Over the course of the past year and a half, Release Engineering, Site Reliability Engineering, and Services have been working on a container based continuous delivery pipeline for MediaWiki related software. The pipeline will empower developers to efficiently and safely move their software changes from their local development environment into the Wikimedia production cluster. The pipeline is driven by technologies such as Jenkins, Docker (via a small abstraction layer we call Blubber), Helm, and Kubernetes to ensure a high level of parity between development, integration, and production environments, and provides for graduated testing and feedback throughout the process.
Please join us to learn how you might contribute to the design, implementation, and testing of the pipeline and how it can support you in your work.
*Public YouTube recording on MediaWiki channel: https://www.youtube.com/watch?v=GvowpttwdqQ
*Commons: https://commons.wikimedia.org/wiki/File:Wikimedia_Continuous_Delivery_Pipeline-_Say_What%3F.webm