Talk about software-defined anything and everything has been a hot topic these past couple of years. Whenever I can spare a few cycles, I brew a hot cup of Earl Grey tea and set out to catch up on the software-defined news. After digesting all that I’ve heard and read, I’ve concluded that beyond the typical high-tech hype, there is something fundamentally right about an infrastructure that uses software as its secret sauce to change faster and keep up with the staggering pace of business.
Why is it that IT infrastructure needs to be so agile and flexible? Business isn’t a one-size-fits-all proposition – think finance, insurance, telecommunications, retail, education, government – these all have unique demands. Likewise, the IT applications and workloads that power these businesses aren’t completely alike either and they’re not static. Applications and data continually change and grow according to the latest business initiatives.
In my research, the giant webscale companies are often held up as prime examples of software-defined infrastructure. They’ve exhibited some pretty impressive results using stripped down architectures where software defines and then redefines infrastructure personality. Everyone in IT infrastructure can learn something from what these companies are doing. So does that mean all infrastructure should emulate the big webscale do-it-yourself giants?
Well in theory, yes; but in approach, probably not exactly. Something I hear much less about is the differences between giant webscale companies and enterprises. Analysts have said that webscale companies typically have 1 to 4 major workloads. Large enterprises may run more than 100 applications, both legacy and next generation. Webscale companies have invested in deep vertical integration and in-house expertise in infrastructure and have massive economies of scale. Enterprises need to focus first and foremost on their core business, have smaller economies of scale and most are just getting started with DevOps.
What I like about working at Hitachi is we recognize these fundamental differences and are prescribing an approach that is software-defined, yet application-led. This means that first enterprises need to think about the application being deployed and how it helps run their business; next, how software can make it run better, easier and faster. There are three major themes when designing a software-defined infrastructure: Automation for simplicity, Access to more data for insight and Abstraction for greater agility. This might include automating and virtualizing legacy applications or using new scale-out architectures for greenfield applications where the intelligence is completely in the software. I anticipate Hitachi will deliver on these three “As” by putting its own twist on webscale-style infrastructure and by extending its more traditional infrastructure solutions through software.
Right now, my Earl Grey is cold and it’s time to pour a fresh cup and get back to work. But stay tuned, because something is definitely brewing here at Hitachi and announcements that fill Hitachi’s prescription for software-defined infrastructure are expected soon. --Paula