Docker truly is a wondrous thing. Not so much the technology (although I have to admit that containers are very cool) but more the ecosystem that it has built up in an incredibly short space of time. On a weekly basis I get an email from a vendor launching themselves into the Docker ecosystem. Of course there is a nagging doubt at the back of my mind that just maybe we’re creating something of a bubble. In this respect it pays to look at the OpenStack ecosystem – a few years ago it had a similar level of frothiness with vendors sprouting up like mushrooms. Unfortunately there has been recent carnage and many of those vendors have come to realize that there simply isn’t enough of a market opportunity for all of them.
Docker strikes me as somewhat different however since most of the initiatives I hear about are tackling different parts of the breadth of functionality needed for Docker to be truly game changing. Of course many of these companies are creating a feature and may well be a target for consolidation in the future – but at least they’re not doing exactly the same as what every other vendor is doing – as was the case with many of the OpenStack vendors.
Today’s example of the growing ecosystem is StackEngine, a company coming out of stealth today to announce its Docker operations product. The company, which recently raised $1 million in seed funding from Silverton Partners and LiveOak Venture Partners, wants to help users deploy, manage and scale their Docker-based applications. Hmmm, where have I heard that before? Is StackEngine the start of the “me too” Docker vendors? StackEngine is currently in private alpha and is planning a GA sometime in Q4.
StackEngine believes that there aren’t the tools that are really needed to run Docker applications. The feel that traditional tools such as Chef and Puppet are systems-automation tools, rather than container automation ones. They are too complex to solve container problems. StackEngine is looking to help operations teams bootstrap Docker into production and ticks off the discovery, visualization and management of Docker deployments.
The StackEngine founders certainly know from their infrastructure management. Bob Quillin and Eric Anderson, formerly of CopperEgg, Hyper9 and VMware, bring more than 30 years collective experience in virtualization and management technologies. Given that fact I was keen to quiz them on where they saw the differentiation lie between StackEngine and the other Docker-management tools. According to the pair:
The StackEngine approach differs from many of the developer-centric or PaaS-inspired solutions that have dominated the early days of Docker. After spending time with many early Docker adopters, [we] saw that the true bottleneck in most organizations was on the operations and devops side versus the developer side. Developers were finding plenty of early tools from the likes of CoreOS (base OS, service discovery, networking), templates and best practices from people like CenturyLink (Panamax), and cluster mgmt scheduling tools from Google Google (Kubernetes) and Mesos. But ops teams had little to work with to manage these apps into production and operate them on a day-to-day basis. But if Docker strives to be the next VMware VMware, they lack the vCenter ecosystem of automation tools for operations that help you discover, monitor, configure, manage change, orchestrate, and adapt containers across your environment whether that’s on-premise or cloud or hybrid. StackEngine’s goal is to bring devops and operations closer into the container community with tools solutions designed for them – not tools intended to design them out.
I agree that Docker is lacking the robust vCenter-esque management platforms, but I suspect that is absolutely the vision that many of those aforementioned vendors have. Whether StackEngine can be a player in building out that platform will be seen in the months and years to come.