- What is the Atomic Developer Bundle (ADB)?
- When would the Atomic Developer Bundle (ADB) typically be used?
- What is the Typical Usage Pattern?
- What does it contain?
- How do I Install and Run the Atomic Developer Bundle (ADB)?
- What is full feature set?
- What are the deliverables and how do I get it?
- Documentation
- Interested in Contributing to this project?
What is the Atomic Developer Bundle (ADB)?
The Atomic Developer Bundle (ADB) is a prepackaged development environment filled with production-grade pre-configured tools that makes container developer's lives easier. The ADB supports the development of multi-container applications against different technologies and orchestrators while providing a path that promotes best practices.
As a container developer, you want to use the ADB for these reasons:
- Pre-Configured: You don't have to spend time building an environment and fighting configuration battles
- Multiple Environment Support: The ADB works on Windows, Linux and Mac OS X. The ADB supports several orchestrators (Open Shift, Kubernetes, Mesos, and plain Docker). The ADB is language independent and supports multiple developer models (IDE, CLI, SSH containment).
- Production-Grade: The components of the ADB are configured to behave just as they will in production. Containers promise seamless delivery, but only if you test them in the right environment. This is that environment.
- Self-Contained: The ADB is ready to go once installed. It comes prepackaged with the most common components ready, in case they are needed.
- Open Source: The ADB leverages existing tools and technologies where ever possible to avoid pushing a developer into an environment that won't be supportable in production or that is tied to a single vendor. This also means it benefits from the stability of existing projects instead of reinventing the wheel.
The ADB is a virtual machine that is executed with Vagrant and some Vagrant plugins.
When would the Atomic Developer Bundle (ADB) typically be used?
A developer begins using the ADB, in most cases, once they have a working application that has been decomposed into micro-services. They then follow this general outline of steps:
- Build a Dockerfile for each microservice that is not able to be used from another source. In other words, build containers for the parts of the application that are unique and reuse for those that aren't (i.e. a database service)
- Confirm the application works in the ADB by manually launching the application's container components or by using the instance of OpenShift in the container to launch the application
- Build a Nulecule description of the application or complete the OpenShift application configuration
- Test the application
The original motivation behind this project may also be of interest.
What is the Typical Usage Pattern?
The ADB supports three basic modes of usage. The modes vary by how much they rely on tools on the developers workstation. From most to least reliant, they are:
Host-based IDE Mode
This mode uses the ADB as a server resource for host-based IDE tools. In this mode, the user will run eclipse or other IDE tools that will access the resources of the ADB.
Host-based CLI Mode
This mode uses the ADB as a server resource for host-based CLI tools. In this mode, the user will run docker and other CLI tools on their workstation and the result will be containers executed inside of the ADB.
SSH Mode
This mode uses the ADB as a Linux virtual machine. The user will use the ssh command to log into the ADB and will directly execute docker and other commands from the command line. This is similar to having installed a Linux virtual machine and then installing and configuring all of the software the ADB ships with.
More information about using the ADB is available.
What does it contain?
The ADB is built on top of CentOS 7 and contains the following:
- Docker: container runtime and packaging
- Atomic CLI: container usage assistance
- Kubernetes: container orchestration
- OpenShift Origin: a next generation PaaS for docker containers.
The ADB supports Atomic App, an implementation of the multi-container application specification, nulecule for multi-container applications.
How do I Install and Run the Atomic Developer Bundle (ADB)?
Below is a quick installation guide using the most common options. Detailed installation instructions are available.
Install VirtualBox for your operating system. Linux users may wish to use their distribution provided package.
sudo dnf install VirtualBox
Install Vagrant for your operating system. Linux users may wish to use their distribution provided package.
sudo dnf install vagrant
Install the vagrant-service-manager Vagrant plugin:
vagrant plugin install vagrant-service-manager
Download the latest ADB from atlas.hashicorp.com. This can be done automatically by Vagrant:
$ vagrant init projectatomic/adb
Start the ADB
$ sudo vagrant up
If you will be using host based tools, setup your environment with vagrant-service-manager
$sudo vagrant service-manager --help
What is full feature set?
Today the box supports the following:
- Providing docker support to unsupported platforms (i.e. Microsoft Windows, Mac OS X, etc.)
- Kubernetes orchestration for local testing of applications
- Application definition using the Nulecule specification
- Additional goals, objectives and work in progress can be found in the architecture and roadmap document and on the Project Atomic trello board
What are the deliverables and how do I get it?
The ADB is delivered as a Vagrant box for various (currently libvirt and VirtualBox) providers. The boxes are built using the CentOS powered Community Build System. Boxes are delivered via Hashicorp's Atlas and are available at cloud.centos.org. These boxes differ from existing Vagrant boxes for CentOS as they have specific build requirements that are not enabled in those boxes.
Documentation
- Architecture and Roadmap
- Building the Vagrant box for Developers
- Installing the ADB
- How to use the ADB
- Updating the ADB
Interested in Contributing to this project?
We welcome issues and pull requests. Want to be more involved, join us:
Mailing List: container-tools@redhat.com
IRC: #atomic and #nulecule on freenode
- Meetings:
Planning and Integration Meeting:
every Wednesday at 1230 UTC in a Bluejean Video Conference. Alternately, a local phone access number may be available. (1 hour)
Team Standup/Review
every Monday at 1300 UTC in IRC freenode #nulecule (.5 hour)
Note: These meetings, mailing lists, and irc channels may include discussion of other Project Atomic components.
Documentation is written using reStructuredText. An online reStructuredText editor is available.