Processors

Understanding the available processors in Serverless OTLP Forwarder.

Overview

Processors are responsible for handling telemetry data from CloudWatch Logs and forwarding it to collectors. The Serverless OTLP Forwarder supports two types of processors, each designed for specific use cases.

OTLP Stdout Processor

The OTLP Stdout processor is the primary processor for handling OpenTelemetry data. It processes OTLP-formatted log entries from CloudWatch Logs and forwards them to OTLP collectors.

Features

  • Processes OTLP data written to stdout by instrumented functions
  • Supports both JSON and protobuf formats (based on application configuration)
  • Handles compressed data (if configured by the application)
  • Forwards to any OTLP-compatible collector
  • Supports multiple collectors in parallel

Authentication

The processor supports two authentication methods:

  1. Custom Headers:
    • Uses headers stored in AWS Secrets Manager
    • Compatible with OTEL_EXPORTER_OTLP_HEADERS format
    • Supports any collector-specific authentication scheme
  2. AWS SigV4:
    • For sending data to AWS Application Signals OTLP endpoint
    • Uses the forwarder’s IAM role
    • No additional credentials needed

For detailed configuration instructions, including how to set up multiple collectors, see the Collector Configuration section.

AWS Spans Processor (Experimental)

This processor is experimental and should not be used in production environments.

The AWS Spans processor provides compatibility between AWS Application Signals and standard OTLP collectors.

Features

  • Reads trace data from the aws/span log group
  • Converts AWS Application Signals format into OTLP JSON
  • Forwards converted data to standard OTLP collectors
  • May have limitations and known issues

Best Practices

  1. Processor Selection:
    • Use the OTLP Stdout processor for standard observability platforms
    • Consider the experimental status of the AWS Spans processor
  2. Authentication:
    • Store collector credentials securely in Secrets Manager
    • Use SigV4 when sending data to AWS Application Signals OTLP endpoint
    • Rotate API keys regularly
  3. Performance:
    • Configure appropriate OpenTelemetry SDK settings in your applications
    • Use the BatchSpanProcessor in your instrumented functions
    • Consider compression for large payload volumes
  4. Monitoring:
    • Monitor Lambda execution metrics
    • Use the generated OpenTelemetry traces for operational insights
    • Set up appropriate alerting on Lambda errors