Docker Has Been in an Acquisitive Mood Again, This Time Pulling in Infinit

Docker, the leading vendor in the container industry, has been dipping into its pockets again, this time to acquire French file sync company Infinit. Yep, Docker, the company whose core product is a container management and deployment technology, is acquiring a consumer file sync company.

At first glance, this is an odd acquisition. Infinit is a company with a product that appears similar to Aspera, Accellion, or even good old-fashioned FTP. If you need to share your files, but email send or receive limits are getting in the way, Infinit allows you to create a Share directly to other Infinit users or a link that can be emailed to non–Infinit users to access your files directly, thereby allowing them to download them for review or further work.

This is even more surprising when you consider that the company that Docker came from, dotCloud, provided a similar service. Then again, there appears to be nothing new under the sun. We all know that containers came from Linux chroot and Solaris Zones; all Docker did was make managing and deploying them simpler.

All in all, a very odd purchase. Then again, not so when we consider what is one of the most requested features for a Docker container: distributed storage.

This again seems odd, as originally the main tenet of a container was statelessness. However, as Solomon Hykes of Docker stated in his blog post informing us of the Infinit acquisition:

“At Docker we believe that tools should adapt to the people using them, not the other way around.”

Users and developers of containers are attempting to shoehorn in applications that the original Docker developers never considered in their original specification, which purely considered next-generation cloud native applications. These legacy applications require storage, networking, and resilience, due to developers’ wish for the same portability they can obtain with virtual machine–based technology at the application layer and the need for multicontainer applications that are deployed in distributed systems. Basically, they need to ensure that all containers have access to the same storage—whether the container is deployed in a local data center, a colo environment, or one of many instances of various public clouds—and also the same networking stack.

The networking stack Docker solved with its acquisition of SocketPlane. The Infinit acquisition will solve the distributed storage issue with its yet-to-be-released infinit-cluster. For a more in-depth look at this solution, watch the video of the session that Infinit’s CTO Quentin Hocquet gave at the Docker Distributed Systems Summit in Berlin in October, where he laid out the architecture and demoed a three-node storage solution built from the technology.

How it scales is a question for the future, but it looks interesting. If—or more precisely, when—it is fully integrated into the Docker engine, it will be a powerful tool for the enabling of legacy designed applications.

Container technology is maturing, but apparently not in the direction the purists desire. It is becoming loaded down with legacy requirements that solve issues that should be designed out with the next-generation applications. The purist position is, however, being deprecated due to the desire of users and developers to continue developing applications in a distributed manner, with stateful entities and legacy resilience concepts. Docker’s position is understandable. It needs to continue to stay relevant, and its desire to move with the needs of the user is laudable and will enable it to do so. The only issue with this is that we will now have a technology that will be tethered to the old world, rather than being the forerunner of a true serverless and stateless paradigm.