Skip to content
Workers
Visit Workers on GitHub
Set theme to dark (⇧+D)

Query Postgres from Workers using a database connector

Overview

In this tutorial you learn how to retrieve data in your Cloudflare Workers applications from a PostgreSQL database using Postgres database connector.

For a quick start, you will use Docker to run a local instance of Postgres and PgBouncer, and to securely expose the stack to the Internet using Cloudflare Tunnel.

Basic project scaffolding

To get started:

  1. Run the following git command to clone a basic Postgres database connector project.
  2. After running the git clone command, cd into the new project.
$ git clone https://github.com/cloudflare/worker-template-postgres/
$ cd worker-template-postgres

Cloudflare Tunnel authentication

To create and manage secure Cloudflare Tunnels, you first need to authenticate cloudflared CLI. Skip this step if you already have authenticated cloudflared locally.

$ docker run -v ~/.cloudflared:/etc/cloudflared cloudflare/cloudflared:2021.11.0 login

Running this command will:

  • Prompt you to select your Cloudflare account and hostname.
  • Download credentials and allow cloudflared to create Tunnels and DNS records.

Start and prepare Postgres database

Start the Postgres server

You can find a prepared docker-compose file that does not require any changes in scripts/postgres with the following services:

  1. postgres
  2. pgbouncer - Placed in front of Postgres to provide connection pooling.
  3. cloudflared - Allows your applications to connect securely, through a encrypted tunnel, without opening any local ports.

Run the following commands to start all services. Replace postgres-tunnel.example.com with a hostname on your Cloudflare zone to route traffic through this tunnel.

$ cd scripts/postgres
$ export TUNNEL_HOSTNAME=postgres-tunnel.example.com
$ docker compose up
# Alternative: Run `docker compose up -D` to start docker-compose detached

docker-compose will spin up and configure all the services for you, including the creation of the Tunnel's DNS record. The DNS record will point to the Cloudflare Tunnel, which keeps a secure connection between a local instance of cloudflared and the Cloudflare network.

Import example dataset

Once Postgres is up and running, seed the database with a schema and a dataset. For this tutorial, you will use the Pagila schema and dataset. Use docker exec to execute a command inside the running Postgres container and import Pagila schema and dataset.

$ curl https://raw.githubusercontent.com/devrimgunduz/pagila/master/pagila-schema.sql | docker exec -i postgres_postgresql_1 psql -U postgres -d postgres
$ curl https://raw.githubusercontent.com/devrimgunduz/pagila/master/pagila-data.sql | docker exec -i postgres_postgresql_1 psql -U postgres -d postgres

The above commands will download the SQL schema and dataset files from Pagila's GitHub repository and execute them in your local Postgres database instance.

Edit Worker and query Pagila dataset

Database connection settings

In src/index.ts, replace https://dev.example.com with your Cloudflare Tunnel hostname, ensuring that it is prefixed with the https:// protocol:

src/index.ts
const client = new Client({
user: 'postgres',
database: 'postgres',
hostname: 'https://REPLACE_WITH_TUNNEL_HOSTNAME',
password: '',
port: 5432,
})

At this point, you can deploy your Worker and make a request to it to verify that your database connection is working.

Query Pagila dataset

The template script includes a simple query to select a number (SELECT 42;) that is executed in the database. Edit the script to query the imported Pagila dataset if the pagila-table query parameter is present.

// Query the database.
// Parse the URL, and get the 'pagila-table' query parameter (which may not exist)
const url = new URL(request.url)
const pagilaTable = url.searchParams.get("pagila-table")
let result
// if pagilaTable is defined, run a query on the Pagila dataset
if ([
"actor",
"address",
"category",
"city",
"country",
"customer",
"film",
"film_actor",
"film_category",
"inventory",
"language",
"payment",
"payment_p2020_01",
"payment_p2020_02",
"payment_p2020_03",
"payment_p2020_04",
"payment_p2020_05",
"payment_p2020_06",
"rental",
"staff",
"store",
].includes(pagilaTable)) {
result = await client.queryObject(`SELECT * FROM ${pagilaTable};`)
} else {
const param = 42
result = await client.queryObject(`SELECT ${param} as answer;`)
}
// Return result from database.
return new Response(JSON.stringify(result))

Worker deployment

In wrangler.toml, enter your Cloudflare account ID in the line containing account_id:

wrangler.toml
name = "worker-postgres-template"
type = "javascript"
account_id = ""

Publish your function:

$ wrangler publish
✨ Built successfully, built project size is 10 KiB.
✨ Successfully published your script to
https://workers-postgres-template.example.workers.dev

Set secrets

Create and save a Client ID and a Client Secret to Worker secrets in case your Tunnel is protected by Cloudflare Access.

$ wrangler secret put CF_CLIENT_ID
$ wrangler secret put CF_CLIENT_SECRET

Test the Worker

Request some of the Pagila tables by adding the ?pagila-table query parameter with a table name to the URL of the Worker.

$ curl https://example.workers.dev/?pagila-table=actor
$ curl https://example.workers.dev/?pagila-table=address
$ curl https://example.workers.dev/?pagila-table=country
$ curl https://example.workers.dev/?pagila-table=language

Cleanup

Run the following command to stop and remove the Docker containers and networks:

$ docker compose down
# Stop and remove containers, networks