hey y'all, can I get confirmation on setup for my DataHub forking process, any feedback on an improved solution would be appreciated:

Deploying DataHub into a k8s cluster in my internal org

- Fork the repo from GitHub 
- Mirror the fork into my organization's BitBucket repository
- Make any code changes to the mirrored repo in BitBucket 
- Build the docker images and deploy into a registry 
- Configure the default helm charts to point to the custom images in my registry 
- Deploy/update DataHub in k8s with the custom helm charts

@famous-waitress-64616 could you check on this? 🙂

If you're making ingestion changes (i.e. in

metadata-ingestion

), then you'll want to run ingestion from the CLI in a virtualenv with your forked python code. Otherwise, the above sounds reasonable to me. @brainy-tent-14503 could you take a look and confirm?

That process seems right to me. While you might want to run all the tests before creating all the images, the 1 line to build all docker containers is

./gradlew dockerTag

(we use a standard docker plugin for gradle). This will build from source code, create images, and tag them with a version either based on the latest git tag or overridden with DATAHUB_VERSION. These image can then be retagged & pushed to your internal private registry for use.

awesome thanks for the confirmation y'all!