We usually think of revolutions as events that occur overnight and change everything radically. But there are many more revolutions that go unnoticed as they slowly transform a market, a technological eco-system or a paradigm.
Look at the way that virtualisation has transformed the server market. VMware, a software company, overcame a number of obstacles, the biggest ones being initial performance and scaling limitations.
Today however, virtualisation is the fundamental building block for all Cloud infrastructures, private, public and hybrid.
Similar transformations are taking place at different levels in the data centre today but no area has been talked about more than storage. For years, hardware had the upper hand, but now it looks as if software is finally drawing level
Hardware vs software – what’s on top?
It’s wrong to look at hardware and software as if they were in a struggle for supremacy. Some of the biggest IT companies on the planet are software and hardware companies alike and software and hardware are symbiotic.
But innovations work in cycles and sometimes a new development on one side makes an innovation possible on the other side. For example: the first iPhone was neither the first smartphone with a touchscreen, nor was it very good at what it did at the time. Apple just managed to do what others failed to do before the iPhone – it added a user-friendly interface, essentially software, to standard hardware.
Together, they changed the paradigm of how smartphones can or should be used, transforming the entire smartphone sector and almost erasing household brands like Blackberry from the market. Ultimately, Apple achieved a healthy balance between innovations on the software and hardware side.
This healthy balance has been lost in the storage market. For years, the only real innovations made by legacy storage vendors involved finding creative ways to charge customers.
Customers could only chose between a range of vendors operating with the same business model that bundled hardware and software in a proprietary way and locked the customer in, over-charging for purchase, service and upgrade. Faulty hard-disk? Ka-ching! Added feature? Ka-ching! Fork-lift upgrade after three years? Kaaaaa-ching!
This business model was not suited to providing what customers really needed: an affordable way to keep their data safe. For years, the innovations of legacy vendors concentrated only on the hardware side, producing faster, bigger and more expensive boxes that needed to be sold quickly.
The real innovation is in the software
It is no secret that the amount of data generated and stored is exploding. As much as 90 per cent of the world's data has been generated in the last two years and further big trends, such as the Internet of Things, Social Media and Mobility will increase the growth even more.
Data centre’s are grappling with the results of this growth even as managers struggle to make do with flat or decreasing IT budgets. One solution to the data deluge comes with the arrival of Software-Defined Storage (SDS).
The principle is easy: rather than buying a proprietary product that bundles hardware and software, customers can choose the best hardware and software and combine them into the most effective product for their needs. SDS gives them the ability to choose less expensive standard hardware and use it with enterprise-level software.
Differentiation of the underlying hardware in storage systems sold by large legacy vendors has also eroded to the point of extinction – storage servers basically come off the shelf and the big vendors have just rebranded standard servers with fancy stickers and face plates.
But on the software side, SDS has proven to be at least equal to proprietary systems and, in many ways, superior. 2014 promises to be very significant for customers as software becomes more important than hardware and provides them with the opportunity to address their critical pain points of scalability, availability, flexibility and cost.
Software-Defined Storage powered infrastructure
As SDS continues to take hold, it will drive fundamental architecture changes in IT infrastructure. The reason is simple: once IT managers understand they can combine the leading storage software with the best available hardware for their needs, only the most superior solutions on either side will succeed.
Some customers are already using SDS to change how they buy storage capacity, effectively matching for storage what VMware has allowed them to do for years on compute.
These customers standardise on a Software-Defined Storage solution and make hardware vendors compete every time they need new capacity, continuously optimising CAPEX without compromising OPEX. For these customers SDS is a source of flexibility and agility. It removes lock-in and puts them back in control.
SDS is also affecting the emerging set of hyper-converged infrastructure solutions combining compute, network and storage. VMware Virtual SAN is a critical component of their EVO:RAIL hyper-converged appliances with leading hardware vendors. Hyper-converged pods deliver ultimate deployment and management simplicity for specific workloads such as Virtual Desktop.
Last but not least, SDS is proving instrumental to the emergence of next generation cloud and big data infrastructure. OpenStack clouds seldom get deployed on legacy storage hardware. Instead, customers favor scale out shared nothing architectures deployed on industry standard x86 servers with a mix of low cost SSDs and HDDs delivering object, block and file services at enormous scale.
Storage is the foundation of any datacentre infrastructure deployment and Software-Defined Storage is fundamentally transforming that foundation, enabling new, better, more flexible, scalable and cost effective data centre infrastructure.
With SDS changing the economics of storage for legacy enterprise applications, SDS enabling simpler hyper-converged infrastructure for entry-level workloads, or SDS providing the webscale foundation to next generation OpenStack clouds, storage has never been more exciting, or more important, as an IT segment.