Small World Big Data
Published: 19 Jan 2017
Yes, it's a brand new year and time to make some Next Big Thing predictions for the year to come. This year, our outline of what's on the immediate horizon is already well known: hybrid cloud adoption, big data applications and containers. Looking a little further out at enterprise IT trends, we might see the first practical persistent storage-class memory begin to disrupt 30 years of traditionally structured data center infrastructure. And expect a hot smoking internet of things mess of requirements to land in the lap of IT folks everywhere.
All of these topics are, of course, highly interrelated. In fact, it wouldn't surprise me to find that many organizations will have to bite the bullet on all five at the same time to handle a new internet of things (IoT) data processing application. But let’s take a quick look at each:
Cloud adoption. I am as guilty as the next pundit in predicting when cloud adoption will finally be considered a "traditional" deployment model. But this time I really mean it! VMware is demonstrating cross-cloud products. Microsoft is making real hay rolling traditional businesses, large and small, into software as a service, likeOffice365, and infrastructure as a service, like Azure. And all our favorite storage vendors are realizing that building in a cloud tier won't shrink on-premises storage needs given the growth in data and hybrid technologies that balance and marry the best benefits of both cloud and on-premises processing.
Big data. Hadoop is a decade old now. With newer generation platforms like Apache Spark making it easier to deploy and consume big data interactively for SQL-friendly business analysis, real-time operations, machine learning and even graph-based applications, it's time for us all to get on board this train. As I've said, all data can grow up into big data someday. One of the top enterprise IT trends we've noticed is less concern about what big data is and more focus on getting maximum value out of all that data. In fact, I predict that data access -- or data paucity -- will become a new corporate key performance indicator in the future.
Containers. Having predicted the fast rise of containers last year, I claim some victory here against naysayers. Containers have won even if they aren't in production everywhere yet. Yes, there are some major issues yet to be resolved for the regular, not quite DevOps, IT organization. Many apps will never transition to containers -- just like how we will have mainframe applications and VM-based appliances hanging around for decades -- but open the hood of every modern application, appliance, cloud or software-defined infrastructure, and you’ll likely find containers. In fact, most of the newest enterprise IT trends covered above – especially cloud and big data -- are internally powered by container-based development and deployment.
Storage-class memory. Memory-speed persistent storage at an affordable price is inevitable. But is this finally its year? It's looking more likely that we'll start to see some availability at significant volumes by the end of 2017 -- at least at rates high enough to start disruptive arguments about what will become the next best infrastructure design and how applications should evolve to make full use of persistent memory. We may even begin debating whether completely new computing paradigms will emerge.
Internet of things. IoT has already crept into my house in the form of Tile tags and other connected devices, and I've written recently about how IoT could impact the data center. If you think about smartphones as gadgets, cars as connected devices, and even crockpots as networked things, you can easily imagine that discussions about IoT concerns -- security, data protection, data access, data use ethics, distributed processing, scalable command and control, big data flows and far more -- will occupy a great deal of the average IT data center architect's time in 2017.
If you haven't yet spent any serious time envisioning how, for example, big data analytics -- now a decade old -- could solve some tough problems, provide competitive business insight or present new opportunities in the future of cloud computing, you are probably missing out. "Follow the leader" was a tried and true strategy when a new technology maturity and adoption curve was slow enough to get on board somewhere in the middle and still benefit. Today, however, that curve has gotten a lot shorter. In fact, let me predict that, like the disappearing middle class, there soon will only be early adopters and those too late to the game to reap the benefits.
Look into the future with these cloud predictions for 2017
Cut your cloud spending before it's too late
Evaluate the best hybrid cloud deployment tips of the year