Monitoring S5 on Azure¶

Guide to monitoring S5 Slidefactory deployed on Azure Container Apps.

Quick Health Check¶

Via Web Interface¶

1. Open https://slidefactory.sportfive.com (or preview URL)
2. Check if login page loads
3. Log in with Azure AD
4. Verify dashboard loads

Expected: < 2 seconds load time

Via Health Endpoint¶

# Check health endpoint
curl https://slidefactory.sportfive.com/health

# Expected response
{"status": "healthy", "database": "connected", "redis": "connected"}

Azure Portal Monitoring¶

Container Apps Overview¶

1. Go to Azure Portal → Resource Groups → rg-slidefactory-prod
2. Select slidefactory-web-prod Container App
3. View Overview page

Key Metrics: - HTTP requests/sec - Average response time - Replica count - CPU/Memory usage

Log Stream (Real-Time Logs)¶

1. Go to Container App → Log stream
2. View real-time logs from application

What to Look For: - ✅ INFO logs: Normal operations - ⚠️ WARNING logs: Potential issues - ❌ ERROR logs: Problems requiring attention

Example Healthy Logs:

INFO:     Application startup complete
INFO:     GET /health 200 OK
INFO:     GET /presentations/ 200 OK

Example Problem Logs:

ERROR:    Database connection failed
ERROR:    Redis connection timeout
ERROR:    500 Internal Server Error

Metrics Dashboard¶

1. Go to Container App → Metrics
2. Add metrics to chart:
3. CPU Usage (target: < 70%)
4. Memory Usage (target: < 80%)
5. HTTP Requests (track traffic)
6. HTTP Response Time (target: < 2s)

Application Insights¶

Access Insights¶

1. Go to Azure Portal → Application Insights → appi-slidefactory-prod
2. View Application Map for dependencies
3. View Performance for request times
4. View Failures for errors

Key Queries (Log Analytics)¶

Failed Requests:

requests
| where success == false
| where timestamp > ago(1h)
| project timestamp, name, resultCode, duration
| order by timestamp desc

Slow Requests:

requests
| where duration > 2000  // > 2 seconds
| where timestamp > ago(1h)
| project timestamp, name, duration, url
| order by duration desc

Exception Summary:

exceptions
| where timestamp > ago(24h)
| summarize count() by type, outerMessage
| order by count_ desc

Database Connection Errors:

dependencies
| where type == "SQL"
| where success == false
| where timestamp > ago(1h)
| project timestamp, target, duration, resultCode

Alerts¶

Configured Alerts¶

Create Custom Alert¶

1. Go to Container App → Alerts → Create alert rule
2. Select metric (e.g., CPU Usage)
3. Set condition (e.g., > 80% for 5 minutes)
4. Add action group (email notification)
5. Name alert and create

User Activity Monitoring¶

Active Users¶

Query (Application Insights):

customEvents
| where name == "UserLogin"
| where timestamp > ago(1h)
| summarize count() by user_id

Presentation Generation¶

Query (Application Insights):

customEvents
| where name == "PresentationGenerated"
| where timestamp > ago(24h)
| summarize count() by bin(timestamp, 1h)
| render timechart

N8N Workflow Executions¶

Query (Application Insights):

dependencies
| where type == "HTTP"
| where target contains "n8n"
| where timestamp > ago(24h)
| summarize count() by bin(timestamp, 1h), success
| render timechart

Database Monitoring¶

Via Azure Portal¶

1. Go to PostgreSQL Flexible Server → Monitoring → Metrics
2. Add metrics:
3. Connections (track connection pool usage)
4. CPU Percent (target: < 70%)
5. Memory Percent (target: < 80%)
6. IOPS (disk I/O)

Query Performance¶

Slow Queries (connect to database):

-- Enable pg_stat_statements extension
CREATE EXTENSION IF NOT EXISTS pg_stat_statements;

-- Find slow queries (> 1 second average)
SELECT
  query,
  calls,
  total_time,
  mean_time,
  max_time
FROM pg_stat_statements
WHERE mean_time > 1000  -- milliseconds
ORDER BY mean_time DESC
LIMIT 10;

Database Size¶

-- Check database size
SELECT
  pg_database.datname,
  pg_size_pretty(pg_database_size(pg_database.datname)) AS size
FROM pg_database
WHERE datname = 'slidefactory';

-- Check table sizes
SELECT
  schemaname || '.' || tablename AS table,
  pg_size_pretty(pg_total_relation_size(schemaname||'.'||tablename)) AS size
FROM pg_tables
WHERE schemaname = 'public'
ORDER BY pg_total_relation_size(schemaname||'.'||tablename) DESC
LIMIT 10;

Redis Monitoring¶

Via Azure Portal¶

1. Go to Azure Cache for Redis → Metrics
2. Add metrics:
3. Connected Clients (should be steady)
4. Server Load (target: < 70%)
5. Cache Hits / Cache Misses (ratio)
6. Used Memory (target: < 80%)

Redis CLI Monitoring¶

# Connect to Redis (with TLS)
redis-cli -h redis-prod.redis.cache.windows.net \
  -p 6380 \
  -a <password> \
  --tls

# Check memory usage
INFO memory

# Check connected clients
CLIENT LIST

# Monitor commands in real-time
MONITOR

Storage Monitoring¶

Via Azure Portal¶

1. Go to Storage Account → Metrics
2. Add metrics:
3. Transactions (request count)
4. Success E2E Latency (response time)
5. Used Capacity (storage usage)
6. Egress (data transfer out)

Storage Analytics¶

Large Objects:

# List largest files in presentations container
az storage blob list \
  --account-name slidefactoryprod \
  --container-name presentations \
  --query "[].{name:name, size:properties.contentLength}" \
  --output table \
  | sort -k2 -rn | head -20

Storage Usage by Container:

# Get size of each container
for container in presentations templates documents; do
  echo -n "$container: "
  az storage blob list \
    --account-name slidefactoryprod \
    --container-name $container \
    --query "sum([].properties.contentLength)" \
    --output tsv | awk '{printf "%.2f GB\n", $1/1024/1024/1024}'
done

Cost Monitoring¶

View Current Costs¶

1. Go to Azure Portal → Cost Management → Cost analysis
2. Filter by resource group: rg-slidefactory-prod
3. Group by: Service name or Resource

Expected Monthly Costs: - Preview: ~\(200/month - Production: ~\)500/month

Cost Alerts¶

Set up budget alerts: 1. Go to Cost Management → Budgets 2. Create budget with threshold (e.g., $600/month for production) 3. Set alerts at 80%, 90%, 100% of budget

Performance Baselines¶

Expected Performance¶

Daily Checks (Production)¶

• Check Container App status (green in Azure Portal)
• Review errors in Application Insights (< 5 errors/hour)
• Check database connections (no connection errors)
• Verify Redis is responsive (< 10ms latency)
• Review overnight deployments (if any)

Weekly Checks (Production)¶

• Review performance trends (response times stable)
• Check storage usage growth (plan cleanup if needed)
• Review database size (plan optimization if > 50GB)
• Audit security alerts (review any security warnings)
• Review costs (within budget)

Troubleshooting Dashboard¶

Quick Links: - Azure Portal - Application Insights - Container Apps - S5 Troubleshooting Guide

Related Documentation¶

• Infrastructure Details
• Deployment Process
• Troubleshooting