This article is more than 1 year old
Docker adds continuous integration to DataKit
Your code tested as soon as you send it to GitHub
Here's a handy tool from Docker's GitHub repository: a continuous integration library to help manage DataKit projects.
DataKit went open source in May last year. At the time, Anil Madhavapeddy explained it was part of the strategy to let users run Docker natively on Windows and Mac machines. DataKit provides the software component pipeline framework, sitting alongside the HyperKit virtualisation tools and VPNKit networking tools.
DataKit CI complements the Unix-like software pipelining framework, by automatically firing up tests based on GitHub activity.
As explained at the repository, it “monitors your GitHub project and tests each branch, tag and pull request”, with test results displayed in the GitHub UI as status indicators.
Logs and state information are stored in DataKit, to simplify code review with Git tools.
Basically: a GitHub bridge monitors the project and sends changes to DataKit; DataKit CI sees the changes and runs user-specified tests, with an interface the user can watch; and the changes are recorded through the bridge back to GitHub.
Policies can be set, for example so that admins can see everything, while a lower-privilege user can only use DataKit CI for their own projects. ®