It seems like every storage company these days is talking about software-defined storage (SDS), and for good reason.

Software-defined storage provides a vast amount of benefits for companies large and small. It gives businesses the capacity to scale their environments at a lower cost and with much more flexibility than alternative solutions. It also allows companies to add new features to existing storage architectures without having to completely replace their hardware. Organizations that deploy SDS solutions benefit from this greater flexibility and choice in hardware selection through simplified storage management and the ability to adjust hardware infrastructure with evolving requirements.

The hype around SDS isn’t appearing out of nowhere. A host of companies are pointing to their technology as software-defined these days, yet most only have a piece of the puzzle or, worse, aren’t software-defined at all. The excitement is building around SDS, but unfortunately so has the confusion.

Since the term was first introduced, it’s been touted by the industry for its ability to reduce storage capital and operational costs. However, despite the laundry list of benefits, industry experts are just as often cited discussing the inherent problems with SDS. Common issues include the lack of cloud economics and complexity leveraging existing infrastructure. While these experts are certainly raising real concerns about the validity of SDS, we must first take a step back to understand the root of these issues; that being said: the promise of the technology will only align with the hype once the industry can agree on a set definition of what it means to be truly “software-defined.”

In order to bridge the gap of understanding these nuances of SDS, we need to agree on a definition once and for all to help us look past the industry jargon to truly get to the heart of the technology and how it can help companies achieve simple and cost-effective data management. Simply put, SDS is an approach to data storage in which the programming that controls software-relate tasks is decoupled from the physical storage hardware. It’s been described by some as virtualized storage with a service management interface. The emphasis is placed on the storage services versus the hardware. At a minimum, SDS must deliver hardware flexibility. To be considered true software-defined storage, a solution must have the following five key attributes:

  1. Abstraction: I/O services should be delivered independent of the underlying hardware, through logical constructs like logical unit numbers (LUNs), volumes, file shares, and repositories. This involves a clear divide of the typical aspects of data storage from the physical hardware.
  2. Transparency: True SDS will give IT teams the ability to monitor and manage their own storage consumption against costs and available resources. Some solutions run as a service, which fundamentally reduces customers’ management and operational involvement to whatever level they want to take for their infrastructure. This can also include a complete white glove-like experience if desired.
  3. Automation and auto-placement: Resources should be consumed using command line interfaces (CLIs) and application program interfaces (APIs) rather than manually allocated through a graphical user interface (GUI). A key example is found in SDS solutions that are managed with restful APIs. These APIs expose the functionality to users and application management frameworks. Built-in, restful APIs allow you to accelerate development time and deliver ease of use and interoperability at all levels.
  4. Policy/service-driven: The service received (IOPS, latency) should be established by policies that implement quality of service, availability, and resiliency. Companies that provide true SDS solutions know that policy is the key to the future of storage. They offer automatic, policy-based data placement to ensure the right performance at the right cost for each byte of data. Policy controls all aspects of true SDS vendors – including how the data is placed in the system, how it is accessed, how the system is monitored, and the overall performance.
  5. Scalability: Solutions should enable performance and capacity scaling independent of I/O latency. Leaders in software-defined storage are able to deliver single, scalable, and simple solutions, all while being able to support new applications that customers want to add at a later date. They offer multiple types of nodes to flexibly align cost and performance across application deployments as well, creating a spot-on scalable solution for customers.

Vendors that have checked these boxes can officially consider their storage solutions to be truly software-defined. And once they’ve mastered these basics, they will be on the way to becoming a leader in the industry.

There are certain features that separate the leaders in SDS from other players in the space. We believe that in order to lead the space, you must take a transformational approach to data management, delivering dramatic reductions in administrative load with automation, cloud-based management, and analytics. We also believe that, for a business to be an SDS leader, it must provide users with a full spectrum of infrastructure support, including multiple types and generations of technology from various vendors. At the same time, it’s also important to have the agility to conduct an immediate massive scale if necessary, because as demand increases, the same infrastructure must be able to scale in capacity and performance at the pace of business. And it’s difficult to be an SDS leader without data access that persists as hardware evolves, given that hardware deployments are far too frequent for any kind of downtime to occur.

Data is growing quickly, and as a result it’s getting more difficult and expensive to store, protect, and migrate. For this reason, the industry is hopeful for the future of SDS and all the benefits it brings – from reduced cost, to its simplicity, and scalability. And it’s up to us as an industry to agree on what SDS truly is in order to reap the most benefits.