Monitoring Kubernetes Clusters with OpenTelemetry is quite mainstream now, In this session we will dive deeper into how to work configure different stateful processors, i.e. Deduplication Processor / GroupByAttr Processor.

Usecases being ...
1. Dropping redundant data : That can help in saving ingestion + storage costs.
2. Aggregating Service-Level Insights : Distributed applications running on Kubernetes often generate scattered spans, With correct grouping of data we can also achieve tenant-aware monitoring
3. Better Alerting and Anomaly Detection :
Once data is free from redundancy (1st point) and have better co-relation (2nd point), The alerts will only be triggered once necessary.

In my talk, I'll cover how to achieve all the points mentioned above. Also, I have been working as a Staff Engineer in a SaaS Observability platform, so I can share some case studies, where we achieved better results with these configurations.