From Reactive to Proactive: Event-Driven Architectures for Modern...
summary
Data Prepper generates alerts on data in motion by producing metrics and events in parallel to ingestion, cutting detection from minutes to seconds while avoiding full indexing. It provides Prometheus sync and SQS sync for alert-friendly outputs, carries exemplars with histogram and APM service‑map metrics for immediate context, and runs anomaly detection plus an ML inference processor to emit similarity‑based events. Custom processors or a Lambda processor let pipelines output nothing until a rule or model actually fires.
AI-written summary · about
chapters (11)
- 0:00Event-driven architectures intro
- 1:05Observability and security goals
- 2:29From reactive to proactive events
- 3:54Data Prepper overview
- 6:22Event-driven pipelines with Data Prepper
- 8:27Metrics and exemplars for alerts
- 10:37APM maps and anomaly detection
- 12:45ML similarity-based alerts
- 13:54Custom processors and Lambda
- 16:21Conclusion and next steps
- 16:41Q&A
related talks
37mSmart Data-driven Alerts (SDdA) and How They Make DevOps Life Easier
26mUsing Data Prepper for Observability Ingest
41mMonitoring Reactive Applications
08mPrometheus in 5 minutes
36mBeyond Sequential: A Recipe for Async Pipeline Observability and Alerting
33mAlerting in Cloud Native Environments
31mAlerting for Distributed Systems—A Tale of Symptoms and Causes, Signals and Noise
18mAlert Analytics