optimize docker images

[jornh]

I would like to introduce a few optimizations of Docker images across the three service repos:

• Reduce image size by switching to python:3-slim as base image, e.g. Airflow does this as well - this means even quicker Quickstart
• Increase caching to speed up builds (mostly by eliminating pypi/npm dependency pulls and builds)
• Add Gunicorn similar to what's already introduced in Updated Dockerfile to be flexible amundsenmetadatalibrary#66 to allow for a more production ready config (Werkzeug still possible if someone is using it)

This issue mostly to have a common place to link to for discussion etc.

[jornh]

According to https://hub.docker.com/r/amundsendev/amundsen-metadata/tags version 1.1.1 of metadata is 72 MB in compressed size. In comparison v1.0.19 was 347 MB.

This means only 1/4 download size for Amundsen Quickstart images.

[feng-tao]

nice!

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.

[feng-tao]

@jornh could this issue be closed now?

[jornh]

Merge of PR #123 is set up to automatically close it. So we’re close now with FE 1.1.0 tagged and pushed to docker!