Elasticsearch in Action (Radu Gheorghe, Matthew Lee Hinman, Roy Russo) (Z-Library)

【One-Line Pitch】 A hands-on guide to building professional-quality search into your applications with Elasticsearch, using its REST API and cURL examples that translate to any language. Best for developers who want working search features—indexing, querying, relevance tuning, and suggestions—without wrestling low-level Lucene internals. 【Book Arc】 - **Opening (~0%–10%)**: Introduces Elasticsearch as a distributed, Lucene-based search engine, its typical use cases (full-text search plus real-time analytics), and how to install and reach it via the REST API on port 9200. - **Early (~10%–32%)**: Covers the logical and physical layout—documents, types, indices, nodes, and primary/replica shards—then walks through indexing your first document, auto-creating indices and mappings, and retrieving data. - **Early–Middle (~32%–48%)**: Moves into mappings and core field types (string, numeric, date, boolean), arrays and multi-fields, predefined fields, updating documents, and optimistic concurrency control through versioning. - **Middle (~39%–48%)**: Explains the anatomy of a search request—query, size, from, _source, and sort—and the crucial query-versus-filter distinction, including the filtered query and cacheable bitsets. - **Middle (~48%+)**: Dives into query types such as match (boolean and phrase behavior with slop) and range, with guidance to prefer filters for performance. - **Late (beyond excerpts)**: The excerpts do not cover the later chapters on analysis, performance, and administration in detail; the table of contents hints at these but the sampled material stops around query construction. 【Key Takeaways】 - **Elasticsearch is document-oriented and schema-free by default** (Early): You can index a document without pre-creating the index or mapping—Elasticsearch auto-generates both, letting you start fast and refine later. - **The logical/physical split matters** (Early): Documents live in types within indices (logical), while nodes and primary/replica shards handle storage and distribution (physical); understanding this explains how indexing and searching scale. - **Queries score, filters don't** (Middle): Queries return relevance scores; filters only ask whether a document matches, making them faster and cacheable. The book's rule of thumb: if unsure, make it a filter. - **Mappings control searchability** (Early–Middle): The `index` option (`analyzed`, `not_analyzed`, `no`) determines whether and how a field is searchable—critical for matching user expectations. - **Versioning enables optimistic concurrency** (Early–Middle): Each update increments a version; conflicting parallel updates fail, and `retry_on_conflict` can reapply automatically. - **Search requests have tunable components** (Middle): `size`, `from`, `_source`, and `sort` shape results; deep pagination is expensive because Elasticsearch must compute all preceding top hits. - **Match queries are flexible** (Middle): Default boolean/OR behavior can be switched to AND, or to phrase behavior with slop for positional matching. - **REST API via cURL is the teaching vehicle** (throughout): Examples are language-agnostic, so you can port them to your stack. 【Reading Tips】 - Deep-read the early chapters on documents, types, indices, and shards—these mental models underpin everything later. - Skim the installation and configuration material if you already have a running cluster; return to it when tuning cluster name or JVM settings. - Pay close attention to the query-versus-filter distinction and the filtered query pattern; it recurs and affects performance. - Treat the cURL snippets as templates: run them, then translate to your language's HTTP client. - Don't skip the versioning/concurrency section if your application performs concurrent updates. 【Coverage Limits】 This guide is based on stratified excerpts covering roughly the first half of the book; later chapters on analysis, performance tuning, and administration are only hinted at and not summarized here.