目录
ALEX WANG

fix(insight): schedule exports off the hook threads (#703)

  • fix(insight): schedule exports off the hook threads

Exporter I/O previously ran inline inside onInvocationStart, onOperationChange, and onInvocationEnd, so a slow exporter delayed the SDK’s checkpoint callbacks in ON_CHANGE mode.

Records are now handed to a background scheduler that exports at most one record at a time. Each record is a complete snapshot, so updates arriving while an export is in flight coalesce into a single pending slot and only the latest is exported next. Exporters for one record run concurrently and independently. onInvocationEnd drains the scheduler and then flushes every exporter once, so the final record is always delivered before the invocation returns; if no worker can be started, the drain exports the pending record itself at that boundary.

  • fix(insight): harden invocation-end delivery guarantees
  • Reject operation-change hooks once invocation end has begun. A checkpoint for an unawaited asynchronous operation can complete while the end record is draining; its RUNNING snapshot must not replace or follow the final record. The closed flag and the final schedule are set under one lock.
  • Complete the pump handle when worker submission fails, so a drain() that already observed the handle wakes up and exports the pending record inline instead of blocking forever.
  • Flush exporters concurrently and independently, matching how exports fan out, so one slow or failing flush never delays the others.

Co-authored-by: Alex Wang wangyb@amazon.com

2天前642次提交

AWS Lambda Durable Execution SDK for Java

License Java Build Core Conformance OpenTelemetry Conformance Javadoc Maven Coverage E2E Tests CodeQL Spotless Roadmap

Build resilient, long-running AWS Lambda functions that automatically checkpoint progress and resume after failures. Durable functions can run for up to one year while you pay only for active compute time.

Key Features

  • Automatic Checkpointing – Progress is saved after each step; failures resume from the last checkpoint
  • Cost-Effective Waits – Suspend execution for minutes, hours, or days without compute charges
  • Configurable Retries – Built-in retry strategies with exponential backoff and jitter
  • Replay Safety – Functions deterministically resume from checkpoints after interruptions
  • Type Safety – Full generic type support for step results
  • Data-Driven Concurrency – Apply a function across a collection with map(), with per-item error isolation and configurable completion criteria

How It Works

Your durable function extends DurableHandler<I, O> and implements handleRequest(I input, DurableContext ctx). The DurableContext is your interface to durable operations:

  • ctx.step() – Execute code and checkpoint the result
  • ctx.wait() – Suspend execution without compute charges
  • ctx.createCallback() – Wait for external events (approvals, webhooks)
  • ctx.waitForCallback() – Simplify callback handling by combining callback creation and submission in one operation
  • ctx.invoke() – Invoke another Lambda function, durable or standard, and wait for the result
  • ctx.runInChildContext() – Run an isolated child context with its own checkpoint log
  • ctx.map() – Apply a function to each item in a collection concurrently
  • ctx.parallel() - Run multiple operations concurrently with optional concurrency control
  • ctx.waitForCondition() – Poll a condition function until it signals done, suspending between polls

Quick Start

Installation

<dependency>
    <groupId>software.amazon.lambda.durable</groupId>
    <artifactId>aws-durable-execution-sdk-java</artifactId>
    <version>VERSION</version>
</dependency>

Your First Durable Function

public class OrderProcessor extends DurableHandler<Order, OrderResult> {

    @Override
    protected OrderResult handleRequest(Order order, DurableContext ctx) {
        // Step 1: Validate and reserve inventory
        var reservation = ctx.step("reserve-inventory", Reservation.class, 
            stepCtx -> inventoryService.reserve(order.getItems()));

        // Step 2: Process payment
        var payment = ctx.step("process-payment", Payment.class,
            stepCtx -> paymentService.charge(order.getPaymentMethod(), order.getTotal()));

        // Wait for warehouse processing (no compute charges)
        ctx.wait(null, Duration.ofHours(2));

        // Step 3: Confirm shipment
        var shipment = ctx.step("confirm-shipment", Shipment.class,
            stepCtx -> shippingService.ship(reservation, order.getAddress()));

        return new OrderResult(order.getId(), shipment.getTrackingNumber());
    }
}

Deployment

See examples/README.md for complete instructions on local testing and running cloud integration tests.

See Deploy and invoke Lambda durable functions with the AWS CLI for more information on deploying and invoking durable functions.

See Deploy Lambda durable functions with Infrastructure as Code for more information on deploying durable functions using infrastructure-as-code.

Documentation

  • AWS Documentation – Official AWS Lambda durable functions guide
  • Best Practices – Patterns and recommendations
  • FAQ - Frequently asked questions for the Java SDK
  • Javadoc - API reference for the Java SDK
  • SDK Design – Details of SDK internal architecture

Core Operations

  • Steps – Execute code with automatic checkpointing and retry support
  • Wait - Pause execution without blocking Lambda resources
  • Callbacks - Wait for external systems to respond
  • Invoke - Call durable functions or standard Lambda functions
  • Child Contexts - Organize complex workflows into isolated units
  • Map - Apply a function across a collection concurrently
  • Parallel - Run multiple operations concurrently with optional concurrency control
  • Wait for Condition - Poll a condition until it’s met, with configurable backoff

Examples

  • Examples - Working examples of each operation

Advanced Topics

Security

See CONTRIBUTING for information about reporting security issues.

License

This project is licensed under the Apache-2.0 License. See LICENSE for details.

关于
12.7 MB
邀请码
    Gitlink(确实开源)
  • 加入我们
  • 官网邮箱:gitlink@ccf.org.cn
  • QQ群
  • QQ群
  • 公众号
  • 公众号

版权所有:中国计算机学会技术支持:开源发展技术委员会
京ICP备13000930号-9 京公网安备 11010802047560号