What's in Store for Converged and Hyperconverged Systems in 2017? (Video)

hyper-converged_racks-1.jpgWe saw strong adoption of both converged and hyerconverged systems in 2016—and we expect that to continue into 2017. While converged systems have been out for a while and have strong market penetration, expect hyperconverged to come on strong this year as these systems continue to mature for tier-1 workload support. The strong performance of the Nutanix IPO certainly drove some attention on the hyperconverged side, as did the recent HPE buyout of SimpliVity

Of course, these systems don't exist in a vacuum, so there are some things to think about when it comes to the hypervisor environment, systems management, and networking. Those issues will get a lot more attention in 2017, and be key discussion points. So, watch this video for my thoughts on 2016 and 2017, but also please check out my Q & A with Edwin Yuen (systems management) and Dan Conde (cloud and networking) in the Cloud Computing Predictions Video - Part 3. While you are at it, check out their Q & A videos too!! Dan interviews Edwin and me in Part 1, and Edwin interviews Dan and me in Part 2!

Video Transcript

Wow, 2016 was certainly an interesting year for converged and hyper-converged systems. We saw significant interest in using these types of systems. In fact, in research we published in early 2016, we saw that more than 80% of the IT managers we surveyed have already adopted, or plan to adopt, converged and/or hyper-converged systems at least somewhere within their environments.

There's a lot of reasons for this traction, but the big driver is that IT organizations are really looking to attain from their on-premises infrastructures the same levels of speed, agility, and simplicity of management that they can get from the cloud. Now, that doesn't mean the cloud goes away. It certainly has a role. In fact, many IT organizations are looking to build hybrid environments and use the cloud when and where it makes sense for their applications and workloads.

And by the way, this doesn't mean that legacy infrastructure approaches go away either. And by legacy, I mean traditional, 3-tier server storage network approaches to building IT. They have a place too. In fact, in that same survey where we saw that 80-plus percent of IT orgs are using or planning to use converged or hyper-converged platforms, we also saw that more than half of those organizations expect to be running primarily in a legacy fashion five years out, because it can't just rip out and replace what you have. It just doesn't work. There's a lot of legacy infrastructure out there that runs great, and it does the job it was installed to do. But given that there's so much legacy-installed equipment out there, seeing that nearly one in three plan to use converged and nearly 10% hyper-converged as their primary approaches to IT five years out, that's the beginning of a major shift underway.

And we're gonna be asking that question again this year, and I do expect to see some significant growth in the plans five years out for both converged and hyper-converged, driven by this need for simplicity and agility and elasticity with on-prem resources. Now, converged platforms, they've been out there for a while, and there are thousands of them out there. It's important you understand, the big differentiator between converged and hyper-converged systems is that converged systems, they're really more of a hardware packaging exercise, with a single support contact and a management software overlay. And hyper-converged systems are really software defined, and they take away much of the hardware management that you have to do in traditional and converged systems, especially on the storage end of things.

I think in 2017, you're gonna see the lines begin to blur between converged and hyper-converged systems as converged management software matures and virtualizes more of the hardware stack. And hyper-converged systems provide more hardware isolation functionality to meet tier-1 performance demands. It's important to talk about tier-1 performance demands, because there's a perception that hyper-converged systems are more suited to tier-2 workloads and converged systems to tier-1 workloads, largely because you can better isolate workloads in converged systems. But I think you'll see converged solutions continue to mature and offer feature sets to make IT more comfortable supporting tier-1 applications. Much of the IT market is growing at a very slow pace, in fact, sub-2% growth. Hyper-converged is growing very, very quickly, well over 100%. In fact, approaching 150% year over year.

As I said, it's a newer market entrance, so that lots of large numbers play a big role here. Hyper-converged is starting from a very install base, so fast growth is easier to attain, but it's really much harder to maintain. Despite that, expect some hyper-converged platforms to continue to see fast adoption, because feature sets are maturing to allow them to address broader workloads. That adoption, it's gonna reveal some under-discussed issues with hyper-convergence. For example, at scale, managing many small appliances can be challenging, so hyper-converged vendors will need to address this. And also at scale, networking can be an issue as hyper-converged solutions take all of the north-south traffic and move it to the main communications network. And last but not least, the IT teams that I talk to really are looking to build hybrid environments, but they want management tools that spanned on an off-prem worlds to get true workload portability. And that's going to be an issue we need to continue to monitor. Fortunately, I work with a really smart analyst here at ESG, and I had the chance to talk to some of our other cloud analysts about these issues. Edwin Yuen, who covers cloud and systems management, and Dan Conde, who covers cloud and networking, and I asked them about these networking workload portability issues. So please, take a look at the other video we've posted here to hear their thoughts on that.

Topics: Converged Infrastructure