DeepRead
Cover of Streaming Systems

About the book

Stream processing of unbounded data: how to compute correct, reliable results on data that arrives continuously and out of order, covering the difference between when events occurred and when the system sees them. It sits between the batch and streaming worlds and draws on the Apache Beam and Google Cloud Dataflow model, with examples of windowing, watermarks, triggers, and state. Batch processing is a special case of streaming, so one model can handle both if it separates the questions of what is computed, where in event time, when in processing time, and how repeated results relate to each other. The book argues that event-time semantics, not arrival order, should define correctness, and that watermarks and triggers let systems trade latency against completeness explicitly. It also argues that stream processing needs durable state and consistent checkpointing to give exactly-once results.

Written by Claude from what it knows of this book.

What Hacker News says 50% recommend · 0% critical

0242020202120222020 peak · 1
This bookAverage book in Databases and distributed data

In the commenters’ own words

What Open Library readers say

No Open Library reader has rated it yet.

  • 2 want to read
  • 0 reading
  • 2 have read

Open Library’s page for the book

What critics say, via Wikipedia

No Wikipedia article about this book yet, so no critics to quote.

Reddit communities — coming later

Details

First published
2018
ISBN
9781491983874
Editions known
2 at Open Library
Open Library
OL19542624W

Big dataData transmission systemsDistributed processingElectronic data processingStreaming technology (Telecommunications)Streaming technology (telecommunications)

Explore in the mosaic

Read next: its neighbours on the mosaic