Looking for patterns with Flink CEP library
summary
Flink CEP 1.3 demonstrates concrete pattern features: optionals, quantifiers (one or more / unbounded), and kleene closures applied to clickstream and offer/payment sequences, with the library turning patterns into a non-deterministic finite automaton using three transition types: ignore, take and proceed. The engine stores matched events in a shared buffer, uses a within operator as a TTL to free partial matches, and shows how dangling computation states grow and how time or event-based stop conditions reclaim memory. Planned: SQL MATCH_RECOGNIZE and dynamic pattern streams.
AI-written summary · about
chapters (9)
- 0:00Introduction and motivation
- 1:28What is complex event processing
- 2:55Use case with insurance process
- 5:50Succession operators in Flink CEP
- 8:19Quantifiers in patterns
- 9:25How Flink CEP works internally
- 16:46Future features of Flink CEP
- 18:09Summary and closing remarks
- 18:52Q&A session
related talks
27mData enrichment patterns with Apache Flink
20mUsing Flink for Versatile Feature Engineering Pipelines
27mDataEngBytes 2023 - MEL-T1-08 - Troy Sellers
20mSemblance: Feature Generation In Real-Time and Batch Without Time-Travel
18mThe role Stream Processing plays for an Opti Channel Experience
39mBetter, Faster, Stronger Streaming: Your First Dive into Flink SQL
25mData: Moving with the flow
20mA Reactive Fraud Monitoring Engine for Instant Payments