Book review of Designing Data Intensive Applications by Martin Kleppmann

Designing Data Intensive Applications


“Designing Data-Intensive Applications” by Martin Kleppmann is a comprehensive guide that dives deep into the architecture of systems that are built to handle large volumes of data. It provides a foundation for understanding the principles and challenges involved in designing such systems. Kleppmann covers a wide array of topics, from databases to distributed systems, focusing on how to make these systems reliable, scalable, and maintainable.


The book is structured around several key themes:

  • Data Models and Query Languages: Discusses the different ways data can be modeled (relational, document, graph, etc.) and the implications of these choices on the way data is queried and manipulated.
  • Storage and Retrieval: Examines how databases internally store and retrieve data, covering data structures like B-trees, LSM-trees, and the trade-offs involved in using them.
  • Encoding and Evolution: Explores how data and schemas change over time and how systems can evolve without causing disruptions to the applications that depend on them.
  • Distributed Systems: Looks into the challenges of consistency, replication, and partitioning in distributed environments, including discussions on CAP theorem and distributed transactions.
  • Consistency and Consensus: Delves deeper into the consistency models and protocols that ensure data integrity across distributed systems, including concepts like linearizability and quorum.
  • Batch and Stream Processing: Covers the processing models for handling data at rest and data in motion, respectively, discussing the architectures and patterns for building robust data processing pipelines.

Examples by the Chapter

The book uses a variety of examples to illustrate the concepts it covers:

  • Chapter 1: Introduces the complexity of modern data systems with examples from real-world applications that face issues of scale and reliability.
  • Chapter 2: Uses examples from popular databases like PostgreSQL and MongoDB to explain data models and query languages.
  • Chapter 3: Discusses storage engines, using MySQL as an example to explain B-trees and RocksDB for LSM-trees.
  • Chapter 4: Focuses on encoding techniques such as Protocol Buffers and Avro, using examples that highlight the importance of schema evolution.
  • Chapter 5: Explores replication in distributed systems with examples from Cassandra and Kafka.
  • Chapter 6: Examines partitioning and sharding strategies, using Amazon DynamoDB as a case study.
  • Chapter 7: Deals with transactions, using examples from Google’s Spanner to explain concepts like serializability and external consistency.
  • Chapter 8: Tackles the challenge of distributed system consistency, with a deep dive into the Raft consensus algorithm.
  • Chapter 9: Discusses the principles behind batch and stream processing, with examples from Hadoop and Apache Flink.


“Designing Data-Intensive Applications” by Martin Kleppmann is an essential read for anyone working with or designing systems that process large volumes of data. The book’s thorough exploration of fundamental concepts, combined with real-world examples, makes it an invaluable resource. It not only equips readers with the knowledge to design robust, efficient, and scalable data systems but also encourages a deeper understanding of the underlying principles that govern these systems. Whether you’re a student, a software engineer, or an architect, this book offers insights that will enhance your approach to data-intensive application design.