environments. You might be asking “why now?” There are several trends that are intersecting today that make it a necessity. Let’s review them:

Data growth is rampant, creating a data deluge that is nearly impossible to afford and manage.

Information growth is yielding larger and larger data stores which have spawned a new phrase “Big Data” and big data is being used by analytics engines to identify trends that may lead to differentiation and incremental revenue.

Storage transformation is occurring as solid state devices rapidly gain traction because their performance makes them a must have addition to existing spinning disk storage. As solid state technology (disk, flash and cache) closes the “cost gap” with spinning disk, the performance gains will be so compelling that rampant adoption will occur and storage will be changed forever!

Virtualization is delivering more optimization of user functionality and ease-of-use that it too is being adopted across enterprises to deliver information transparently enabling faster –better – more efficient customer interactions and increased revenue generation.

Consumerization of technology is being added to an increasing virtual desktop infrastructure (VDI) adoption as “once personal devices” are being integrated into the enterprise to harness worker cycles 24/7. VDI delivers access to business information that is seamless for the worker enabling work anywhere/anytime in our 24/7 business world.

So how does deduplication impact all of these you might ask? Let’s take a look:

  • Data growth is all about scale and efficiency. We are saving huge amounts of data, but can we afford it? With data deduplication it can be reduced by 5-35X making it very affordable considering the alternative.
  • Transforming storage technology is about the fact that storage has not kept up with the advances in processors creating a performance gap that slows down overall worker and business efficiency. Solid state devices (SSDs) have a tremendous performance edge over existing spinning disk technology. However, they are hampered in their adoption because their price on a per GB basis is still too high compared to spinning disk. Deduplication closes that gap because it reduces the data stored on SSDs yielding a much better “effective cost.” In addition, deduplication addresses the other Achilles heel of SSD’s in that the less data written to the device, the longer the devices life cycle becomes.
  • Virtualization’s promise of efficiency can be optimized by deduplication. From a virtual server perspective dedupe of the server images can save both in the storage of the images, but also in the network utilization. From a storage virtualization view with deduplication data need only be deduped one time. As data’s importance evolves from hot to cool to cold, data is moved from expensive (SSD) to less costly hybrid storage and eventually to HDD and possibly tape. With the right deduplication technology, it never needs to be rehydrated between storage venues across the entire virtual storage environment. That will establish a virtual storage pool that will be more efficient (smaller) and more responsive to users as well reducing OPEX.
  • Virtual desktop infrastructure (VDI) basically creates a virtual “work” desktop that can be accessed from any device. The user sees their applications, and has access to their data no matter what portal they use. Deduplication optimizes that VDI creation at the server and the network access to the portal device. Dedupe saves the storage at the server and the network consumption by 20-35X in many cases.

Adoption of deduplication is a necessity as we transform information technology. Without it we cannot afford the data, access the trends that enable business differentiation or deliver access to data seamlessly on any device in the office, on the road or in your home.

There is one “gotcha” however, not just any deduplication can do this. Dedupe initially designed for backup isn’t even close, it was built just for backup. Deduplication that is needed to do this job must perform at I/O speeds and scale to multiple petabytes that first generation dedupe cannot meet. The dedupe engine that is needed to be broadly adopted across the IT landscape needs to deliver “performance at scale with efficiency.” Simply said, this next-gen dedupe engine must deliver:

  • Performance that meets today’s I/O requirements and does not incur latency 
  • Scalability that can scale to meet today’s high data growth environments 
  • Resource efficiency (RAM and Processors) which is critical to enable scale and performance 
All three of these – performance, scalability and resource efficiency – are needed to deploy a deduplication engine that can address the needs of business today and into the future. With increased data efficiency that deduplication can deliver, IT efficiency will improve markedly and an IT transformation will occur because costs will be reduced, data growth will be more affordable and will be in line with today’s IT budget constraints.