Add Observability for Spring Data Cassandra with Micrometer Integration

[ranzyblessings]

Summary

We need to implement full observability for Spring Data Cassandra by integrating Micrometer for monitoring and trace collection. This will enable us to track Cassandra operations and expose them to observability platforms such as Jaeger, Prometheus, and Grafana.

Goals

1. Integrate Micrometer with Spring Data Cassandra for Cassandra command tracing and metrics.
2. Enable full distributed tracing with Jaeger to visualize Cassandra interactions alongside application traces.
3. Expose Cassandra metrics to Prometheus for scraping and Grafana for visualization.

Requirements

1. Jaeger Integration:

   • Ensure Cassandra queries (select, insert, update, delete, etc.) are captured and traced in Jaeger.
   • Ability to see the performance and latency of Cassandra operations alongside application traces in the Jaeger UI.

2. Prometheus & Grafana Integration:

   • Expose Cassandra command metrics (e.g., operation count, latency, error rates) through Micrometer.
   • Ensure Prometheus can scrape these metrics for long-term storage and analysis.
   • Create Grafana dashboards to visualize Cassandra-specific metrics, such as:
     • Operations per second (insert, query, update, delete)
     • Latency for Cassandra queries
     • Error rates for Cassandra operations
     • Connection counts and resource utilization (memory, CPU, disk I/O)

3. Configuration Details:

   • Use Micrometer’s Cassandra integration to capture the necessary metrics and traces.
   • Ensure Spring Micrometer Tracing is configured with Jaeger as the tracing system.
   • Configure Prometheus to scrape metrics from Spring Boot Actuator endpoints, specifically Cassandra metrics.

Expected Outcome

1. Cassandra operation traces should appear in Jaeger with detailed spans for each command type (select, insert, update).
2. Prometheus should collect and store Cassandra-specific metrics, and Grafana should display them in a usable dashboard.
3. Team members will have comprehensive visibility into both Cassandra performance and the overall health of the application.

Priority

This is a high priority as it will improve observability, help identify bottlenecks, and ensure better monitoring for our production systems.

References

• Spring Data Cassandra Observability Documentation
• OpenTelemetry
• Jaeger Tracing

Notes

Please refer to Micrometer’s documentation for details on Cassandra integration, Jaeger, and Prometheus configuration options. Ensure the tracing configurations are optimized to minimize overhead.