This repository contains Ruby integrations for Elasticsearch.
The elasticsearch
library is a wrapper for two separate libraries:
elasticsearch-transport, which provides a low-level Ruby client for connecting to an Elasticsearch clusterelasticsearch-api, which provides a Ruby API for the Elasticsearch RESTful API
require 'elasticsearch'
client = Elasticsearch::Client.new log: true
# if you specify Elasticsearch host
# client = Elasticsearch::Client.new url: 'http://localhost:9200', log: true
client.transport.reload_connections!
client.cluster.health
client.search q: 'test'
# etc.Both of these libraries are extensively documented.
Please read the elasticsearch-transport and the elasticsearch-api documentation carefully.
See also doc/examples for some practical examples.
For optimal performance, you should use a HTTP library which supports persistent ("keep-alive") connections, e.g. Patron or Typhoeus. These libraries are not dependencies of the Elasticsearch gems. Ensure you define a dependency for a HTTP library in your own application.
This repository contains these additional Ruby libraries:
elasticsearch-extensions, which provides a set of extensions to the base library,elasticsearch-dsl, which provides a Ruby API for the Elasticsearch Query DSL,elasticsearch-xpack, which provides a Ruby API for X-Pack APIs. This API is going to be merged intoelasticsearch-apion v8.0.
Please see their respective READMEs for information and documentation.
For integration with Ruby models and Rails applications, see the elasticsearch-rails project.
The Elasticsearch client is compatible with currently maintained Ruby versions. See Ruby Maintenance Branches. We don't provide support to versions which have reached their end of life.
The gem's version numbers follow Elasticsearch's major versions. The master branch is compatible with the Elasticsearch master branch, which is the next major version.
| Gem Version | Elasticsearch | |
|---|---|---|
| 0.90 | → | 0.90 |
| 1.x | → | 1.x |
| 2.x | → | 2.x |
| 5.x | → | 5.x |
| 6.x | → | 6.x |
| 7.x | → | 7.x |
| master | → | master |
Use a release that matches the major version of Elasticsearch in your stack. Each client version is backwards compatible with all minor versions of the same major version. The client's API is compatible with Elasticsearch's API versions from 0.90 till current.
Check out Elastic product end of life dates to learn which releases are still actively supported and tested.
Install the elasticsearch package from Rubygems:
gem install elasticsearch
To use an unreleased version, either add it to your Gemfile for Bundler:
gem 'elasticsearch', git: 'git://github.com/elasticsearch/elasticsearch-ruby.git'
or install it from a source code checkout:
git clone https://github.com/elasticsearch/elasticsearch-ruby.git
cd elasticsearch-ruby/elasticsearch
bundle install
rake install
To work on the code, clone and bootstrap the project first:
git clone https://github.com/elasticsearch/elasticsearch-ruby.git
cd elasticsearch-ruby/
bundle exec rake setup
bundle exec rake bundle
This will clone the Elasticsearch repository into the project, and run bundle install in all subprojects. There are a few tasks to work with Elasticsearch. Use rake -T and look for the tasks in the elasticsearch namespace. You can build elasticsearch with rake elasticsearch:build after having ran setup.
To run tests, you need to start a testing cluster on port 9250, or provide a different one in the TEST_CLUSTER_PORT environment variable.
There's a Rake task to start the testing cluster. By default this is going to try and use the elasticsearch command on your system:
rake test:cluster:start
You can also configure where the elasticsearch startup script is found with the TEST_CLUSTER_COMMAND environment variable, e.g.:
TEST_CLUSTER_COMMAND=~/elasticsearch/bin/elasticsearch rake test:cluster:stop
You also can configure the port, number of nodes, and other settings with environment variables:
TEST_CLUSTER_COMMAND=./tmp/builds/elasticsearch-7.10.0-SNAPSHOT/bin/elasticsearch \
TEST_CLUSTER_PORT=9250 \
TEST_CLUSTER_NODES=2 \
TEST_CLUSTER_NAME=my_cluster \
ES_JAVA_OPTS='-Xms500m -Xmx500m' \
TEST_CLUSTER_TIMEOUT=120 \
rake test:cluster:start
You can stop the cluster with a rake task, passing in the TEST_CLUSTER_COMMAND variable:
TEST_CLUSTER_COMMAND=./tmp/builds/elasticsearch-7.10.0-SNAPSHOT/bin/elasticsearch \
rake test:cluster:stop
There's also a rake task for starting up Elasticsearch in a Docker container:
rake docker:start[version] - E.g.: rake docker:start[7.x-SNAPSHOT]. To start the container with X-Pack, pass it in as a parameter: rake docker:start[7.x-SNAPSHOT,xpack].
To run tests against unreleased Elasticsearch versions, you can use the rake elasticsearch:build Rake task to build Elasticsearch from the cloned source (use rake elasticsearch:update to update the repository):
Note: If you have gems from the elasticsearch family installed system-wide, and want to use development ones, prepend the command with bundle exec.
rake elasticsearch:build
This is going to create the build in ./tmp/builds/.
You can pass a branch name (tag, commit, ...) as the Rake task variable:
rake elasticsearch:build[origin/1.x]
To run all the tests in all the subprojects, use the Rake task:
time rake test:client
By default, tests will atempt to use http://localhost:9200 as a test server. If you're using a different host/port, set the TEST_ES_SERVER environment variable, e.g.:
$ TEST_ES_SERVER='http://localhost:9250' be rake test:client
This software is licensed under the Apache 2 license.