06 - Performance Efficiency

Caching strategies, compression, routing optimization, and latency reduction for Azure Front Door

---

🎯 Performance Efficiency Principles

Principle	Front Door Implementation
Optimize for latency	Edge caching, anycast routing, compression
Scale automatically	Global edge network, no capacity planning
Cache strategically	Right TTL, query string handling, cache key
Reduce payload size	Compression, minification, image optimization

---

✅ Performance Checklist

#	Recommendation	Impact
1	Enable caching for static content	🔴 High
2	Configure appropriate cache TTLs	🔴 High
3	Enable compression (gzip/brotli)	🔴 High
4	Use latency-based routing	🟡 Medium
5	Optimize query string caching	🟡 Medium
6	Set appropriate origin timeouts	🟡 Medium
7	Minimize redirects	🟢 Low
8	Use HTTP/2	🟢 Low

---

⚡ Latency Optimization

How Front Door Reduces Latency

Latency Components

Component	Typical Latency	Optimization
Client → Edge POP	10-50ms	Anycast routing (automatic)
TLS Handshake	20-50ms	TLS 1.3, session resumption
WAF Processing	1-5ms	Tune rules, exclusions
Cache Lookup	<1ms	High cache hit ratio
Edge → Origin	50-200ms	Cache more, Private Link
Origin Processing	Variable	Optimize application

Latency-Based Routing

resource originGroup 'Microsoft.Cdn/profiles/originGroups@2023-05-01' = {
  name: 'og-latency-optimized'
  parent: frontDoor
  properties: {
    loadBalancingSettings: {
      sampleSize: 4
      successfulSamplesRequired: 3
      additionalLatencyInMilliseconds: 50  // Route to origin within 50ms of fastest
    }
    healthProbeSettings: {
      probePath: '/health'
      probeProtocol: 'Https'
      probeRequestType: 'HEAD'
      probeIntervalInSeconds: 30
    }
  }
}

Latency Sensitivity Settings:

Value	Behavior	Use Case
0ms	Always route to fastest origin	Latency-critical apps
50ms	Route to any origin within 50ms of fastest	Balanced (recommended)
200ms+	Distribute more evenly	Load distribution priority

---

💾 Caching Strategies

Cache Hit Ratio Impact

Caching Configuration

resource staticRoute 'Microsoft.Cdn/profiles/afdEndpoints/routes@2023-05-01' = {
  name: 'route-static'
  parent: endpoint
  properties: {
    originGroup: { id: originGroup.id }
    patternsToMatch: ['/static/*', '/assets/*', '/images/*']
    cacheConfiguration: {
      // Cache behavior
      cacheBehavior: 'OverrideAlways'        // Override origin cache headers
      cacheDuration: '7.00:00:00'            // 7 days
      
      // Query string handling
      queryStringCachingBehavior: 'IgnoreQueryString'  // Ignore for static
      
      // Compression
      compressionSettings: {
        isCompressionEnabled: true
        contentTypesToCompress: [
          'text/html'
          'text/css'
          'text/javascript'
          'application/javascript'
          'application/json'
          'image/svg+xml'
          'application/xml'
        ]
      }
    }
  }
}

resource apiRoute 'Microsoft.Cdn/profiles/afdEndpoints/routes@2023-05-01' = {
  name: 'route-api'
  parent: endpoint
  properties: {
    originGroup: { id: originGroup.id }
    patternsToMatch: ['/api/*']
    cacheConfiguration: {
      cacheBehavior: 'HonorOrigin'           // Respect origin Cache-Control
      queryStringCachingBehavior: 'UseQueryString'  // Include query in cache key
    }
  }
}

Query String Caching Behaviors

Behavior	Cache Key	Use Case
IgnoreQueryString	/page.html	Static assets (same content regardless of query)
UseQueryString	/page.html?v=1.2.3	Versioned assets, API responses
IgnoreSpecifiedQueryStrings	Exclude specified params	Ignore tracking params like utm_source
IncludeSpecifiedQueryStrings	Only include specified params	Cache only on relevant params

Cache TTL Recommendations

Content Type	TTL	Cache-Control Header
Static assets (CSS, JS)	7-30 days	max-age=2592000, immutable
Images	7-30 days	max-age=2592000
Fonts	30-365 days	max-age=31536000, immutable
HTML pages	1-5 minutes	max-age=300, s-maxage=60
API responses (cacheable)	10-60 seconds	max-age=60, s-maxage=10
User-specific content	Don't cache	private, no-store

---

🗜️ Compression

Compression Impact

Enable Compression

compressionSettings: {
  isCompressionEnabled: true
  contentTypesToCompress: [
    // Text-based formats (high compression ratio)
    'text/plain'
    'text/html'
    'text/css'
    'text/javascript'
    'text/xml'
    
    // Application formats
    'application/javascript'
    'application/json'
    'application/xml'
    'application/xhtml+xml'
    'application/rss+xml'
    'application/atom+xml'
    
    // Other compressible formats
    'image/svg+xml'
    'image/x-icon'
    'font/ttf'
    'font/otf'
    'application/vnd.ms-fontobject'
  ]
}

Compression Best Practices

Do	Don't
✅ Compress text-based content	❌ Compress already-compressed formats (JPEG, PNG, MP4)
✅ Use brotli for modern browsers	❌ Compress small files (< 1KB overhead)
✅ Set Vary: Accept-Encoding	❌ Compress encrypted content

---

🔄 Routing Optimization

Path-Based Routing for Performance

// Optimized multi-route configuration
var routes = [
  {
    name: 'static-assets'
    patterns: ['/static/*', '/assets/*', '/_next/static/*']
    originGroup: staticOriginGroup
    cacheDuration: '30.00:00:00'
    queryString: 'IgnoreQueryString'
  }
  {
    name: 'api'
    patterns: ['/api/*']
    originGroup: apiOriginGroup
    cacheDuration: null  // Honor origin
    queryString: 'UseQueryString'
  }
  {
    name: 'default'
    patterns: ['/*']
    originGroup: webOriginGroup
    cacheDuration: '00:05:00'
    queryString: 'UseQueryString'
  }
]

Origin Selection Strategy

Strategy	Configuration	Best For
Lowest Latency	additionalLatencyInMilliseconds: 0	Real-time apps
Balanced	additionalLatencyInMilliseconds: 50	General web apps
Weighted	weight: 1000/500	Gradual rollout
Priority Failover	priority: 1/2	DR scenarios

---

📊 Performance Monitoring

Key Performance Metrics

Metric	Target	Query
Cache Hit Ratio	> 80%	See below
Origin Latency	< 200ms	See below
Total Latency	< 100ms (cached)	See below
5XX Error Rate	< 0.1%	See below

Performance Queries (KQL)

Cache Hit Ratio

AzureDiagnostics
| where Category == "FrontDoorAccessLog"
| where TimeGenerated > ago(24h)
| summarize 
    Total = count(),
    Hits = countif(cacheStatus_s == "HIT"),
    Misses = countif(cacheStatus_s == "MISS"),
    Other = countif(cacheStatus_s !in ("HIT", "MISS"))
| extend HitRatio = round((Hits * 100.0) / Total, 2)
| project HitRatio, Total, Hits, Misses, Other

Latency Percentiles

AzureDiagnostics
| where Category == "FrontDoorAccessLog"
| where TimeGenerated > ago(1h)
| extend LatencyMs = toint(timeTaken_s) * 1000
| summarize 
    P50 = percentile(LatencyMs, 50),
    P75 = percentile(LatencyMs, 75),
    P95 = percentile(LatencyMs, 95),
    P99 = percentile(LatencyMs, 99)
    by bin(TimeGenerated, 5m)
| render timechart

Slow Requests Analysis

AzureDiagnostics
| where Category == "FrontDoorAccessLog"
| where TimeGenerated > ago(1h)
| extend LatencyMs = toint(timeTaken_s) * 1000
| where LatencyMs > 1000  // Requests > 1 second
| project 
    TimeGenerated,
    requestUri_s,
    LatencyMs,
    cacheStatus_s,
    originName_s,
    httpStatusCode_s
| order by LatencyMs desc
| take 100

Cache Miss Analysis

// Identify high-volume cache misses (optimization opportunities)
AzureDiagnostics
| where Category == "FrontDoorAccessLog"
| where cacheStatus_s == "MISS"
| summarize MissCount = count() by requestUri_s
| order by MissCount desc
| take 50

---

⚙️ Rules Engine for Performance

URL Rewrite (Reduce Origin Load)

resource ruleSet 'Microsoft.Cdn/profiles/ruleSets@2023-05-01' = {
  name: 'performance-rules'
  parent: frontDoor
}

// Redirect to CDN for known static patterns
resource staticRedirectRule 'Microsoft.Cdn/profiles/ruleSets/rules@2023-05-01' = {
  name: 'redirect-static-to-cdn'
  parent: ruleSet
  properties: {
    order: 1
    conditions: [
      {
        name: 'UrlPath'
        parameters: {
          typeName: 'DeliveryRuleUrlPathMatchConditionParameters'
          operator: 'BeginsWith'
          matchValues: ['/legacy-static/']
        }
      }
    ]
    actions: [
      {
        name: 'UrlRewrite'
        parameters: {
          typeName: 'DeliveryRuleUrlRewriteActionParameters'
          sourcePattern: '/legacy-static/'
          destination: '/static/'
          preserveUnmatchedPath: true
        }
      }
    ]
  }
}

Cache Override Rules

// Force cache for specific paths regardless of origin headers
resource forceCacheRule 'Microsoft.Cdn/profiles/ruleSets/rules@2023-05-01' = {
  name: 'force-cache-assets'
  parent: ruleSet
  properties: {
    order: 2
    conditions: [
      {
        name: 'UrlFileExtension'
        parameters: {
          typeName: 'DeliveryRuleUrlFileExtensionMatchConditionParameters'
          operator: 'Equal'
          matchValues: ['js', 'css', 'woff2', 'woff', 'ttf']
        }
      }
    ]
    actions: [
      {
        name: 'CacheExpiration'
        parameters: {
          typeName: 'DeliveryRuleCacheExpirationActionParameters'
          cacheBehavior: 'Override'
          cacheType: 'All'
          cacheDuration: '30.00:00:00'
        }
      }
    ]
  }
}

---

🚀 Performance Optimization Patterns

Pattern 1: Static Asset Optimization

Implementation:

• Use versioned file names or query strings
• Set Cache-Control: max-age=2592000, immutable
• Configure Front Door to IgnoreQueryString or UseQueryString

Pattern 2: API Response Caching

// Short-lived cache for API responses
cacheConfiguration: {
  cacheBehavior: 'OverrideIfOriginMissing'
  cacheDuration: '00:01:00'  // 1 minute
  queryStringCachingBehavior: 'UseQueryString'
}

Best Practices:

• Cache GET requests only
• Include query params in cache key
• Use Vary header for different response formats
• Set short TTL (30-60 seconds)

Pattern 3: Stale-While-Revalidate

Configure origin to return:

Cache-Control: max-age=60, stale-while-revalidate=300

This allows Front Door to:

1. Serve cached content immediately
2. Fetch fresh content in background
3. Update cache for next request

---

📋 Performance Summary

Optimization	Implementation	Expected Impact
Enable caching	Route-level cache config	50-90% latency reduction
Compression	Gzip/brotli for text content	60-80% bandwidth reduction
Latency routing	additionalLatencyInMilliseconds: 50	10-30% latency reduction
Query string optimization	IgnoreQueryString for static	Higher cache hit ratio
Path-based routing	Separate routes by content type	Better cache efficiency
Rules Engine	Force cache headers	Consistent caching

---

🔗 References

Resource	Link
Performance Efficiency	WAF Performance Efficiency
Caching	Front Door Caching
Compression	Enable Compression
Rules Engine	Front Door Rules Engine

---

Previous: 05 - Operational Excellence