Another EMC World is in the books and as Edward Sharp of AVI Networks proclaimed, these elephants can dance! EMC is starting to attract a new crowd and it was rightfully seen throughout the EMC {code} breakout sessions.

On the first day, Joseph Heck of the RackHD team and myself (Kenny Coleman) presented an introductory session titled “Automating your Data Center with RackHD“. The room was at capacity with 120+ people. The energy was palpable and crowds were gathered around the RackHD booth all week wanting to dive deeper. People were eagerly awaiting to learn more about this open source technology that is creating quite the stir.

The major take-away from our session is that physical infrastructure can be consumed from higher powers such as Kubernetes, Cloud Foundry and more. Alleviating the need to have domains of management, which blurs the lines between new and traditional infrastructure.


So ask yourself a question. What do you want your data center to be like when it grows up? It’s a pretty easy assumption that many of us want to run our data centers like AWS, Google or Azure. For a few years now, we’ve been moving more operations to the cloud. We are starting to embrace it. However back in March of 2016, the Wired talked in-depth about Dropbox’s mass exodus from AWS. Apple is doing the same thing. There are lots of factors, such as scale, that play into it but many companies are moving back to on-prem because new tools are developed that run your data center just as efficiently as the big dogs.

The goal is to treat our infrastructure as code. How do we operate infrastructure in a way that is more hands-off? You can’t be like a cloud provider unless you orchestrate at the very lowest levels. We have to be able to treat physical components as if they were virtual machines. The biggest problem to solve is what to do with a server after it gets rolled into the data center and fitted with power and a network connection? What happens after pressing that power button? The goal to accomplish is getting this server operational as fast as possible. That may be to add it to an existing cluster of servers for additional resources or perhaps running a bare-metal application.

RackHD is a technology stack for enabling automated hardware management and orchestration through cohesive APIs. It serves as an abstraction layer between other M&O layers and the underlying physical hardware.

The real take-away is that physical infrastructure provisioning can be consumed and managed by other orchestration tools. This elevates the understanding of the underlying infrastructure to a new layer. It allows tools to start consuming physical infrastructure in the same way that we used to consume virtual machines. Pretty powerful stuff.

During this presentation we examined a few integrations. First is the OpenStack Shovel project. Shovel is a server with a set of APIs that wraps around RackHD/Ironic’s existing APIs allowing users to find bare-metal compute nodes that are dynamically discovered by RackHD and register them with Ironic. Shovel also uses the SEL pollers service in RackHD to monitor compute nodes and logs errors from SEL into the Ironic Database. Also provided is a Shovel Horizon plugin to interface with the Shovel service. The Plugin adds a new Panel to the admin Dashboard called RackHD that displays a table of all the bare-metal systems discovered by RackHD. It also allows the user to see the node catalog in a nice table view, register/unregister node in Ironic, display node SEL and enable/register a failover node.


The second integration is the Bosh RackHD CPI, which enables Bosh to provision and monitor software deployment on bare-metal machines. It would work on any Bosh release, including Cloud Foundry. For years people have been asking for the functionality to deploy Cloud Foundry on bare-metal; the RackHD CPI finally makes this possible. Users and administrators of Cloud Foundry no longer need to waste their efforts setting up IaaS prior to setting up Cloud Foundry.


The third integration is the latest v0.1.0 release of the Docker Machine Driver for RackHD. This part of the presentation featured a hands-on live demo. The demonstration showed how easy it was to create a “docker-ready” node with Docker Machine and we began deploying containers in less than 5 minutes. You know what’s even more awesome? The Day 3 Keynote featuring Rancher Labs deploying hosts using RackHD used this driver to make the magic happen. Be sure to read the in-depth guest blog post Building an enterprise-grade container service with RackHD, Rancher, REX-Ray and ScaleIO Part 1 & Part 2


Screen Shot 2016-05-09 at 8.08.34 PM


The final integration that was talked about is a brand new release by Travis Rhoden of the {code} team who has built an integration with Kubernetes. The script will use a RackHD SKU as a pool of nodes for Kubernetes. This is still very experimental but will be stabilized in the very near future.

Screen Shot 2016-05-09 at 8.19.45 PM

Want to see the presentation? You can check it out below or download it direct from Slideshare.