Skip to main content

Best Practices for Observability with Grafana Cloud using OpenTelemetry

Introduction​

This document outlines the best practices for implementing observability in your applications using Grafana Cloud and the OpenTelemetry standard.

Logging Best Practices​

Log Level​

  • Default Log Level: Set the log level for applications to debug by default. This ensures comprehensive logging, which is crucial for effective monitoring and troubleshooting.
  • Log Level Filtering: If log level filtering is required for cost considerations, perform this filtering in the Grafana Alloy pipeline rather than at the application level.

Log Labels​

  • Label Format: Use PascalCase for all log labels. This ensures consistency and readability across your logs.

Metrics Best Practices​

  • Standard Metrics: Use standard metrics provided by OpenTelemetry to ensure compatibility and ease of integration with Grafana Cloud.
  • Custom Metrics: When defining custom metrics, ensure they are well-documented and follow a consistent naming convention.

Tracing Best Practices​

  • Trace Context Propagation: Ensure trace context is propagated across all services to maintain a complete view of transactions.
  • Span Naming: Use descriptive names for spans to make traces easier to understand and analyze.

Conclusion​

Following these best practices will help you achieve effective observability with Grafana Cloud and OpenTelemetry, ensuring your applications are well-monitored and issues can be quickly identified and resolved.