Software-defined Storage (SDS) is an inescapable term right now...and looks to become an inescapable reality soon. That is, of course, if you buy into it as being something that's brand new and not already all around us. I just posted what I hope is a fun, insightful, and intentionally somewhat provocative ESG Video Capsule on just this contention. Please take a look—as with all our Video Capsules, it's under 140 seconds:
Following on from our recent Next-generation Storage Architectures qualitative research, we just completed a comprehensive, quantitative Storage Trends survey. While the latter is not yet published (it will be soon!.... and there'll be lots more insights from it in this blog in the coming weeks and months), one thing stood out in both pieces of research...and that is the almost universal interest in SDS (whether "interest" is as existing adoption, active planning, or committed investigation).
Is everything about the current SDS offerings perfect? Of course not. Does it replace all "traditional" storage instantly? Of course not. But does it have the potential to genuinely and dramatically change the accepted manifestation of "the storage industry"? Emphatically, yes. It's not a ridiculous stretch to compare the scale of the potential change to what has happened since "software definition" to Uber hitting the taxicab industry. We still get around in cars, but it's way more flexible, scalable, easy, and predictable; and the existing taxi vendors are having to decide whether to fight or join the new reality. Of course, storage managers tend towards the conservative—that's both understandable and smart when you are responsible for data, which is really the only impossible-to-easily-reproduce element within IT, and which for many organizations, is actually their business. Hence, for many, the perceived newness of SDS is itself a reason for caution. But as (a) the underlying truth that modern SDS is mainly an extension and evolution of what we have had for years comes to be realized; and (b) as the big "yellow cab" storage vendors embrace and promote their own versions of "Uber," you can be sure that SDS is the existing wave of the future of storage!
Mark: It's probably another few years before I get to stop talking about software defined storage, SDS, as new. It will be a norm.
With so much emphasis already placed upon it by vendors of all types, sizes, and even backgrounds, the writing is on the wall for a big chunk of the traditional storage systems world. It's fascinating to watch this play out. Users are actually a bit nervous, despite SDS having existed in some form for ages.
Meantime, the big traditional vendors are in a Catch-22 situation. Damned if they do, and damned if they don't. Of course the smart ones realize that the IT industry has never rewarded those vendors that don't move, and so their decisions are more about timing, emphasis, and degree, than about intent and direction.
So to be clear, SDS is a convenient contemporary phrase for something that only appears to be new. The industry has not suddenly embraced a new software first religion. The evangelism might be happening now, but the missionary work has been around for years.
Let's face it. No hardware is more than a lump of silicon, plastic, and metal, without the intelligence to make it useful, meaning that all hardware has essentially been software defined from the start. And most apparently proprietor storage systems today are in reality already X86 servers running some storage software. Sounds like SDS to me, or typically we refer to those elements as controllers running microcode, since that protects both the mystique, and the margins.
The advances that SDS offers in heterogeneity, orchestration, and so on, are just that. Advances are not brand new, even if they've not been common to date. Software defined storage is a new phrase, but that's the only truly new thing about it.