-
Notifications
You must be signed in to change notification settings - Fork 92
Description
Would it be possible to provide some further explanation on what the local development cycle for the gateway server looks like?
This page describes the build process.
This page describes the local execution process.
When attempting the latter, I've been unable to establish communication with the Kubernetes cluster described in my kube config (a local deployment of Docker Desktop Kubernetes).
I've implemented a local development cycle that leverages Skaffold in conjunction with Helm. Skaffold watches the source files. When a change is detected, it automatically rebuilds the image (I wrote a Dockerfile to COPY and pip install the source code), stores it locally (no pulling to and from a remote registry), and deploys an updated helm chart. The process takes 12 seconds from saved file to functional deployment. It works, and I'd be happy to contribute it upstream if anyone else finds it useful, but I'm wondering what the existing development cycle when interacting with the Kubernetes backend looks like, and if it's possible to implement a cycle that runs the gateway server directly on the host for improved iteration time.
Thanks!