Figure out better tooling for Airbyte devs to develop on Cloud Kube Clusters.

[davinchia]

Tell us about the problem you're trying to solve

Today we have a rough process of developing on Kube. Devs usually build local images and rely on the local docker caching to pull those images into local kube clusters (either minikube or docker for desktop).

This breaks down if the kube cluster is in the cloud, and testing on cloud kube clusters is essential since we've noticed subtle but breaking differences between them and local kube clusters. In addition to building the image, a dev also has to tag the image, push the image and bump the deployment's image to the new tag. This is very error prone and manual e.g. I've tagged the wrong image before and not realised till half an hour later.

Describe the solution you’d like

A naive solution would be to reduce this to one command.

A better solution is to point the cloud storage cluster to the dev's local docker cache, so image pulls are automatically, if possible.

[davinchia]

Since then:

1. we have various Cloud Env clusters.
2. we now have a Github workflow to build and push images.

Going to close this.