A practical guide to using the Kubernetes API from Python in production.
Batteries are included: Python code, service account manifests, Dockerfile and all commands to deploy directly to Google Kubernetes Engine (no affiliation).
Kubernetes has ranked as one of the top 25 Go projects on Github for a while now. However, the Kubernetes API itself is language agnostic and auto-generated clients exist for many languages, including Python.
In this talk I will show how you can query deployments, watch for changes, and launch your own deployments with only a few lines of Python code.
The presentation comes with batteries included. It will not only show the code but include everything you need to run it inside Kubernetes itself, like Dockerfiles and service account manifests. I will use the Google Kubernetes Engine (GKE) for the demo.
By the end of the talk, you will know how to interact with the Kubernetes API from Python and how to deploy that program onto Kubernetes itself. From there it is only a small step to modify the code and create, watch or update pods, ingresses, namespaces, replica sets or any other Kubernetes resource.
Watch 'Create and Watch Kubernetes Resources With Python' on PyCon AU's YouTube account
Oliver Nagy
Ucroo