SQL and NoSQL Databases (Michael Kaufmann, Andreas Meier)（Z-Library）

【One-Line Pitch】 A structured textbook that teaches relational and non-relational data management side by side, so you can model, query, secure, and architect both SQL and NoSQL systems for big-data work. Best for students and practitioners who want one coherent comparison instead of scattered tutorials. 【Book Arc】 - **Opening (~0%–15%)**: Frames databases as information systems and business-critical assets, then introduces the relational model, SQL, and the RDBMS as the baseline vocabulary for everything later. - **Early (~15%–35%)**: Builds modeling skills across paradigms—entity-relationship design and its mapping into SQL schemas, then graph and document modeling with their own mapping rules. - **Middle (~35%–60%)**: Covers languages and security: relational algebra and SQL, Cypher for graphs, MQL for documents, plus access control, integrity constraints, ACID transactions, and soft-consistency ideas like BASE and the CAP theorem. - **Late (~60%–80%)**: Moves into system architecture—indexes, hashing, BSON, index-free adjacency, query optimization, MapReduce, layered architecture, and cloud databases. - **Ending (~80%–100%)**: Surveys post-relational territory: federated, temporal, multi-dimensional, object-relational, knowledge, and fuzzy databases, alongside data warehouse and data lake systems. 【Key Takeaways】 - **SQL remains foundational, not obsolete** (Early): The authors argue relational technology is more relevant than ever because big-data analysis still leans on SQL aggregation, even as NoSQL handles volume, variety, and velocity. - **Modeling differs by paradigm** (Early): Entity-relationship modeling maps to SQL schemas, while graph and document databases need their own mapping rules—so schema design is not a one-size-fits-all skill. - **Three query languages, one comparative lens** (Middle): SQL, Cypher, and MQL are presented in parallel, letting you see how relational, graph, and document thinking diverge in practice. - **Security is a first-class chapter** (Middle): Access control, integrity constraints, and transaction consistency are treated separately for SQL, Cypher, and MQL, not as an afterthought. - **ACID vs. BASE is a design trade-off** (Middle): The book contrasts strict transaction guarantees with soft consistency, the CAP theorem, and nuanced consistency settings for distributed data. - **Architecture explains performance** (Late): Indexes, hashing, BSON, index-free adjacency, query trees, and cost-based optimization connect data models to real system behavior. - **MapReduce bridges SQL and big data** (Late): Parallel processing with MapReduce is shown as a practical link between relational querying and distributed computation. - **Beyond the big three, many database families exist** (Ending): Federated, temporal, multi-dimensional, object-relational, knowledge, and fuzzy databases show the field's breadth beyond SQL, graph, and document systems. 【Reading Tips】 - Deep-read the first five chapters if you need a solid conceptual base; they carry the core modeling, language, security, and architecture material. - Skim the post-relational survey chapters on a first pass, then return when a specific database family becomes relevant to your work. - Use the parallel language sections (SQL, Cypher, MQL) as a comparison exercise—write the same query in two paradigms to internalize the differences. - Treat the security and transaction chapters as reference material; revisit them when designing access control or consistency strategies. - The book's companion site, sql-nosql.org, offers slides, tutorials, case studies, and a workbench for MySQL and Neo4j—use it for hands-on practice. 【Coverage Limits】 This guide is based on stratified excerpts covering the front matter, table of contents, and early-to-middle chapters; detailed content from later chapters and specific examples is not fully represented. Claims about chapter-level depth rely on the provided structure rather than complete chapter text.