Introduction to Atlas Search

In today's digital age, full-text search capabilities are a necessity for modern applications. The efficiency of search mechanisms plays a pivotal role in enhancing user experience and ensuring easy availability of relevant information. Building relevance-based application features is a complex endeavor. You don't just need to sift through vast amounts of data, but also ensure that the results are tailored to the user's intent and context. The significance of these features is profound because they streamline user interactions, fostering engagement and retention.

Atlas Search is an embedded full-text search feature in MongoDB Atlas, providing a seamless and scalable experience for building relevance-based applications, among other advanced search features. Built on Apache Lucene, Atlas Search eliminates the need to operate a separate search system alongside your database.

This chapter will cover the following topics:

• The features...

MongoDB Atlas Search

Before MongoDB Atlas Search, developers had to rely on various other solutions to implement search functionality in their applications. Some of these solutions included:

• Algolia: This is a powerful and flexible search and discovery solution. Even though it's used in many applications, it can be quite costly. This cost is driven by the quantity of records stored and the volume of API actions performed. If you have a large dataset, Algolia might not be the best fit for you.
• Elasticsearch: This is a distributed, RESTful search and analytics engine suitable for a growing number of use cases. Just like Atlas Search, it is based on Apache Lucene.
• Solr: This is another powerful search platform built on Apache Lucene. It's highly reliable, scalable, and fault-tolerant, providing distributed indexing, replication, and load-balanced querying.
• The built-in $text index of MongoDB: This is the easiest solution to implement but was not suitable...

Technical overview of Atlas Search indexes

Atlas Search indexes are data structures that categorize data in an easily searchable format. They establish mapping between the terms themselves and the documents that contain those terms, enabling you to retrieve documents faster by using certain identifiers.

An Atlas Search index can be established on a single field or multiple fields. It is advisable to index fields that are frequently used to sort or filter your data, as this allows for the rapid retrieval of documents containing pertinent data during a query.

Index creation and management

When you set up one or more Atlas Search indexes, the mongot process is activated on the nodes in the cluster. Each mongot process communicates with the mongod process on the same node. The mongot process interacts with the backend database and opens change streams for each index to create and update search indexes.

When you create an Atlas Search index, you can specify the fields using...

Apache Lucene

Atlas Search is a full-text search engine built on top of Apache Lucene, showcasing the versatility of Lucene in a range of search-based applications. Apache Lucene itself is an open-source project focused on the development of search software, and offers a core search library known as Lucene Core.

Apache Lucene Core is a high-performance, full-featured search engine library written entirely in Java. It is a technology that's suitable for nearly any application that requires structured search, full-text search, faceting, nearest-neighbor search across high-dimensionality vectors, spell correction, or query suggestions. The key features of Lucene include the following:

• Scalable, high-performance indexing:
  • Small RAM requirements
  • Incremental indexing as fast as batch indexing
  • Index size roughly 20-30% the size of the indexed text
• Powerful, accurate, and efficient search algorithms:
  • Ranked searching (best results returned first)
  • Numerous powerful query types...

Summary

MongoDB Atlas Search is a powerful, integrated full-text search solution that simplifies the development of relevance-based application features, and ensures real-time data synchronization. Its seamless integration with MongoDB Atlas and Apache Lucene provides a streamlined and efficient solution for implementing full-text search functionality.

Atlas Search features, such as rich query DSL, fuzzy search, synonyms, custom scoring, a visual editor, autocomplete, fast faceting, and highlighting, allow you to build sophisticated search logic, return relevant search results despite typos or spelling errors, define similar search terms, control how search results are scored and ranked, and more.

Atlas Search integrates with Apache Lucene by embedding a Lucene node within each Atlas cluster, which is responsible for handling search indexing and querying. The integration is achieved through a dedicated mongot Java web process that runs alongside mongod on each node in the Atlas...