Note: In the past months we have worked on the new Elasticsearch Node.js client, and if you want you can already try it by following the instructions below, while if you're going to use the legacy one or report an issue, please check out elastic/elasticsearch-js-legacy.
The official Node.js client for Elasticsearch.
- One-to-one mapping with REST API.
- Generalized, pluggable architecture.
- Configurable, automatic discovery of cluster nodes.
- Persistent, Keep-Alive connections.
- Load balancing (with pluggable selection strategy) across all available nodes.
- TypeScript support out of the box.
npm install @elastic/elasticsearch
The library is compatible with all Elasticsearch versions since 5.x, but you should use the same major version of the Elasticsearch instance that you are using.
# Elasticsearch 7.x
@elastic/elasticsearch@7
# Elasticsearch 6.x
@elastic/elasticsearch@6
# Elasticsearch 5.x
@elastic/elasticsearch@5
You can find the full documentation in the docs folder.
const { Client } = require('@elastic/elasticsearch')
const client = new Client({ node: 'http://localhost:9200' })
// promise API
const result = await client.search({
index: 'my-index',
body: { foo: 'bar' }
})
// callback API
client.search({
index: 'my-index',
body: { foo: 'bar' }
}, (err, result) => {
if (err) console.log(err)
})
The returned value of every API call is formed as follows:
{
body: object | boolean
statusCode: number
headers: object
warnings: [string]
meta: object
}
The client is designed to be easily configured as you see fit for your needs, following you can see all the possible options that you can use to configure it.
{
// the Elasticsearch endpoint to use
node: string | string[];
// alias of above
nodes: string | string[];
// custom connection class
Connection: typeof Connection;
// custom connection pool class
ConnectionPool: typeof ConnectionPool;
// custom transport class
Transport: typeof Transport;
// custom serializer class
Serializer: typeof Serializer;
// max number of retries for each request
maxRetries: number;
// max request timeout for each request
requestTimeout: number;
// max ping timeout for each request
pingTimeout: number;
// perform a sniff operation every `n` milliseconds
sniffInterval: number;
// perform a sniff once the client is started
sniffOnStart: boolean;
// custom sniff endpoint, defaults `_nodes/_all/http`
sniffEndpoint: string;
// perform a sniff on connection fault
sniffOnConnectionFault: boolean;
// configurethe node resurrection strategy, default `ping`
resurrectStrategy: 'ping' | 'optimistic' | 'none';
// adds `accept-encoding` header to every request
suggestCompression: boolean;
// enable gzip request body compression
compression: 'gzip';
// ssl configuraton
ssl: http.SecureContextOptions;
// http agent options
agent: http.AgentOptions;
// filters which node not to use for a request
nodeFilter: nodeFilterFn;
// custom selection strategy, defaults `round-robin`
nodeSelector: nodeSelectorFn | string;
}
If needed you can pass request specific options in a second object:
// promise API
const result = await client.search({
index: 'my-index',
body: { foo: 'bar' }
}, {
ignore: [404],
maxRetries: 3
})
The supported request specific options are:
{
ignore: [number], // default `null`
requestTimeout: number, // client default
maxRetries: number, // default `5`
asStream: boolean, // default `false`
compression: string, // default `false`
headers: object, // default `null`
querystring: object // default `null`
}
This software is licensed under the Apache 2 license.