Hatchet The Distributed Task Queue for More Resilient Web Applications

What is Hatchet?

Hatchet is a modern distributed task queue system engineered to enhance web application resilience. It operates as a low-latency queue with an average start time of 25ms, combining real-time interaction capabilities with enterprise-grade reliability. The platform supports multiple queuing strategies including FIFO, LIFO, Round Robin, and Priority Queues.

Built with a focus on scalability and reliability, Hatchet offers comprehensive features such as customizable retry policies, built-in error handling, and full observability of task runs. The system supports multiple programming languages through declarative SDKs, including Python, Typescript, and Go, making it versatile for various development environments.

Features

Low Latency Scheduling: 25ms average start time for real-time task execution
Concurrency Management: Support for FIFO, LIFO, Round Robin, and Priority Queues
Workflow Orchestration: Full DAG-style execution capabilities
Error Handling: Customizable retry policies and built-in recovery mechanisms
Observability: Searchable runs with log streaming and metrics tracking
Multi-language Support: SDKs for Python, Typescript, and Go
Cron Scheduling: Support for recurring task execution
Spike Protection: Traffic smoothing and system load management

Use Cases

Generative AI request distribution
Large-scale document batch processing
Multi-modal system workflow orchestration
Event-based architecture management
Background task processing
Scheduled task execution

FAQs

What is Hatchet?

Hatchet is a managed low-latency queue for web apps to solve scaling issues like concurrency, fairness, and rate limiting.
Does Hatchet Cloud run my compute?

Not yet, but soon. Currently Hatchet Cloud provides a managed control plane for workers. Users can run workers on any cloud provider or on-premises and connect them to Hatchet Cloud.
Is there a self-hosted version of Hatchet?

Yes, Hatchet is an MIT-licensed Open Source project and instructions for self-hosting docker containers can be found in the documentation.