Software-defined storage: data always outlives hardware

Industry analyst house, IDC, defines SDS as:
‘Any storage software stack that can be installed on any commodity resources (x86 hardware, hypervisors, or cloud) and/or off-the-shelf computing hardware and used to offer a full suite of storage services and federation between the underlying persistent data placement resources to enable data mobility of its tenants between these resources.’

Essentially, software-defined storage is a cost-effective method for a business to ensure its data is preserved beyond the lifespan of its physical storage hardware, be that DAS, SAN, NAS, MAID or tape. It commoditises general purpose hardware and uses software to literally ‘define’ the storage solution that is being implemented. Now that we have a workable definition for SDS, we can examine what the drivers and benefits SDS are compared to more traditional based storage.

Data sprawl dilemma
CIOs rank data growth as their top concern, and making sense of that data as their number one priority, according to a 2012 Gartner report on the CIO Agenda. In addition, over 60 percent of Fortune 1,000 storage professionals cited managing storage growth as their main pain point over the coming years.

The problem is that as data grows, a company’s existing primary storage system gets full and more storage hardware needs to be added, which can potentially create data silos - different performance classes, or ages of hardware that become disconnected islands of data. With the exponential growth in the amount of data, IT must either continue to add high cost primary storage, or chose which data to archive, and decide which users to impact. However, the majority of corporate users often need their data to be online and accessible at all times. As a result of this approach, a lot of companies find that their data became difficult to administer, find and protect and therefore expensive to keep online.

SDS to the rescue: A cross-platform solution
Storage systems have traditionally been defined by dedicated platforms, with features mainly delivered via single purpose hardware. As a result of this approach, once a system reached capacity or maxed out on performance previously, another had to be rolled in alongside – and managed entirely separately. IT was then faced with the prospect of implementing a gruelling upgrade and migration process and on occasion, would be forced to do both.

At the same time conventional archives could be disconnected from day-to-day workflows and often confused with backup. In some instances, this would lead to IT staff being required to retrieve files no longer online as users demanded. The resulting infrastructure sprawl increases costs and causes significant management challenges where IT admin intervention is needed for silo load balancing and archive access. In an ideal world, storage solutions should accommodate next-gen technologies in an open architecture with little or no impact to user access and this is where SDS is vital.

A software-defined storage solution takes the approach of taking commoditised hardware that is tailored to its storage requirement by software.
It allows much greater flexibility in architecture, and reduces costs. In one example, leveraging this concept delivers virtualised data management to lower the cost of high volume storage. In this way, the software-defined storage solution’s biggest benefit lies in its heterogeneity. Products such as SGI’s InfiniteStorage Gateway allow businesses to align data with appropriate storage over its lifecycle, whilst maintaining seamless data access and enabling the ability to connect to heterogeneous storage environments.

It is this ability to take general purpose hardware and turn it into a data management platform using software that provides the most value and sets SDS apart from traditional storage. This means that businesses need not be tied into lengthy vendor-centric contracts that limit their ability to be more agile when dealing with the data sprawl. Abstracting the storage from the hardware means that it can become a shared pool, without the constraints of physical systems, which creates the most efficient utilisation of resources.

Through the long-term lens
There was a certain amount of short-termism in how businesses procured storage hardware in the past, largely due to the technology not having caught up to the challenges imposed by the data explosion. Previously, IT buyers would procure an entire computing system in the knowledge that the hardware would be obsolete in three to five years; therefore need replacing, upgrading or both. Using SDS, businesses are freed from this short-termist trap and can embrace a longer-term view of their storage requirements in the knowledge that their data will not outlive the hardware it resides on.

The data sprawl problems that have been faced by SGI customers in the life-science and digital media space for over 20 years are now becoming commonplace in general IT. As this explosion in corporate data gains momentum, it will continue to put a strain on companies’ existing storage environments. If left unchecked, this can result in fragmented infrastructures and storage silos where expensive disk is being used to hold inactive data. By turning to SDS, a business can ensure that software becomes the most important aspect of the storage eco-system, rather than the hardware, giving IT the ability to work across vendor platforms, which will ultimately free up resources.