The last week has been big in the world of Software-Defined Storage. VMware has launched the final pillar of the Software Defined Enterprise, Virtual SAN, and just the day after Nutanix announced that it had been granted the patent for its ground breaking software defined storage technology. The VMware launch completes its vision of the Software-Defined Datacenter and Enterprise with all of the major pillars in place, and at the same time intends to shake up the existing SDS market. I have a lot of friends working at VMware and I congratulate them on this major milestone. This launch will broaden the options for customers and keep the other companies in the SDS and hyper-converged field, including the industry leader (where I work) Nutanix, in check. We will all have to innovate and differentiate in different ways, and try to keep ahead and stay focused. Fortunately there is plenty of market out there for us all and the market is still growing. More than anything though this major launch again validates the entire hyper-converged, scale out software-defined storage mixed with compute architecture and the market as a whole. So how do I think this is going to play with the VMware partnership and other storage vendors, including the likes of Nutanix?
Disclaimer: This article represents my thoughts and opinions only and not those of anyone else.
I think Chad Sakac made some excellent points in his article – A Few Thoughts and Opinions on VSAN and Hyperconvergence. The only companies that are worried about what VSAN represents are those companies that can’t adapt to change and can’t move to a new paradigm while at the same time providing support for the legacy architectures. It’s absolutely true that there is no one size fits all. Just like with Mainframes that are still here today, there will be monolithic storage arrays here for quite some time to come and for some use cases. Where compute and storage capacity and performance don’t always balance, or grow at a very dissimilar pace. A prime example is, as Chad pointed out, large file storage for millions upon millions of files and massive amounts of raw capacity where not much compute capacity is required. Significantly though I think those workloads will change over time. I also think if you have the compute growing when the capacity does you’ll find other workloads to take up the compute capacity while the monsters eat the storage capacity.
There is a shift going on in infrastructure, and we can see it, this is why VMware and other companies such as Nutanix are disrupting that market. A lot of workload types work very well in scale out architectures, even current workloads of today (hence the stunning success of Nutanix). But that’s not the big picture. The big picture is the ‘Pivotal’ change in applications architecture that is the driver behind the new infrastructure architectures, such as mobility, big data etc, and this will happen over time. Applications lead and infrastructures generally follow. Anyway, the only reason we have infrastructure is to run the applications, not the other way around. So I hope VMware is successful, because if they are, the rest of the software defined storage and hyperconverged market will be also. Especially companies like Nutanix that have a few years head start with mature products in the market.
So what about the inevitable competition? If I were a dinosaur storage company that couldn’t adapt, I’d be worried. Otherwise I’ll be looking to develop around the use cases of the future that will be successful. When it comes to Nutanix though, there is only one area where VMware competes, and we’re looking at things differently. Nutanix is competing successfully with so many other players already (including vBlock, FlexPod, PureFlex, etc), this is just one more. But in reality there are another dozen storage and hyperconverged startups all in stealth mode at the moment that are also going to be competing and trying to eat all of our lunch. This is just the way of the world. With VMware there will undoubtedly be areas of friction, in VDI and in lower ends of the market. Maybe with this launch, Nutanix and the other players, all providing extremely cost effective and low TCO View solutions, this year really will be the year of the Desktop, EUC, or whatever the latest term is? There are certainly a lot of options that will make it successful.
VMware is not a one product company. They make a lot of great products and the Nutanix architecture is a great fit for everything else, and we cooperate, partner, and integrate with all of those other products. vCloud Suite, vCenter Operations Management Suite, NSX, Horizon Suite, and even some of VMware’s other storage related products etc all run exceptionally well on Nutanix. I’m also seeing a lot of traction with customers migrating from traditional Unix systems over to Nutanix, and their existing physical business critical applications, which is the area that you know I focus on. Unix to Nutanix on vSphere Migrations increases the net virtualizable market for everyone, and there is a massive market opportunity here with a lot of low hanging fruit. So although there is competition in one area, there is cooperation and partnering in a lot of other areas, just like there will be with the other storage partners.
Nutanix is focused on getting from nothing to installed and running infrastructure in the shortest possible time (<30 minutes), in the simplest and most automated way (with Nutanix Foundation), out of the box, with applications running, and all of the storage features that benefit those applications all built in. Running Oracle, SAP, Enterprise Java, Exchange, SQL, Sharepoint, and migrating over traditional Unix workloads as well as everything else, and being the best location for VDI. Over the past couple of weeks I’ve been testing Monster Oracle RAC 12c VM’s on Nutanix and it’s been running superbly well. That is easy to operate, easy to design, easy to scale, and all with a low entry point and low cost unit of scalability, completely avoiding disruptive future forklift upgrades. This is easy to do with a strong foundation. Your datacenter can really be uncompromisingly simple. Don’t need two clicks when no clicks will do.
VMware, vSphere and the whole vCloud Sutie is a great choice for high performance workloads, especially with it’s ability to run high IO VM’s. If the hypervisor was so bad at running high performance storage IO, why would anyone run an Oracle database or high IO monster VM on vSphere? They do, and they are successful, I’ve done many an Oracle database and SAP migration to vSphere and Unix to vSphere migrations. The reasons why are not solely related to performance though, manageability and other aspects are also important.
We are heading into a new era of IT, once where things will change at an even more rapid pace, but also one where uncompromisingly simple is the next big thing. Achieving business results faster, but in a less complex way, that is much easier to manage, and much more scalable and agile.
I will leave you with a link to the TechTarget article on the launch, which contains some great thoughts from the Nutanix CEO Dheeraj Pandey on the impact of VSAN.
Congratulations VMware on a great launch. I wish all my friends at VMware the best success. I look forward to catching up with many of you at events throughout this year and partnering with you closely in the areas where we can mutually benefit everyone.
Disclaimer Reminder: This article represents my thoughts and opinions only and not those of anyone else, no offence is intended.
This post appeared on the Long White Virtual Clouds blog at longwhiteclouds.com, by Michael Webster +. Copyright © 2014 – IT Solutions 2000 Ltd and Michael Webster +. All rights reserved. Not to be reproduced for commercial purposes without written permission.