And there’s your quantitative evidence that SDS is real.
Joke aside, I sat through session after packed session on the future of storage and Software-Defined Storage’s place in that future during this year’s Gartner Data Center, Infrastructure, & Operations Management Conference (DCIM for short).
For some leading edge companies, the relevance of SDS is crystal clear. Dailymotion eliminated their capacity scaling challenges without throwing tons of appliances at the problem. RTL II addressed performance at scale. Other new Scality customers in Manufacturing and Healthcare were finally able to address data durability at petabyte scale.
But Gartner reminded me that many mainstream businesses are looking at SDS with this lens:
It’s still hype to them.
But in the room at the DCIM show, more than 30% of participants across multiple sessions will be taking a hard look at SDS in the next year, and 50% within the next two years.
Why are storage customers looking? Do they even know what SDS is?
They’re looking because the majority of storage customers are somewhat dissatisfied, or very dissatisfied with the current TCO of storage. They’re looking because “web-scale” architectures are becoming increasingly relevant to the modern data center, and current storage doesn’t support them. They’re looking because infrastructure flexibility and agility are increasingly critical in an era of rapidly changing business models.
Storage customers may have limited, differing, or rough definitions of SDS, but that isn’t stopping them from considering it.
Gartner continues to refine their own SDS definition:
SDS leverages software that separates and abstracts storage capabilities dynamically derived from the physical and/or virtual storage devices, regardless of location or class of storage to improve agility and deliver quality of service (QoS) while optimizing or controlling costs.
Gartner struggles a bit with the different representations of SDS that are out there. They get confused by vendors who claim that SDS should virtualize and manage current storage appliances, when that notion is antithetical to both the notion of “improving agility” and a significant amount of the “cost control” that comes from shifting to industry standard x86 servers. On one hand, they urged caution with adoption of SDS. On the other, they predicted that:
“By 2016, server-based storage solutions will lower storage hardware costs by 50% or more.” In his session on the “Economics of SDS”, Roger Cox of Gartner proceeded to back that up with multiple TCO models showing cost savings ranging from 50% to almost 70%
In the end, there were valid points to be considered from multiple analysts. Are there feature sets that need more maturity? Yes. Do applications, processes, and procedures need to be considered before SDS adoption? Certainly. Is there inertia to overcome? There always is, but the benefits are too great to ignore.
If you haven’t at least considered SDS, now is the time. And this is coming from the most mainstream of technology analysts and consultants.