« back

Readings

Below you can find a list of recommended readings and resources the lectures are based on. Parts of the resources listed here will be given as reading assignments during the course. You should be able to access all of these for free when connected to the campus network. Please contact the instructor if any of the links below is broken.

Books

• Data Stream Management: Processing High-Speed Data Streams
• Introduction to Apache Flink: Stream Processing for Real Time and Beyond
• Mining of Massive Datasets, Chapter 4.

Papers

Each team will be assigned one of the following papers to review and present in class:

• BriskStream: Scaling Data Stream Processing on Shared-Memory Multicore Architectures
• TimeCrypt: Encrypted Data Stream Processing at Scale with Cryptographic Access Control
• Move Fast and Meet Deadlines: Fine-grained Real-time Stream Processing with Cameo
• Ananke: A Streaming Framework for Live Forward Provenance
• Analyzing Efficient Stream Processing on Modern Hardware
• LightSaber: Efficient Window Aggregation on Multi-core Processors
• EdgeWise: A Better Stream Processing Engine for the Edge
• Shared Arrangements: practical inter-query sharing for streaming dataflows
• FineStream: Fine-Grained Window-Based Stream Processing on CPU-GPU Integrated Architectures
• WASP: Wide-area Adaptive Stream Processing
• Turbine: Facebook’s Service Management Platform for Stream Processing
• Flextream: Adaptive Compilation of Streaming Applications for Heterogeneous Architectures

Paper assignments will be coordinated via Piazza. You are allowed and encouraged to propose a paper you like if it is not in the list above.

Blog posts

• Streaming 101: The world beyond batch
• Streaming 102: The world beyond batch
• Watermarks, Tables, Event Time, and the Dataflow Model
• Kafka streams take on watermarks and triggers
• Why I Recommend My Clients NOT Use KSQL and Kafka Streams
• A Deep-Dive into Flink’s Network Stack
• Managing Large State in Apache Flink: An Intro to Incremental Checkpointing
• An Overview of End-to-End Exactly-Once Processing in Apache Flink (with Apache Kafka, too!)
• A Deep Dive into Rescalable State in Apache Flink
• How To Size Your Apache Flink Cluster: A Back-of-the-Envelope Calculation
• Tracking progress in Timely Dataflow