Monitor MediaTailor CDN operations and performance - AWS Elemental MediaTailor
Essential CDN performance metricsSet up monitoring toolsImplementation checklistRelated topics

Monitor MediaTailor CDN operations and performance

Effective monitoring of your AWS Elemental MediaTailor and content delivery network (CDN) integration ensures reliable content delivery, optimal performance, and quick issue detection. This monitoring approach applies to all MediaTailor implementations including server-side ad insertion (SSAI), channel assembly, and combined workflows.

Monitoring your CDN integration enables you to:

  • Detect and resolve issues before they impact viewers

  • Track key performance indicators and maintain service quality

  • Maintain optimal cache performance and reduce origin load

  • Ensure ad insertion success rates meet business requirements

Essential CDN performance metrics

Track these core metrics to evaluate your CDN's effectiveness with MediaTailor implementations:

Cache hit ratio

What it measures: The percentage of requests served from the CDN cache versus from origin.

Target values:

  • Content segments: 95% or higher cache hit ratio

  • Ad segments: 90% or higher cache hit ratio

  • Manifests: Varies by implementation (personalized manifests should not be cached)

Why it matters: Higher cache hit ratios reduce origin load, improve response times, and lower bandwidth costs.

Origin request volume

What it measures: The number of requests reaching your MediaTailor origin servers.

Target pattern: Should remain low and stable, with occasional spikes for cache misses or new content.

Why it matters: High origin request volumes indicate caching inefficiencies and can impact MediaTailor performance.

Response latency

What it measures: How quickly the CDN responds to viewer requests.

Target values:

  • Cached content: Less than 100ms (P95)

  • Origin requests: Less than 500ms (P95)

  • Manifest requests: Less than 100ms (P95)

  • Segment requests: Less than 50ms (P95)

Why it matters: Low latency ensures smooth playback and good viewer experience.

Error rates

What it measures: The percentage of requests that result in HTTP errors.

Target values:

  • 4xx errors: Less than 0.1% of total requests

  • 5xx errors: Less than 0.01% of total requests

  • Origin errors: Less than 0.05% of origin requests

Why it matters: High error rates indicate configuration issues or service problems that impact viewer experience.

MediaTailor specific metrics

Monitor these MediaTailor metrics alongside CDN metrics for complete visibility:

Ad fill rates

Key metrics: Avail.FillRate and AdDecisionServer.FillRate

Target values: Above 90% for both metrics

Why it matters: Directly impacts ad revenue and viewer experience

Manifest generation performance

Key metrics: GetManifest.Latency and GetManifest.Errors

Target values: Latency under 200ms, error rate under 1%

Why it matters: Affects playback startup time and reliability

Ad decision server health

Key metrics: AdDecisionServer.Latency, AdDecisionServer.Errors, and AdDecisionServer.Timeouts

Target values: Latency under 1000ms, error rate under 5%, minimal timeouts

Why it matters: ADS performance directly affects ad insertion success

Set up monitoring tools

Configure these tools to monitor your MediaTailor and CDN integration effectively:

Amazon CloudWatch integration

Amazon CloudWatch provides the foundation for monitoring your MediaTailor and CDN integration:

MediaTailor metrics

MediaTailor automatically publishes metrics to CloudWatch that track requests, responses, and errors. Key metrics include:

  • RequestCount: Total number of requests to MediaTailor

  • ResponseTime: MediaTailor response latency

  • 4xxErrorCount and 5xxErrorCount: Error tracking

For a complete list of MediaTailor metrics, see Monitoring AWS Elemental MediaTailor with Amazon CloudWatch metrics.

CDN metrics

Enable CDN metrics collection in CloudWatch including:

  • Cache hit ratios for different content types

  • Origin request counts and response times

  • Error rates by status code

For CloudFront specific metrics, enable detailed monitoring in your distribution settings.

Dashboard configuration

Create dashboards that provide visibility into your MediaTailor and CDN performance:

  1. Create a unified dashboard: Combine MediaTailor and CDN metrics in a single CloudWatch dashboard for complete visibility.

  2. Organize by workflow: Group metrics by implementation type (SSAI, channel assembly, or combined workflows).

  3. Include key performance indicators:

    • Cache hit ratio trends over time

    • Response latency percentiles (P50, P95, P99)

    • Error rate trends and spikes

For detailed dashboard creation guidance, see CloudWatch dashboards.

Alert configuration

Configure alerts to detect issues before they impact viewers:

Critical alerts

Configure immediate alerts for severe issues:

  • High error rates: Alert when 5xx errors exceed 0.1% of requests over 5 minutes

  • Origin server issues: Alert when origin response time exceeds 2 seconds

  • Cache hit ratio drops: Alert when cache hit ratio falls below 70% for manifests or 85% for segments

Warning alerts

Configure early warning alerts for performance degradation:

  • Latency increases: Alert when P95 response time exceeds 200ms

  • Cache efficiency decline: Alert when cache hit ratio drops below 90% for segments

Implementation checklist

Use this checklist to ensure comprehensive monitoring coverage:

  1. Metrics collection:

    • ✓ MediaTailor metrics enabled in CloudWatch

    • ✓ CDN detailed monitoring enabled

  2. Dashboard setup:

    • ✓ Unified CloudWatch dashboard created

    • ✓ Key metrics visualized with appropriate time ranges

    • ✓ Dashboard shared with relevant teams

  3. Alert configuration:

    • ✓ Critical alerts configured with immediate notification

    • ✓ Warning alerts set up for early detection

    • ✓ Alert escalation procedures documented

  4. Operational procedures:

    • ✓ Incident response procedures documented

    • ✓ Regular review schedule established

    • ✓ Team training completed

For additional guidance on specific monitoring scenarios: