As we count down the final days of 2017, we would like to bring you the final installment of the top 5 blogs of 2017. On day 5, we take a look back DockerCon EU, when we announced Kubernetes support in the Docker platform. This blog takes an in-depth look at the industry-leading container platform and the addition of Kubernetes.
The Docker platform is integrating support for Kubernetes so that Docker customers and developers have the option to use both Kubernetes and Swarm to orchestrate container workloads. Register for beta access and check out the detailed blog posts to learn how we’re bringing Kubernetes to:
Docker is a platform that sits between apps and infrastructure. By building apps on Docker, developers and IT operations get freedom and flexibility. That’s because Docker runs everywhere that enterprises deploy apps: on-prem (including on IBM mainframes, enterprise Linux and Windows) and in the cloud. Once an application is containerized, it’s easy to re-build, re-deploy and move around, or even run in hybrid setups that straddle on-prem and cloud infrastructure.
The Docker platform is composed of many components, assembled in four layers:
- containerd is an industry-standard container runtime implementing the OCI standards
- Swarm orchestration that transforms a group of nodes into a distributed system
- Docker Community Edition providing developers a simple workflow to build and ship container applications, with features like application composition, image build and management
- Docker Enterprise Edition, to manage an end to end secure software supply chain and run containers in production
These four layers are assembled from upstream components that are part of the open source Moby Project.
Docker’s design philosophy has always been about providing choice and flexibility. This is important for customers that are integrating Docker with existing IT systems, and that’s why Docker is built to work well with already-deployed networking, logging, storage, load balancers and CI/CD systems. For all of these (and more), Docker relies on industry-standard protocols or published and documented interfaces. And for all of these, Docker Enterprise Edition ships with sensible defaults, but those defaults can be swapped for certified third party options for customers that have existing systems or prefer an alternative solution.
In 2016, Docker added orchestration to the platform, powered by the SwarmKit project. In the past year, we’ve received lots of positive feedback on Swarm: it’s easy to set up, is scalable and is secure out-of-the-box.
We’ve also gotten feedback that some users really like the integrated Docker platform with end-to-end container management, but that they want to use other orchestrators, like Kubernetes, for container scheduling. Either because they’ve already designed services to work on Kubernetes or because Kubernetes has particular features they’re looking for. This is why we are adding Kubernetes support as an orchestration option (alongside Swarm) in both Docker Enterprise Edition, and in Docker for Mac and Windows.
We’re also working on innovative components that make it easier for Docker users to deploy Docker apps natively with Kubernetes orchestration. For example, by using Kubernetes extension mechanisms like Custom Resources and the API server aggregation layer, the coming version of Docker with Kubernetes support will allow users to deploy their Docker Compose apps as Kubernetes-native Pods and Services.
With the next version of the Docker platform, developers can build and test apps destined for production directly on Kubernetes, on their workstation. And ops can get all the benefits of Docker Enterprise Edition – secure multi-tenancy, image scanning and role-based access control – while running apps in production orchestrated with either Kubernetes or Swarm.
The Kubernetes version that we’re incorporating into Docker will be the vanilla Kubernetes that everyone is familiar with, direct from the CNCF. It won’t be a fork, nor an outdated version, nor wrapped or limited in any way.
Through the Moby Project, Docker has been working to adopt and contribute to Kubernetes over the last year. We’ve been working on containerd (now 1.0) and cri-containerd for the container runtime, on InfraKit for creating and managing Kubernetes installs, and on libnetwork for overlay networking. See the Moby Project blog post for more examples and details.
Docker and Kubernetes share much lineage, are written using the same programming language and have overlapping components, contributors and ideals. We at Docker are excited to have Kubernetes support in our products and into the open source projects we work on. And we can’t wait to work with the Kubernetes community to make containers and container-orchestration ever more powerful and easier to use.
While we’re adding Kubernetes as an orchestration option in Docker, we remain committed to Swarm and our customers and users that rely on Swarm and Docker for running critical apps at scale in production. To learn more about how Docker is integrating Kubernetes, check out the sessions “What’s New in Docker” and “Gordon’s Secret Session” at DockerCon EU.
Where to go from here?