Introduction
This is meant to demonstrate scalable cloud architectures in a way that still doesn't cause me to go bankrupt.
The long-term goal is to demonstrate a "record viewer" type of application that lets people see their own record, scaling to ~500K users, while putting the scaling burden on a cloud provider instead of requiring a huge dedicated team.
Since I still have free tier left, I'm trying for GCP but I envision that the same concepts would apply fairly equally.
Dependencies
- Intended to run in Google Cloud Platform
- Node.js to support use of Cloud Functions
Development Infrastructure
See the separate rabornrecord-docker-devenv module for a simplistic Docker setup (Dockerfile and setup/run scripts) that are meant to support modules like this. In particular, the GCP Cloud Functions emulator.
Roadmap
[ ]: Get a scalable record viewer microservice impl stood up.
[ ]: Show off a record at all (completely fake).
[ ]: Add authentication and authorization using a scalable IDaaS.
[ ]: Specify APIs in use using Swagger / OAS 2.0.
[ ]: Add a ton of fake records of the right format (compliant with API).
[ ]: Load testing.