pgmq-rest

Because sometimes your messages need a queue, and your queues need a REST API.

Overview

pgmq-rest provides a REST API for PGMQ (PostgreSQL Message Queue), making it easy to integrate message queues into your applications.

Features

• 🚀 Simple REST API for PGMQ (PostgreSQL Message Queue)
• 🔄 Support for sending and receiving messages
• 📦 Batch operations for better performance
• 📊 Queue metrics and monitoring
• 🐳 Docker and Docker Compose support
• 📝 Swagger documentation included

Quick Start

The fastest way to get started is using Docker Compose:

cd example
docker-compose up -d

To stop the stack:

docker-compose down -v

Usage Example

Here's a quick example of how to use the API:

# Send a single message
curl -X POST http://localhost:8080/api/v1/send \
  -H "Content-Type: application/json" \
  -d '{"queue_name": "my_queue", "msg": {"task": "process_data"}}'
# Response: [123]  # Returns the message ID

# Send multiple messages
curl -X POST http://localhost:8080/api/v1/send_batch \
  -H "Content-Type: application/json" \
  -d '{"queue_name": "my_queue", "msgs": [{"task": "process_data"}, {"task": "process_data"}]}'
# Response: [123, 124]  # Returns message IDs for each message

# Read messages
curl -X POST http://localhost:8080/api/v1/read \
  -H "Content-Type: application/json" \
  -d '{"queue_name": "my_queue", "vt": 30, "qty": 1}'
# Response: [["123", 1, "2024-04-15T12:00:00Z", "2024-04-15T12:00:30Z", {"task": "process_data"}, {}]]
# Format: [msg_id, read_ct, enqueued_at, vt, message, headers]

# Read messages with polling
curl -X POST http://localhost:8080/api/v1/read_with_poll \
  -H "Content-Type: application/json" \
  -d '{"queue_name": "my_queue", "vt": 30, "qty": 1}'
# Response: [["123", 1, "2024-04-15T12:00:00Z", "2024-04-15T12:00:30Z", {"task": "process_data"}, {}]]
# Format: [msg_id, read_ct, enqueued_at, vt, message, headers]

API Reference

The API provides the following main endpoints:

• POST /api/v1/send - Send a single message to a queue
• POST /api/v1/send_batch - Send multiple messages to a queue
• POST /api/v1/read - Read messages from a queue
• POST /api/v1/read_with_poll - Read messages with polling
• GET /api/v1/metrics - Get queue metrics

For detailed API documentation, visit http://localhost:8080/docs after starting the service.

Configuration

Variable	Description	Default
DB_HOST	PostgreSQL host	postgres
DB_PORT	PostgreSQL port	5432
DB_NAME	Database name	postgres
DB_USER	Database user	postgres
DB_PASSWORD	Database password	postgres
DB_POOL_SIZE	Connection pool size	20

Performance Considerations

• Connection Pooling: Adjust DB_POOL_SIZE based on your application's concurrency needs
• Batch Operations: Use send_batch for better performance when sending multiple messages
• Visibility Timeout: Set appropriate visibility timeout when reading messages to prevent message loss
• Queue Size: Monitor queue metrics to prevent queue overflow

Security

• Always use secure connections (HTTPS) in production
• Change default credentials in production
• Use environment variables or secrets management for sensitive configuration
• Consider network isolation for the PostgreSQL instance
• Regularly update to the latest version for security patches

Development

1. Start a local PostgreSQL instance with pgmq:

   # Cleanup existing container and volumes
   docker stop pgmq 2>/dev/null || true
   docker rm pgmq 2>/dev/null || true
   docker volume rm pgmq_data 2>/dev/null || true
   
   # Start PGMQ
   docker run -d --name pgmq \
     -p 5432:5432 \
     -e POSTGRES_USER=postgres \
     -e POSTGRES_PASSWORD=postgres \
     -e POSTGRES_DB=postgres \
     -v pgmq_data:/var/lib/postgresql/data \
     -v $(pwd)/init-pgmq.sql:/docker-entrypoint-initdb.d/init-pgmq.sql \
     tembo.docker.scarf.sh/tembo/pg17-pgmq:latest

2. Run the development server:

   DB_HOST=localhost DB_PORT=5432 DB_USER=postgres DB_PASSWORD=postgres DB_NAME=postgres DB_POOL_SIZE=20 bun dev

3. Run the tests:

   DB_HOST=localhost DB_PORT=5432 DB_USER=postgres DB_PASSWORD=postgres DB_NAME=postgres DB_POOL_SIZE=20 bun test

Manual Setup

If you prefer to run the services manually instead of using Docker Compose:

1. Start the PGMQ container:

   docker run -d --name pgmq \
     -p 5432:5432 \
     -e POSTGRES_USER=postgres \
     -e POSTGRES_PASSWORD=postgres \
     -e POSTGRES_DB=postgres \
     -v $(pwd)/init-pgmq.sql:/docker-entrypoint-initdb.d/init-pgmq.sql \
     tembo.docker.scarf.sh/tembo/pg17-pgmq:latest

2. Run the pgmq-rest container:

   docker run -d --name pgmq-rest \
     -p 8080:8080 \
     -e DB_HOST=pgmq \
     -e DB_PORT=5432 \
     -e DB_NAME=postgres \
     -e DB_USER=postgres \
     -e DB_PASSWORD=postgres \
     --link pgmq:postgres \
     eichenroth/pgmq-rest:latest

3. Cleanup:

   docker stop pgmq-rest pgmq
   docker rm pgmq-rest pgmq
   docker volume rm pgmq_data

Troubleshooting

Common issues and solutions:

1. Connection Issues

   • Verify PostgreSQL is running and accessible
   • Check network connectivity between services
   • Verify credentials and permissions

2. Queue Operations Fail

   • Check if the queue exists
   • Verify message format
   • Check PostgreSQL logs for errors

3. Performance Issues

   • Monitor queue metrics
   • Check PostgreSQL performance
   • Adjust connection pool size if needed

License

This project is licensed under the MIT License.

Acknowledgments

• Built with Elysia.js
• Powered by PGMQ