RackHD: An Open Source Solution for a Hyper-Scale Problem

You might recall that we took our first major step into open source back at EMC World, rolling out the CoprHD project – derived directly from EMC’s ViPR Controller product. At the time, I said that this was just the beginning of what you would see from EMC on this front.

Today, I’m delighted to announce that we’ve now got our second major open source project out there. What is it? RackHD.

What problem does RackHD solve? Well, have you ever experienced the pain of updating firmware and BIOS or installing a low-level OS on a bunch of hosts? In a 2nd platform world, this problem is frustratingly manual, but manageable. But when you consider the scale of the 3rd platform, along with the increased variety of the commodity hardware that underpins it, the problem becomes much worse – and at hyper-scale, it becomes impossible.

Today’s hyper-scale players have solved this through highly constrained hardware specs or by implementing “bare metal as a service” standards – but these options don’t really work for the rest of us. And I do say “us” because we face this problem ourselves. We want to get ahead of the problem. So we’ve created the RackHD open source project.

What’s RackHD do? It’s a physical hardware Management & Orchestration (M&O) layer that automates discovery, description, provisioning and programming on a broad range of hardware – servers, switches and storage. What’s that mean? It means it becomes very easy to update firmware and BIOS and install software components like vSphere, KVM, ScaleIO, and CoreOS – for a data center of any size… most critically those at hyper-scale. It’s secure, platform agnostic, and programmable via APIs. Does this step on the toes of higher-level infrastructure M&O or software like Puppet, Chef, and Ansible? Nope. But it can certainly integrate or even be incorporated by them.

For EMC, RackHD is already playing a key role in things like VxRack, which of course runs on industry standard servers, enclosures and switches at massive scale. Virtustream and Pivotal are making use of it too. But we don’t expect this just to be an EMC thing. It’s an industry-wide challenge. We know our partners feel this pain, as do many of our large customers and service provider partners. Our open sourcing of the project and their participation will deliver more features, faster, without lock-in.

So where can you find it? In the RackHD community on GitHub. There you’ll find the bits, the docs, and the experts. Don’t hesitate: click on over, collaborate, contribute and yes, criticize, so together we can make RackHD better.

For more on RackHD, check out the official press release and a Pulse blog post by John Roese, EMC Senior Vice President & Chief Technology Officer.

About the Author: Jeremy Burton