Deis – an open-source PaaS for Docker
Having just debuted with its first major release, Deis is an open-source micro-PaaS that makes it easier to deploy and manage Dockerised applications.
A year ago, it was another interesting, new technology. Today, Docker has grown so large that is the heart of an ecosystem of Docker-related tools and technologies. One of these is Deis, a micro-PaaS technology for deploying and managing applications which brings out-of-the-box support for all major JVM languages.
This open-source application platform is designed to simplify the management of applications on self-managed servers. Similar to the open-source PaaS Flynn, Deis is based on both Docker and CoreOS and provides a private Apache-licensed PaaS with a Heroku-inspired workflow.
Deis brings a scalable systems layer below its familiar PaaS surface, making it “more than a Heroko clone”, as others have commented. One of its most significant claims to fame is that it can reduce the effort required for large-scale platform operations, although it has been said that the tool is slightly lacking in system monitoring.
Happy first birth-Deis
This first major-release milestone for Deis, which has already gathered more than 3,000 commits on GitHub, brings a stable API and a solid, scalable architecture of components. Here’s an overview of its main features and the Git Push workflow.
DEIS Git Push Workflow / © DEIS
Compact installation – On a CoreOS Cluster the installation takes less than 30 minutes and only requires one command-line interface. Deis itself is also delivered as a series of Docker images.
High availability – Deis claims to be able to withstand the failure of any one host in the cluster.
Familiar workflows – Deis provides three deployment workflows: Heroku Buildpacks, Dockerfiles and native Docker Images.
Extensive documentation – Deis has expanded its documentation for developers and summarised it on documentation page.
It runs everywhere – Deis is supported by Amazon Web Services, Google Compute Engine, Digital Ocean, Rackspace, OpenStack and VMware. It also runs on both open and private clouds and even on Bare Metal.
