Keynote | Flink Ahead 2.0: The Sequel
summary
Flink SQL powers continuous materialized views and Ververica Platform 2.3 ships an integrated editor, catalog, and autopilot for horizontal scaling. Live demos show JSON orders read from Kafka, converted to Avro, watermarked with 15s lateness, tumbling/hop windows (15s/1m with 10s slide), and an over-window running total. Text is tokenized by a table function and scored by a scalar UDF (-10..10) for sliding sentiment, pattern-matched via MATCH_RECOGNIZE, and published with a statement set to Kafka and Elasticsearch; Jupyter WP 0.1 runs Flink SQL cells. Requirements: Kubernetes + blob storage.
AI-written summary · about
chapters (16)
- 0:00Introduction to Flink SQL
- 2:17What is Flink SQL?
- 4:02Stream processing use cases
- 5:28Demo setup in Verica Platform
- 6:54Creating and querying tables
- 9:20Data cleansing with Flink SQL
- 12:13Streaming analytics with windows
- 13:37Running totals with over windows
- 14:42Sentiment analysis demo
- 22:24Advanced SQL features
- 23:50Joining with Hive tables
- 26:39Verica Platform enhancements
- 28:43Jupyter notebook support
- 29:55Getting started guide
- 30:15Future of Flink SQL
- 33:22Q&A and resources
related talks
35mKeynote: Stream analytics made real with Pravega and Apache Flink
29mA brief history of time with Apache Flink
27mArchitecting Sub-Second Analytics Pipelines
32mDiscover the Modern Way to Build Real-Time Analytics Solutions at Any Scale
36mBuilding a Self-Service Streaming Platform at Pinterest
20mBuilding realtime efficient queries in an IOT platform
43mBuilding a community fountain around your data stream
22mCloudera's Data-in-Motion vision