Automated deployment of Python packages for development
summary
A reproducible pipeline converts each approved Git change into deployed packages: Gerrit + Jenkins CI run tests and build sdist/wheel and conda packages, Jenkins copies artifacts into an internal PyPI via Apache and installs them into an Anaconda environment promoted dev→test→prod; versioneer derives PEP 440 versions from git tags and Ansible manually promotes production releases. This yields instant access to latest dev builds, pinned test collections per tag, and deliberate production updates — a multi‑year practice that now runs reliably.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 4:54Development and deployment
- 10:03Code review with Gerrit
- 12:05Gerrit configuration
- 15:39Continuous integration
- 17:19Automated build process
- 20:10Standardization templates
- 21:26Release management
- 24:05Conda and environment setup
- 27:07Questions and answers
related talks
39mA Pythonic Approach to Continuous Delivery
30mImplementing Continuous Deployment with Jenkins & Gerritt
34mContinuous Deployment for webapps based on Django
28mBuilding reproducible Python applications for secured environments
27mRelease Management with Devpi
29mCan packaging improve Django deployments?
01h 12mPitfalls of Continuous Deployment
36mPython Packaging from Init to Deploy