I think everyone knows by now that I am a huge fan of HyperConverged Infrastructure. The biggest reason for that is that I love per-node scalability. I still love the way that EqualLogic has taught me to design infrastructures for its current use and only scale later (read: Storage That Grows Cheaper). There are two reasons for postponing acquisition for future use:

• you don’t know today what your demands will be in 2 years from now 
• you don’t know today what the technology will be 2 years from now

In the last two years we have seen a huge upcoming of HyperConverged architectures. The leader both from a technology and market share perspective is Nutanix and their biggest challenger today would be SimpliVity. We are in a buyers market here so both companies have a huge potential going forward. And then there are some smaller challengers in the space like Pivot3 and even the ‘build your own datacenter’ players like Maxta and VMware VSAN now play in that same market segment.

 

Scale Computing

 

Scale Computing is the odd one out here. Where all the players in the market (leaving VMware aside for a moment) have chosen a VSA type architecture, Scale Computing has chosen to build a kernel based storage and virtualisation cluster. As a platform they chose KVM. If I look around I don’t hear that many people talking about KVM today other than some Linux freaks (sorry for the offence) so why would a vendor go in the SME market with something nobody is talking about?

 

The trickery is in the User Interface. Before I ever worked on ESX I was no more than a Windows Admin / SE. And even after that I still was. I didn’t really have to know the ESX kernel because VMware made that all easy and hidden for me. And still today the only people that need interaction with the kernel are architects or 2nd/3rd line support engineers.

 

And that’s exactly why it shouldn’t matter what kernel is driving your virtualization from an admin perspective. But KVM today is nothing more than a hypervisor engine. If VMware would’t have built a decent UI on top of ESX no one would have used it. And if they wouldn’t have built vCenter to cluster the whole thing it would never have gotten the market share it has today. Scale Computing has built that virtualization and storage cluster for KVM. That’s it, dead simple.

The Project

I’ll put this section in RED as this post is a clear disclaimer that as of this week ScaleComputing is a customer. Building stuff is one thing, talking about it is another 🙂 So Scale Computing reached out to me and asked some help to spread their message. For at least the next 4 weeks I will dedicate my time to Scale, learn about their technology, talk to some of their customers and see if I can create some new collateral that helps them talk to more new customers. I’ll even fly to their HQ in Indianapolis for a few days as well.

It’s going to be really interesting for me. I know one or two things about storage and virtualisation in general but it will be my first connection into KVM. I don’t know today how much I will be sharing on my blog about it but if you want to keep in touch during this project give me a heads-up on twitter or so. I will have a cluster at my disposal for testing purposes as well 😉

Disclaimer: SImpliVity is also a client.