aws-durable-execution-sdk-python

What it is and what it does

This is an AWS SDK for building durable, long-running workflows inside Lambda functions that can survive Lambda pauses, timeouts, and restarts. It provides decorators and context objects to define steps with retry logic, waits for time or external callbacks, and parallel fan-out operations, all with automatic checkpointing so execution can resume from where it left off rather than restarting from the beginning. The SDK handles deterministic replay internally so that side effects don't duplicate and logging is de-duplicated.

You write Lambda handlers using the `@durable_execution` decorator and define individual work units as `@durable_step` functions, then call them through the context's `step()`, `wait()`, and parallel methods. The SDK depends only on boto3 and runs as pure Python, making it straightforward to add to existing Lambda projects. It's in active development (Beta status) and supported on Python 3.11, 3.12, 3.13, and 3.14.

Use it for:

• Multi-step order processing workflows that need to pause for approval or payment confirmation without blocking Lambda execution
• ETL pipelines that checkpoint progress and resume after failures without reprocessing completed batches
• Approval chains or human-in-the-loop processes that wait for external callbacks before proceeding to the next step
• Parallel fan-out tasks with configurable completion criteria across multiple concurrent operations
• Complex state machines that require structured subflows and isolated execution contexts

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Builds long-running AWS Lambda workflows with automatic checkpointing, durable steps, waits, callbacks, and parallel execution that can resume after Lambda pauses or restarts.

Yes, if you are building multi-step workflows on AWS Lambda and need automatic checkpointing and replay. The low install friction, active maintenance, permissive license, and zero known vulnerabilities make it a safe addition. Beta status means the API may evolve, but the SDK is in active use. Not necessary for simple single-step Lambda functions or if you already use alternative orchestration tools.

Install

aws-durable-execution-sdk-python on PyPI

pip install aws-durable-execution-sdk-python

uv add aws-durable-execution-sdk-python

poetry add aws-durable-execution-sdk-python

Installing aws-durable-execution-sdk-python

Before you install

Low install friction; single pure-Python wheel dependency on boto3. Active maintenance with recent releases; repo is not archived and receives commits. Beta status (Development Status :: 4) indicates the SDK is still stabilizing but in active use.

License in practice

Apache-2.0 permissive license allows commercial and private use with minimal restrictions; suitable for most production and proprietary projects.

Quickstart

pip install aws-durable-execution-sdk-python

from aws_durable_execution_sdk_python import DurableContext, durable_execution, durable_step

@durable_step
def validate_order(step_ctx, order_id: str) -> dict:
    return {"order_id": order_id, "valid": True}

@durable_execution
def handler(event: dict, context: DurableContext) -> dict:
    validation = context.step(validate_order(event["order_id"]), name="validate")
    return {"status": "approved", "order_id": event["order_id"]}

Requires Python 3.11 or later; designed for AWS Lambda runtime environment with boto3 available.

Verify before relying

• Whether checkpointing and replay work correctly across all Lambda pause/restart scenarios in production
• Performance overhead of automatic checkpointing on workflow latency and cost
• Compatibility with Lambda layers, container images, and other deployment models beyond standard zip packages
• Maximum timeout limits for durable workflows and whether they can exceed typical Lambda execution windows

Package facts

Evidence: aws_durable_execution_sdk_python-1.7.0-py3-none-any.whl

Tags

More Distributed Computing packages