AWS Lambda无法发送EC2状态变更通知至Slack求助

我们的需求是：当用户停止运行中的EC2实例、创建新实例或终止实例时，需向Slack频道发送通知。

本地测试验证 #

curl命令测试成功 #

curl -X POST -H 'Content-type: application/json' --data '{"text":"Hello, World!"}' https://hooks.slack.com/services/XXXXX/XXXXX/XXXXX

Python代码测试成功 #

import requests
import os

# Define the data to send in the Slack message
data = {
    "text": "Hi, This is a test from Local"
}

# Get the Slack webhook URL from environment variables
webhook = os.environ.get("webhook_slack")

# Send the POST request to Slack
response = requests.post(webhook, json=data)

# Check and print the response status for debugging
print(f"Slack Response Status: {response.status_code}")
print(f"Slack Response Text: {response.text}")

Lambda执行异常情况 #

AWS EventBridge可正常调用Lambda函数，但Lambda无法向Slack发送通知，Lambda代码如下：

import json
import os
import urllib3

SLACK_WEBHOOK_URL = os.environ['SLACK_WEBHOOK_URL']
http = urllib3.PoolManager()

def lambda_handler(event, context):
    # Log the entire event for debugging
    print("Received event:", json.dumps(event))
    
    # Extract details from EventBridge event
    detail = event.get("detail", {})
    instance_id = detail.get("instance-id")
    state = detail.get("state")
    region = event.get("region")
    
    # Log extracted details
    print(f"Instance ID: {instance_id}, State: {state}, Region: {region}")

    # Define message based on instance state
    if state == "running":
        message = f"🚀 EC2 Instance {instance_id} has been created and is now running in {region}."
    elif state == "stopped":
        message = f"🛑 EC2 Instance {instance_id} has been stopped in {region}."
    elif state == "terminated":
        message = f"💥 EC2 Instance {instance_id} has been terminated in {region}."
    else:
        print("State does not match expected values. Exiting.")
        return  # Exit if the state is not relevant

    # Send notification to Slack
    slack_message = {
        "text": message
    }
    response = http.request(
        "POST",
        SLACK_WEBHOOK_URL,
        body=json.dumps(slack_message),
        headers={"Content-Type": "application/json"}
    )

    # Log the HTTP response status and data for debugging
    print(f"Slack Response Status: {response.status}")
    print(f"Slack Response Data: {response.data}")

    if response.status != 200:
        raise ValueError(
            f"Request to Slack returned an error {response.status}, response is: {response.data}"
        )

    return {"status": "Notification sent successfully"}

当前已配置内容 #

• 为Lambda角色配置了EC2只读访问权限与Lambda执行角色
• 已在Lambda配置中保存环境变量SLACK_WEBHOOK_URL
• Lambda日志中无明确错误信息

排查步骤 #

1. 检查Lambda网络访问权限

   • 默认Lambda运行在VPC外可直接访问公网，若配置了VPC，需确保Lambda所在子网有NAT网关/互联网网关，且安全组、网络ACL允许出站访问Slack的HTTPS端口（443）。
   • 可添加测试代码尝试访问https://www.google.com，验证公网请求能力。

2. 验证环境变量正确性

   • 确认SLACK_WEBHOOK_URL与本地测试使用的URL完全一致，无多余空格或字符。
   • 临时添加日志打印URL（生产环境需删除），确认配置无误：print(f"Slack Webhook URL: {SLACK_WEBHOOK_URL}")

3. 检查urllib3使用细节

   • urllib3默认需处理SSL证书，可临时添加cert_reqs='CERT_NONE'测试（仅用于排查）：

     response = http.request(
         "POST",
         SLACK_WEBHOOK_URL,
         body=json.dumps(slack_message),
         headers={"Content-Type": "application/json"},
         cert_reqs='CERT_NONE'
     )
     

   • 或替换为requests库（需打包依赖或配置Lambda层），兼容本地测试逻辑。

4. 查看完整CloudWatch日志

   • 检查CloudWatch日志组的所有日志流，包括初始化日志和截断的错误信息，确认是否有遗漏的异常记录。

5. 验证EventBridge事件格式

   • 核对打印的Received event日志，确认detail.state字段值是否与代码判断的running/stopped/terminated完全匹配，避免字段名或大小写差异。

6. 检查Slack Webhook有效性

   • 确认Webhook未过期，对应Slack频道权限未变更，可重新生成URL测试。

内容的提问来源于stack exchange，提问作者Antham