sqlmesh-test-project

Basic Setup for Open Source Deployment. Points at BigQuery and uses GCP Cloud SQL Postgres instance for the backend.

GCP Setup

1. Have BigQuery Enabled
2. Create a Cloud SQL Postgres Instance
3. Create a database for the state backend -- Get the connection string
4. Create a SQLMesh User for the database -- Username / Password
5. Create a GCP service account and download a json keyfile
6. Rename it to keyfile.json and place it in the root of the repo

Repo Setup

1. Download uv
2. Run make install
3. Fill out .env file -- reference .env.example file

Current Configuration

See the config.py file. It MUST be named config.py and have a config object in it.

Production Gateway

• Connection: BigQuery
• State Connection: Cloud SQL (Postgres)
• Test Connection: DuckDB

Linter

• Enabled
• Rules:
  • Ambiguous Invalid Columns
  • Invalid Select Star Expansion
  • No Ambiguous Projections
  • No Missing Audits
  • No Missing Unit Tests