Skip to content

Latest commit

 

History

History
264 lines (194 loc) · 5.51 KB

WEBHOOK_TESTING.md

File metadata and controls

264 lines (194 loc) · 5.51 KB

Testing Webhook → Queue → Worker → AI Flow

🎯 Overview

This guide shows you how to test the complete CI/CD analysis flow:

GitHub Webhook → Backend → BullMQ Queue → Worker → AI Analysis → Database

📋 Prerequisites

1. Start Redis (Required for Queue)

docker-compose up -d

Verify:

redis-cli ping
# Should return: PONG

2. Start Backend Server

npm run start:backend

Should see:

  • Log processing queue initialized.
  • Server is running on http://localhost:3001
  • ✅ Google Gemini AI initialized (gemini-1.5-flash)

3. Start Worker Process

npm run start:worker

Should see:

  • Log processing worker started.

🧪 Quick Test (Automated)

Run the complete test script:

./test-webhook-flow.sh

This script will:

  1. ✅ Check Redis is running
  2. ✅ Check backend server
  3. ✅ Check worker process
  4. ✅ Simulate GitHub webhook
  5. ✅ Verify job in queue
  6. ✅ Monitor worker processing

🔧 Manual Testing (Step-by-Step)

Step 1: Prepare Webhook Payload

Create test-webhook.json:

{
  "action": "completed",
  "workflow_run": {
    "id": 123456789,
    "name": "CI Build",
    "conclusion": "failure",
    "html_url": "https://github.com/youruser/yourrepo/actions/runs/123456789"
  },
  "repository": {
    "full_name": "youruser/yourrepo",
    "name": "yourrepo"
  },
  "installation": {
    "id": 12345
  }
}

Step 2: Calculate Webhook Signature

# Your webhook secret from .env
SECRET="your_webhook_secret_here"

# Calculate signature
SIGNATURE=$(cat test-webhook.json | openssl dgst -sha256 -hmac "$SECRET" | sed 's/^.* //')
echo "sha256=$SIGNATURE"

Step 3: Send Webhook Request

curl -X POST http://localhost:3001/api/webhooks/github \
  -H "Content-Type: application/json" \
  -H "X-GitHub-Event: workflow_run" \
  -H "X-Hub-Signature-256: sha256=$SIGNATURE" \
  -d @test-webhook.json

Expected Response:

Event received

Step 4: Verify Job in Queue

# Check queue size
redis-cli llen "bull:log-processing:wait"

# View job details
redis-cli --raw lrange "bull:log-processing:wait" 0 -1 | jq '.'

Step 5: Watch Worker Logs

In your worker terminal, you should see:

Processing job for run ID: 123456789 in repo: youruser/yourrepo
Successfully fetched log URL for run 123456789
Downloading logs from: https://...
🤖 Sending request to Google Gemini AI...
✅ Received response from Gemini AI
Saving analysis results to database...
Analysis saved for run 123456789

Step 6: Check Database

npm run prisma:studio

Open http://localhost:5555 and check:

  • WorkflowRun table - should have entry for run ID 123456789
  • AnalysisResult table - should have AI analysis

🔍 Troubleshooting

Queue Issues

Problem Solution
Redis connection refused Run docker-compose up -d
No jobs in queue Check webhook signature is correct
Worker not processing Ensure worker is running with npm run start:worker

Webhook Issues

Problem Solution
401 Invalid signature Check GITHUB_WEBHOOK_SECRET in .env matches
404 Not Found Verify backend is running on port 3001
Job not queued Check backend logs for errors

Worker Issues

Problem Solution
GitHub authentication failed Set GITHUB_APP_ID and GITHUB_PRIVATE_KEY in .env
AI error Set GEMINI_API_KEY in .env
WorkflowRun not found Worker expects run to exist in DB first

🎬 Real GitHub Webhook Setup

To receive real webhooks from GitHub:

1. Install GitHub App

  • Go to your repo → Settings → GitHub Apps
  • Install your CI/CD Analyzer app

2. Configure Webhook

  • Webhook URL: https://your-domain.com/api/webhooks/github
  • Events: Select "Workflow runs"
  • Secret: Set your GITHUB_WEBHOOK_SECRET

3. Use ngrok for Local Testing

# Install ngrok
npm install -g ngrok

# Expose local server
ngrok http 3001

Use the ngrok URL in GitHub webhook settings:

https://abc123.ngrok.io/api/webhooks/github

📊 Monitoring Queue

View Queue Status

# Jobs waiting
redis-cli llen "bull:log-processing:wait"

# Jobs completed
redis-cli llen "bull:log-processing:completed"

# Jobs failed
redis-cli llen "bull:log-processing:failed"

Clear Queue (if needed)

# Clear all queues
redis-cli del "bull:log-processing:wait"
redis-cli del "bull:log-processing:completed"
redis-cli del "bull:log-processing:failed"

✅ Success Criteria

You know everything is working when:

  1. ✅ Webhook returns 200 Event received
  2. ✅ Backend logs show Dispatched job for run ID: ...
  3. ✅ Worker logs show Processing job for run ID: ...
  4. ✅ Worker logs show 🤖 Sending request to Google Gemini AI...
  5. ✅ Worker logs show ✅ Received response from Gemini AI
  6. ✅ Worker logs show Analysis saved for run ...
  7. ✅ Database has entry in AnalysisResult table

🚀 Quick Reference

# Start everything
docker-compose up -d        # Redis
npm run start:backend       # Backend server
npm run start:worker        # Worker process

# Test webhook flow
./test-webhook-flow.sh      # Automated test

# Monitor
redis-cli monitor           # Watch Redis commands
npm run prisma:studio       # View database

That's it! Your webhook → queue → worker → AI flow is ready to test! 🎉