Docker containers: How Flocker's multimillion injection marks the rise of the ecosystem

Flocker firm ClusterHQ points to its series A funding round announced today as proof of the long-term potential of Docker and container technology.
Written by Toby Wolpe, Contributor
Mark Davis: A signal to users and developers.
Image: ClusterHQ

ClusterHQ, the startup behind the open-source Flocker container manager, says a new $12m funding injection sends an important signal about the emergence of the Docker ecosystem.

The investment, led by Accel Partners London, with participation from Canaan Partners and existing investors, will enable the 17-strong, Bristol, UK-based firm to expand the size of its software engineering team and step up efforts in the developer community.

"This chunk of money we raised is a lot for 17 people to spend. It is to my knowledge the largest series A investment in any Docker-related company besides Docker itself," ClusterHQ CEO Mark Davis said.

"Such a series A investment in a nascent market suggests a lot about the viability of the market itself. It's also a signal to users and to developers that there's going to be a lot more software built around what Linux containers and what Docker have built that is going to make this stuff really viable in production very quickly."

Davis, who is mainly based in Silicon Valley, joined the board of ClusterHQ last June and was appointed CEO in August.

By automating the creation and deployment of apps in containers - a lighter-weight form of virtualisation - Docker is seen as freeing developers from software and infrastructure dependencies, cutting costs and creating efficiencies in the process.

ClusterHQ's Flocker is designed to make all the components of a containerised application portable, including stateful elements such as databases.

"Without Flocker today, if you want to build a Dockerised or containerised application, any significant application today probably has not just one but several stateful services that it deals with: a database and in some cases multiple databases. The way Docker works is these need to be essentially delivered as external services that the container connects to. That external service is outside the domain of Docker and the container system," Davis said.

"It works fine. You can do all sorts of cool stuff with your containers but the benefit of containers, like the ability to think about deploying them super-rapidly and maybe in different places, is not something that is built into Docker today for the stateful parts of services."

That limitation rules out, for example, live-migrating a stateful application from server A to server B or from cloud A to cloud B inside a Docker environment.

"For many production environments [that's] essential. Particularly over the past decade or so, the folks at VMware and with other hypervisor systems have taught us that one of the cool tricks you can do with hypervisors is live-migrate applications as they're running, which is really powerful," he said.

"It allows us to do lots of things operationally inside a datacentre, for high availability reasons, disaster recovery, load balancing, maintenance. Today, you can't do that with containers because the stateful part of the application can't be moved while the state is maintained."

According to ClusterHQ, Flocker is a data volume and container manager that enables a container holding a stateful component to be live-migrated from one server to another without any loss of transaction, service, or downtime.

"Containers are exciting and probably the next big wave of how we build infrastructure applications, how we build the server side of mobile distributed applications. But it's still a technology base that's in its infancy," Davis said.

"It's almost like containers today are what virtualisation was 12 to 15 years ago. Advanced developers can actually build production systems out of them but most people don't have the top one percent of software engineers working for them. It's really hard to make this stuff work in production that's robust and scalable, particularly around some of the more prosaic parts of infrastructure."

These parts include areas such as backups, disaster recovery and compliance, which occur outside the Docker environment.

"It can be made to work but what customers want is to be able to Dockerise everything and take advantage of all the coolness of containers and Docker for a whole app, not just the stateless parts," Davis said.

"If we can enable that, we can enable everyone to take advantage of the benefits of Docker and deliver the kinds of enterprise-grade robustness, scalability, deployability and ongoing manageability in the production environment, which is today probably cloud-based and probably not tied to any particular hardware infrastructure."

ClusterHQ aims to double employee numbers by the middle of the year, mainly through hiring software engineers and some developer evangelists, with the emphasis on ensuring Flocker matches Docker's ease of use and appeal to developers. For the foreseeable future all its technology will continue to be open source.

It expects its primary source of revenue to come from subscription licence fees from companies that want to run Flocker in production environments with the support of the software's architects, for security, bug fixes and enhancements.

More on Docker

Editorial standards