Database Connected GCF

Prisma + Google Cloud Function

Connect your GCF to Prisma and make Cloud SQL interactions simpler & cleaner!

Cloud Account Prerequisites:

• Google Cloud Console account
• Cloud SQL instance
• 2nd-gen Cloud Function
• Cloud Run API enabled
• Cloud SQL Admin API enabled

General Knowledge Prerequisites:

• Google Cloud IAM privileges & roles
• Google Cloud Service Accounts
• Google Cloud Run
• Google Cloud Secrets
• Prisma
• Node 16

Local Tool Recommendations:

• VSCode
• NPX (npm install -g npx)
• NVM (or a similar node manager)

To test a GCF locally:

1. nvm use 16 - Switch to Node 16 (nvm install 16.5.1 if not installed)
2. npm install - Install all node_modules for the GCF
3. **DATABASE_URL=mysql://<user>:<pass>@localhost/<database> npx prisma migrate dev --name init - Run migrations to create tables inside the database.
4. DATABASE_URL=mysql://<user>:<pass>@localhost/<database> npx @google-cloud/functions-framework --target=run - Serve the example GCF locally
5. http://localhost:8080/ - Visit the locally served GCF

*A note on easily moving ready components from local to the cloud:

After cloning this repository, you can upload a zip of the example GCF directory directly to a created Cloud Function. Alternatively, you can manually create & copy GCF files and their contents into your created Cloud Function.

*A note on Prisma schema datasource and reconciling localhost issues.

The url property for the datasource object is typically set to env("DATABASE_URL"). This requires your Cloud Function to have access to a stored secret named DATABASE_URL. This variable is formatted as such for MySQL: mysql://<user>:<pass>@localhost/<database>

*TODO: Determine a cleaner way to pass DATABASE_URL into functions-framework for local testing.