Why X-Ray Customers Should Adopt Application Signals + Transaction Search

The Evolution of Observability Needs

As applications have grown in complexity and scale, customer observability requirements have evolved significantly. While AWS X-Ray has served as a reliable distributed tracing solution, the modern application landscape demands more comprehensive visibility.

Technical Architecture Differences

X-Ray Traditional Approach:

Application Signals + Transaction Search:

Key Migration Benefits

Capability	X-Ray	Application Signals + Transaction Search
Data Ingestion	100% of transactions (when configured)	100% of transactions (when configured)
Throughput Limits	Subject to X-Ray service quotas at high volume	Higher throughput capacity with CloudWatch Logs
Cost Model	Per-trace pricing (expensive at 100%)	Application Signals Bundled pricing
Storage Format	X-Ray proprietary format	OpenTelemetry standard format
Storage Backend	X-Ray optimized storage	CloudWatch Logs with selective indexing
Analytics	X-Ray console only	Transaction Search + X-Ray trace analytics
Query Capabilities	X-Ray console and APIs	Transaction Search visual analytics + X-Ray
Indexing	All traces indexed	Selective indexing (configurable %)
Business Context	Limited custom attributes	Rich OTEL span attributes + business context

Primary Value Propositions

1. Higher Throughput and Scalability

• CloudWatch Logs handles higher throughput than X-Ray, enabling customers to track all application events without hitting service limits
• Logs as storage for trace data removes X-Ray's throughput constraints for high-volume applications
• Scalable infrastructure designed for massive log ingestion volumes

2. Enhanced Analytics and Integration Capabilities

• Native CloudWatch Logs features available for span data analysis:
  • Metrics Filters: Create custom metrics from span attributes and patterns
  • Subscription Filters: Stream span data to other AWS services (Lambda, Kinesis, etc.)
  • Log Insights: Advanced querying capabilities beyond traditional trace analysis
• Transaction Search provides advanced visual query interface for span-level analytics
• OTEL format enables richer business context in spans with custom attributes

3. Cost Effective 100% Sampling

• Bundled pricing makes complete visibility cost-effective compared to per-trace X-Ray pricing. Please see Example 13 in CloudWatch pricing page
• Predictable costs based on data volume, not trace count
• Selective indexing optimizes storage costs while maintaining complete data access

Leveraging CloudWatch Logs Features with Span Data

Since Transaction Search stores span data in CloudWatch Logs (aws/spans log group), you can leverage all native CloudWatch Logs capabilities:

Metrics Filters:

# Create custom metrics from span attributes
aws logs put-metric-filter \
  --log-group-name "aws/spans" \
  --filter-name "HighLatencyRequests" \
  --filter-pattern '[timestamp, request_id, span_id, trace_id, duration > 5000]' \
  --metric-transformations \
    metricName=HighLatencySpans,metricNamespace=CustomApp/Performance,metricValue=1

Subscription Filters:

# Stream span data to Lambda for real-time processing
aws logs put-subscription-filter \
  --log-group-name "aws/spans" \
  --filter-name "ErrorSpanProcessor" \
  --filter-pattern '[..., status_code="ERROR"]' \
  --destination-arn "arn:aws:lambda:region:account:function:ProcessErrorSpans"

Log Insights Queries:

-- Find all spans with specific business attributes
fields @timestamp, attributes.customer_id, attributes.order_value, duration
| filter attributes.service_name = "payment-service"
| filter attributes.customer_tier = "premium"
| stats avg(duration) by attributes.customer_id
| sort avg(duration) desc

Integration Opportunities:

• Real-time Alerting: Use subscription filters to trigger Lambda functions for immediate incident response
• Business Intelligence: Export span data to analytics platforms via Kinesis Data Streams
• Custom Dashboards: Create CloudWatch dashboards using metrics derived from span attributes
• Compliance Auditing: Use Log Insights to query spans for regulatory compliance reporting