Using Elasticsearch

Abstract

Elasticsearch is a real-time, full-text search and analytics engine based on RESTful API using JSON over HTTP. Some of the features of Elasticsearch are distributed, highly available, document-oriented, and schema-free. Couchbase Plugin for Elasticsearch makes it feasible to index and search data stored in Couchbase Server in real-time using Elasticsearch. With the plugin, data streams from Couchbase Server to Elasticsearch in real-time. Couchbase data gets indexed in Elasticsearch and may be queried using a RESTful API. To make use of the plugin, two clusters are required to be created: an Elasticsearch cluster and a Couchbase cluster. The plugin is installed in the Elasticsearch cluster. Using the Cross Datacenter Replication (XDCR) in Couchbase Server, the Couchbase data is replicated and streamed to Elasticsearch cluster. The Elasticsearch cluster may be queried to get results as document IDs. The document IDs may be used to retrieve the document from Couchbase Server directly. The Couchbase Server data is kept in sync with the Elasticsearch cluster index. Any changes in the Couchbase data are streamed in real-time to Elasticsearch.