AWS Lambda durable functions

Build fault-tolerant workflows in code with AWS Lambda durable functions.

AWS Lambda durable functions extend the Lambda programming model with built-in capabilities for fault-tolerant, long-running workflows. Checkpoint progress, recover from failures, and suspend execution for up to one year — all without compute charges. You write sequential code in your preferred language, and the open source durable execution SDK handles checkpointing, retries, and state management automatically.

Durable functions introduce two core primitives: Steps (context.step()) for automatic retries and checkpointing, and Waits (context.wait()) for suspending execution without compute charges. Additional operations include waitForCallback() for human-in-the-loop approvals, waitForCondition() for polling, and parallel() or map() for concurrency patterns.

Durable functions currently support Python, Node.js (JavaScript/TypeScript), and Java runtimes, and integrate with existing Lambda event sources, AWS SAM, AWS CDK, CloudFormation, and Terraform.

Hands on with AWS Lambda Durable Functions

Build multi-step applications and AI workflows with AWS Lambda durable functions

AWS News Blog

The launch announcement walks through creating a durable function, using the SDK primitives, and building a production-ready order processing workflow with callbacks for external approvals and configurable retry strategies.

Building fault-tolerant applications with AWS Lambda durable functions

AWS Compute Blog

A deeper technical walkthrough covering the checkpoint and replay mechanism, how durable operations are persisted, and a user onboarding example with email verification and 24-hour timeout using waitForCallback.

Durable Functions vs Step Functions: when to use which

Architecture guidance

Durable functions focus on application logic orchestration using familiar programming languages. Step Functions are suited for workflow orchestration across AWS services with visual state machines. Choose based on whether your coordination is code-centric or service-centric.

re:Invent 2025 - Deep Dive on AWS Lambda durable functions (CNS380)

re:Invent 2025

Eric Johnson and Michael Gasch walk through the evolution from monolithic architectures to durable functions, covering the checkpoint and replay mechanism, compensation patterns, the testing framework, and pricing model.

Common use cases

What to build

Durable functions are well suited for multi-step workflows that need reliability and long-running coordination.

Durable Execution SDK

Developer resources

The SDK provides the primitives for checkpointing, retries, and execution flow management. Available for Python and Node.js/TypeScript.

re:Invent 2025 - Building the future with AWS Serverless

re:Invent 2025

This session covers three major launches: the Serverless MCP server, Lambda Managed Instances, and Lambda Durable Functions, with a live-built application demonstrating each capability.

Presented by:

Eric Johnson

Eric Johnson

Eric Johnson is a Principal Developer Advocate for Serverless Applications at Amazon Web Services and is based in Northern Colorado. Eric is a fanatic about serverless and enjoys helping developers understand how serverless technologies introduces a major paradigm shift in how they approach building and running applications at massive scale with minimal administration overhead. Prior to this, Eric has worked as a developer, solutions architect and AWS Evangelist for an AWS partner company.

Follow on LinkedIn