树莓派3上AWS Greengrass Lambda函数验证失败：未收到MQTT消息求助

Let’s walk through common issues that might be blocking your Lambda from sending messages to the hello/world topic, since you’ve already checked off most of the core setup steps.

1. Verify Lambda Function & Deployment Configuration #

First, double-check the basics in your Greengrass group:

• Confirm your Lambda is set to Run on Greengrass core (not a cloud-only function) in the group’s deployment settings. It’s easy to overlook this toggle when adding the function to the group.
• Check the Lambda code for typos: Ensure the topic name in the code exactly matches hello/world (MQTT topics are case-sensitive, so Hello/World or hello-world won’t work). Also, confirm the code actually publishes the message—official hello-world examples use the Greengrass SDK to send to the local MQTT broker, which forwards to AWS IoT.
• Make sure your Lambda package is compatible with Raspberry Pi 3’s ARMv7 architecture. If you built the package on an x86 machine, you might need to cross-compile or use an ARM-optimized Lambda layer to avoid runtime errors.

2. Confirm Greengrass Deployment Succeeded #

• Head to the AWS IoT Greengrass console, navigate to your group, and check the Deployment history. Ensure the latest deployment shows a status of Completed. If it’s stuck or failed, re-deploy the group and watch for errors (network blips or permission issues often cause partial deployments).
• On your Raspberry Pi, verify the Lambda was synced to the core: Look for your Lambda’s folder in /greengrass/ggc/deployment/lambda/packages/—it should contain your function code and dependencies.

3. Dig Into Lambda & Core Logs for Errors #

Skip the system logs for a minute and focus on Lambda-specific logs:

• Navigate to /greengrass/ggc/var/log/user/<your-lambda-function-name>/ and open the latest log file. Look for errors like missing dependencies, permission denied, or failed MQTT publish calls.
• Check core system logs for context:
  • GGCore.log: Look for entries about Lambda initialization failures or deployment sync issues.
  • GGShadowManager.log: Check for connectivity problems between the core and AWS IoT.

4. Validate MQTT Subscription Setup #

• Ensure your AWS IoT MQTT client is subscribed to the exact topic hello/world (not hello/# or a misspelled variant). Even a single extra space will prevent messages from showing up.
• Confirm the MQTT client is connected to the same AWS region as your Greengrass group. Cross-region MQTT connections won’t receive messages from your core.

5. Test Lambda Execution Directly on the Core #

If you still don’t see messages, trigger the Lambda manually to isolate the issue:

1. SSH into your Raspberry Pi and navigate to the Greengrass core directory:

   cd /greengrass/ggc/core
   

2. Invoke the Lambda with the Greengrass CLI:

   sudo ./greengrass-cli function invoke --function-name <your-lambda-function-name> --payload '{}'
   

3. Check the Lambda logs right after invoking—if the function runs without errors, the issue might be with MQTT forwarding; if it errors out, you’ll see the root cause in the logs.

You can also verify the Greengrass core’s overall health with:

sudo ./greengrass-cli status

This will show if the core is running and if your Lambda is listed as a deployed function.

6. Check Network & Certificate Permissions #

• Ensure your Raspberry Pi has a stable internet connection and can reach AWS IoT endpoints. Test with:

  ping iot.<your-region>.amazonaws.com
  

• Verify the certificate and key files in /greengrass/certs have correct permissions:
  • The private key (privkey.pem) should have permissions 600 (only readable by root).
  • The certificate (cert.pem.crt) and root CA (root.ca.pem) should have permissions 644.

内容的提问来源于stack exchange，提问作者ashwal