ElastomerClient

Making a stupid simple Elasticsearch client so your project can be smarter!

Client

The client provides a one-to-one mapping to the Elasticsearch API endpoints. The API is decomposed into logical sections and accessed according to what you are trying to accomplish. Each logical section is represented as a client class and a top-level accessor is provided for each.

Cluster

API endpoints dealing with cluster level information and settings are found in the Cluster class.

require 'elastomer_client/client'
client = ElastomerClient::Client.new

# the current health summary
client.cluster.health

# detailed cluster state information
client.cluster.state

# the list of all index templates
client.cluster.templates

Index

The methods in the Index class deal with the management of indexes in the cluster. This includes setting up type mappings and adjusting settings. The actual indexing and search of documents are handled by the Docs class (discussed next).

require 'elastomer_client/client'
client = ElastomerClient::Client.new

index = client.index('books')
index.create(
  :settings => { 'index.number_of_shards' => 3 },
  :mappings => {
    :_source => { :enabled => true },
    :properties => {
      :author => { :type => 'keyword' },
      :title  => { :type => 'text' }
    }
  }
)

index.exists?

index.delete

Docs

The Docs class handles the indexing and searching of documents. Each instance is scoped to an index and optionally a document type.

require 'elastomer_client/client'
client = ElastomerClient::Client.new

docs = client.docs('books')

docs.index({
  :_id    => 1,
  :author => 'Mark Twain',
  :title  => 'The Adventures of Huckleberry Finn'
})

docs.search({:query => {:match_all => {}}})

Performance

By default ElastomerClient uses Net::HTTP (via Faraday) to communicate with Elasticsearch. You may find that Excon performs better for your use. To enable Excon, add it to your bundle and then change your ElastomerClient initialization thusly:

ElastomerClient::Client.new(url: YOUR_ES_URL, adapter: :excon)

Retries

You can add retry logic to your Elastomer client connection using Faraday's Retry middleware. The ElastomerClient::Client.new method can accept a block, which you can use to customize the Faraday connection. Here's an example:

retry_options = {
  max: 2,
  interval: 0.05,
  methods: [:get]
}

ElastomerClient::Client.new do |connection|
  connection.request :retry, retry_options
end

Compatibility

This client is tested against:

• Ruby version 3.2
• Elasticsearch versions 5.6 and 8.13

Development

Get started by cloning and running a few scripts:

• ElastomerClient
  • Client
    • Cluster
    • Index
    • Docs
    • Performance
  • Compatibility
  • Development
    • Bootstrap the project
    • Start an Elasticsearch server in Docker
    • Run tests against a version of Elasticsearch
  • Releasing

Bootstrap the project

script/bootstrap

Start an Elasticsearch server in Docker

To run ES 5 and ES 8:

docker compose --project-directory docker --profile all up

To run only ES 8:

docker compose --project-directory docker --profile es8 up

To run only ES 5:

docker compose --project-directory docker --profile es5 up

Run tests against a version of Elasticsearch

ES 8

ES_PORT=9208 rake test

ES 5

ES_PORT=9205 rake test

Releasing

1. Create a new branch from main
2. Bump the version number in lib/elastomer/version.rb
3. Update CHANGELOG.md with info about the new version
4. Commit your changes and tag the commit with a version number starting with the prefix "v" e.g. v4.0.2
5. Execute rake build. This will place a new gem file in the pkg/ folder.
6. Run gem install pkg/elastomer-client-{VERSION}.gem to install the new gem locally
7. Start an irb session, require "elastomer/client" and make sure things work as you expect
8. Once everything is working as you expect, push both your commit and your tag, and open a pull request
9. Request review from a maintainer and wait for the pull request to be approved. Once it is approved, you can merge it to main yourself. After that, pull down a fresh copy of main and then...
10. [Optional] If you intend to release a new version to Rubygems, run rake release
11. [Optional] If necessary, manually push the new version to rubygems.org
12. 🕺 💃 🎉