Serverless Backend with AWS Lambda & API Gateway · 课时

测试与监控生产环境

实施无服务器应用测试策略，并为生产环境设置完善的监控和告警

测试与监控生产环境 是 CoddyKit 上的免费 Serverless Backend with AWS Lambda & API Gateway 课时。这是第 3 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 Serverless Backend with AWS Lambda & API Gateway 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Serverless Backend with AWS Lambda & API Gateway 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

Beyond Dev: Production Monitoring

When your serverless application goes live, testing in development isn't enough. You need robust production monitoring to ensure it's always available, performing well, and serving your users correctly.

Production monitoring focuses on real-time insights, detecting issues before they impact users, and understanding the system's health in a live environment.

Observability: Logs, Metrics, Traces

To effectively monitor a serverless application, we rely on three key pillars of observability:

Logs: Detailed records of events and errors from your functions.
Metrics: Numerical data points that show performance and usage trends.
Traces: End-to-end views of requests as they flow through multiple services.

These pillars help you understand what is happening, how well it's performing, and where problems are occurring.

Essential CloudWatch Metrics

AWS CloudWatch automatically collects metrics for your Lambda functions and API Gateway. Key metrics to monitor for Lambda include:

Invocations: How many times your function is called.
Errors: The number of times your function returns an error.
Duration: How long your function runs (latency).
Throttles: When Lambda rejects invocations due to concurrency limits.

Monitoring these helps you quickly spot performance degradation or failures.

Get Notified: CloudWatch Alarms

Metrics alone aren't enough; you need to be alerted when something goes wrong. CloudWatch Alarms allow you to set thresholds on your metrics.

When a metric breaches its threshold (e.g., Error count > 0 for 5 minutes), an alarm can trigger actions like sending notifications via Amazon SNS (Simple Notification Service) to your email or a chat application.

Deep Dive with CloudWatch Logs

When an alarm goes off, or you notice an issue, CloudWatch Logs are your go-to for debugging. Every print() statement or logger message from your Lambda function is sent here.

You can search, filter, and analyze these logs to understand the exact sequence of events that led to an error. Good logging practices are crucial for production debugging.

Try running this example and check its logs in CloudWatch:

import json
import logging

logger = logging.getLogger()
logger.setLevel(logging.INFO)

def lambda_handler(event, context):
    logger.info(f"Received event: {json.dumps(event)}")
    # Simulate some processing
    try:
        if 'fail' in event:
            raise ValueError("Simulated error for logging")
        message = "Processing successful!"
        status_code = 200
    except Exception as e:
        logger.error(f"Error during processing: {e}")
        message = f"Error: {e}"
        status_code = 500
        
    return {
        'statusCode': status_code,
        'body': json.dumps(message)
    }

Trace Requests with AWS X-Ray

Serverless applications often involve multiple services (API Gateway, Lambda, DynamoDB). When a request fails, it's hard to pinpoint where the issue occurred.

AWS X-Ray provides distributed tracing, giving you an end-to-end view of how requests travel through your application. It helps identify performance bottlenecks and errors across different services.

X-Ray: Enabling & Instrumenting

To use X-Ray, you enable it for your Lambda function and API Gateway. For Lambda, you can optionally instrument your code using the X-Ray SDK to add custom annotations or subsegments.

This allows you to capture specific details about your function's execution steps or business logic within the trace.

Run this Python Lambda with X-Ray SDK enabled:

import json
import os
from aws_xray_sdk.core import xray_recorder
from aws_xray_sdk.core.lambda_context import LambdaContext

xray_recorder.configure(service='MyServerlessApp')
xray_recorder.set_stream_strategy(LambdaContext())

def lambda_handler(event, context):
    # X-Ray automatically captures basic Lambda info
    # We can add custom subsegments or annotations
    with xray_recorder.in_segment('my_custom_processing'):
        xray_recorder.put_annotation('transaction_id', 'xyz123')
        xray_recorder.put_metadata('input_event', event)
        
        print("Function processing with X-Ray...")
        # Simulate some work
        result = {'message': 'Hello from X-Ray enabled Lambda!'}
        
    return {
        'statusCode': 200,
        'body': json.dumps(result)
    }

Understanding X-Ray Service Map

Once X-Ray is collecting data, it visualizes your application's components and their connections in a Service Map. This map shows:

The services involved (e.g., API Gateway, Lambda, DynamoDB).
The average latency between them.
Any errors or faults.

You can then drill down into individual traces to see the exact timeline of a request, including all subsegments and any errors.

Proactive Checks: CloudWatch Canaries

Beyond reactive monitoring, CloudWatch Synthetics Canaries offer proactive testing. Canaries are configurable scripts that run 24/7 from outside your application.

They simulate user interactions—like calling an API endpoint, loading a web page, or submitting a form—to check availability and performance. If a canary fails, it can trigger an alarm, alerting you to potential issues before your users notice them.

Monitoring Knowledge Check

Which of the following are key benefits of using AWS X-Ray in a serverless application?

Recap: Production Ready!

Congratulations! You've learned how to make your serverless applications production-ready through robust monitoring and testing strategies.

We covered the pillars of observability (logs, metrics, traces), using CloudWatch for metrics and alarms, deep-diving with CloudWatch Logs, and gaining end-to-end visibility with AWS X-Ray. Finally, we explored proactive testing with CloudWatch Synthetics Canaries.

Implementing these practices will significantly improve your application's reliability and your ability to respond to issues effectively.

常见问题解答

「测试与监控生产环境」课时是免费的吗？

是的 — 「测试与监控生产环境」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Serverless Backend with AWS Lambda & API Gateway 课程的其余内容，请升级到 CoddyKit PRO。 Serverless Backend with AWS Lambda & API Gateway 课程共包含 4 节课。

「测试与监控生产环境」这节课中我会学到什么？

实施无服务器应用测试策略，并为生产环境设置完善的监控和告警你通过在浏览器中直接运行的动手代码来练习 Serverless Backend with AWS Lambda & API Gateway，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 Serverless Backend with AWS Lambda & API Gateway 需要有经验吗？

无需任何先前经验。CoddyKit 上的 Serverless Backend with AWS Lambda & API Gateway 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 3 节课，共 4 节。

「测试与监控生产环境」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 Serverless Backend with AWS Lambda & API Gateway 课中编写并运行代码吗？

能。每节 Serverless Backend with AWS Lambda & API Gateway 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 Serverless Backend with AWS Lambda & API Gateway