Shipping Docker

It's no secret that we are huge fans of Docker, the container packaging that's currently busy redefining the way we ship and develop applications. In fact we've been using Docker for production workloads since summer 2013.

It’s no secret that we are huge fans of Docker, the container packaging that’s currently busy redefining the way we ship and develop applications. In fact we’ve been using Docker for production workloads since summer 2013, as well as the terrific CoreOS (a minimalistic, auto-updating and clusterable container-host). RedHat’s apparent answer to CoreOS - Atomic - looks amazing as well, even though we have not been able to test-drive it.

Complemented with our toolchain of Puppet, Chef and Ansible for host provisioning (or CoreOS for the absence of it) and CloudFormation for AWS infrastructure provisioning our Devops teams are almost at the point where the only thing that is not 100% automated about our setups is the reading of a projects README.md. Exciting times!

Good news, everyone!

But our setup is not what this post is about, because this week was really all about these amazing breakthroughs for the Docker community:

  • Docker has finally reached 1.0 - while we certainly never shared the sentiment that Docker was not ready for production, it’s a hugely important milestone
  • RedHat 7 has reached general availability and ships with Docker - combined with the strategy of RedHat to support container virtualization wholeheartedly are terrific news for our long-term support and certification requiring enterprise customers
  • Datadog (one of our favourite monitoring services) announced full circle support for their monitoring agents: now we can monitor hosts and Docker containers from within Docker containers running on hosts running nothing but Docker containers!

Congratulations to everyone involved and thanks a lot for your collective efforts! The future is around the corner, and it’s configured in a Dockerfile.