A practical framework for choosing a database, starting with the recommendation to default to PostgreSQL unless there's a measurable reason not to. Covers how to analyze workloads (OLTP vs OLAP, access patterns, sargable predicates), consistency models (ACID vs BASE, CAP vs PACELC, the spectrum from linearizability to eventual consistency), and scaling strategies. Includes a comparison table of database families (relational, document, key-value, wide-column, graph, time-series, vector, search, distributed SQL) with honest notes on tradeoffs. Emphasizes hidden operational costs of polyglot persistence, licensing risks, and the value of 'boring technology.' Ends with an 8-question decision checklist.
Table of contents
Start with the workload, not the databaseConsistency and scale: where the real tradeoffs liveThe database families and what each is actually forThe costs nobody budgets forA decision checklistKey takeaways922 Impressions