Book notes · Martin Kleppmann

Designing Data-Intensive Applications

Chapter-by-chapter notes biased towards what to do rather than how it works: a quick look-up when you need a detail on the fly, or a one-hour recap of the whole book before an interview.

12 chapters about 5 hours end to end

Start reading

1Foundations of data systems

The ideas that apply to any data system, on one machine or many.

4 chapters

Chapter 1 26 minReliable, Scalable, and Maintainable ApplicationsFaults versus failures, load parameters and tail latency, scaling up versus out, and the three habits (operability, simplicity, evolvability) that keep a system maintainableRead
Chapter 2 35 minData Models and Query LanguagesRelational, document and graph models, the object-relational mismatch, schema-on-read versus schema-on-write, declarative versus imperative queries, and why the network model failedRead
Chapter 3 48 minStorage and RetrievalHash indexes, SSTables and LSM-trees versus B-trees, secondary and in-memory indexes, OLTP versus OLAP, star schemas and column-oriented storage with compressionRead
Chapter 4 23 minEncoding and EvolutionBackward and forward compatibility, JSON vs Thrift, Protocol Buffers and Avro, schema evolution rules, and dataflow through databases, RPC and message brokersRead

2Distributed data

What changes when the data lives on more than one machine.

5 chapters

3Derived data

Integrating systems with batch and stream processing.

3 chapters

Designing Data-Intensive Applications

1Foundations of data systems

2Distributed data

3Derived data

Search

Designing Data-Intensive Applications

1Foundations of data systems

2Distributed data

3Derived data