Book
Database Internals
Alex Petrov
Summary
Alex Petrov wrote this because most engineers evaluate databases by vendor marketing rather than by what their storage and replication subsystems actually do, which is usually where the real differences sit. Part I builds up storage engines from B-Trees to log-structured merge trees, covering write-ahead logging and crash recovery along the way; Part II turns to distributed systems, working through failure detection, leader election, replication and consistency models, and consensus algorithms including Paxos and Raft. The book stays close to the mechanics rather than any single product, so the mental model still holds once the database in front of you has changed.
Target Readers
- Senior engineers choosing between database engines who want to evaluate storage and replication subsystems directly, not by marketing claims
- Architects making consistency and replication tradeoffs for a distributed data system who need the mechanics behind terms like leaderless replication
- Developers who have used B-Tree databases for years and want to understand why LSM-tree engines make different tradeoffs
Tags
Colophon
- Publisher
- オライリー・ジャパン
- ISBN
- 978-4873119540
- Published
- Jul 2021
- List price
- ¥4,180incl. taxMay differ from the actual selling price on Amazon
Get this book
* The link above is an advertisement via Amazon Associates.Related Books
Next Books
- Prerequisite
Designing Data-Intensive Applications
Martin Kleppmann
Reason: Only after understanding single-node storage-engine internals do discussions of consistency, replication, and partitioning across multiple nodes become grounded. Kleppmann's 'Designing Data-Intensive Applications' is the capstone that systematizes the principles of distributed data systems, placing internals knowledge as a required prerequisite.
- Recommended
Database Reliability Engineering
Laine Campbell, Charity Majors
Reason: Understanding the internals of storage engines and replication lets you predict, from structure, what to monitor and what can break in production. 'Database Reliability Engineering' bridges that internal knowledge into operational design—capacity planning, incident response, data integrity.