jaeger

Best Practices

• Configure appropriate sampling strategies (probabilistic, rate-limiting, adaptive)
• Use Elasticsearch or Cassandra for production storage backends
• Set span storage TTL for automatic data expiration
• Enable OTLP receiver for OpenTelemetry compatibility
• Configure query service with appropriate max_clock_skew_adjustment
• Use TLS for collector endpoints in production
• Set appropriate queue sizes and worker counts for collector throughput
• Use OTLP receivers for OpenTelemetry compatibility
• Configure appropriate batch processor settings for throughput
• Set index prefix for Elasticsearch to avoid conflicts
• Enable health check endpoints for liveness probes
• Use production-grade storage backends, not in-memory

Documentation Sources