December 2019 Summaries

The text explores the creation of Full Text Search (FTS) indexes in Couchbase, focusing on best practices and use cases such as simple search, field-independent search, nearest neighbor search, and dynamic search. Using the Travel Sample dataset, it details the indexing of various fields like hotel descriptions, names, aliases, addresses, reviews, and geolocation coordinates. Each use case involves creating specific indexes with tailored type mappings and field settings to optimize search performance. It highlights the importance of configuring JSON type fields, analyzers, and dynamic mappings, while demonstrating how to execute and test these indexes using both the Couchbase UI and REST API. The document emphasizes the benefits of using the Scorch index type for its improved performance and reduced size, and provides detailed instructions on how to run searches and obtain results, including highlighted matches and geospatial searches.

Couchbase has introduced the Index Advisor Service to optimize query performance by providing index recommendations for queries, especially for those using N1QL (SQL for JSON). This service helps DBAs, developers, and architects by suggesting appropriate indexes to improve query efficiency without the need to download the latest Couchbase server. The Index Advisor is part of Couchbase Server 6.5 and can offer advice for both regular and covering indexes, even for older server versions. It analyzes queries and recommends indexes based on the order of predicate types and other factors, enhancing performance by suggesting minimal and covering indexes that avoid unnecessary data retrieval. The tool, which currently outputs recommendations in JSON format, aims to expand its capabilities, including UI improvements and support for user-provided data and existing indexes.