The Elastic Search Experience (ESE) provides an intuitive search interface for your data. Many people know how to use a search engine, and often turn to this simple interface to begin their quest for information. The Elastic Search Experience satisfies this need, as a powerful tool to increase productivity by making large sets of data easily accessible.
The Elastic Search Experience is a customizable application that is configured to the needs of your users. Templates are used to customize the search results based on the data a user is searching. Often times, data has to be presented in a particular way so that a user searching it can make sense of the results. The Elastic Search Experience application is designed to be tailored towards these needs.
It's powered by Elasticsearch, which is a mature, scalable, open source, search engine. Elasticsearch provides a flexible way to search over a wide variety of data. The Elastic Search Experience application sits on top of it, and provides a friendly interface any user can understand with zero training.
This repository is for Search Teams looking to run a universal search interface on top of large sets of data. It requires familiarity with Ruby on Rails to configure, customize, and run the application. If your team isn't familiar with Ruby on Rails, I'd encourage you to try Elastic App Search. It comes with a nice React front-end called Search UI and an Admin UI.
This application depends on an Elastic cluster. You can use a local instance or one running inside Elastic Cloud.
Run a local Elastic cluster in Docker:
Run a cluster in Elastic Cloud:
Clone the repo to your local environment:
$ git clone github.com/gose/elastic-search-experienceSetup a new credential file:
$ rails credentials:editAdd the following contents (setting values where necessary):
elastic_local:
user: "elastic"
password: "xxx"
scheme: "https"
host: "localhost"
port: "9200"
elastic_cloud:
user: "elastic"
password: "xxx"
scheme: "https"
host: "xxx"
port: "443"
elastic_apm:
secret_token: "xxx"
server_url: "xxx"
user:
password: 'demo'The application comes with a built-in demo user to simulate authenticated searches. When deploying this application to production, using OAuth is recommended.
The data you have in Elasticsearch can be queried via the search_controller.rb. It has many concerns that contain the logic for the different indices you'll be querying in Elasticsearch. For example, the lunch_search.rb concern will contain most of the Elastic domain logic.
Results are rendered by the document type returned from Elasticsearch, using a partial view in app/views/search. Check out the partial views for Lunch results in _results-lunch.html.slim. It provides a good reference for creating your own custom results based on your data.
- The
search_controlleris the main entry point for a query - Based on the index being queried, a Rails Concern is called
- The concern will use a persistence layer pattern to talk to Elasticsearch
- After results are retrived and mapped to an object, they're handed off to the
showview - The bulk of the
showlogic is captured inside the partial template_results.html.slim - Based on the index being queried, different views for documents are rendered (seen screenshot below)
Mixed Result Types:
If you're not familiar with Elasticsearch terminology (e.g., "documents", "indexes", "fields", etc.), you can read more in Data in: documents and indices. Another great overview is the What is Elasticsearch? page.
To add a new data source to your Elastic Search Page, follow these general steps.
Some logs in Elastic can contain many fields. As you think about adding a new data source to your search page, sometimes it helps to write down the specific fields from a document that you want to search or render. This can make it easier to onboard the data source without being overwhelmed by the number of fields some data sources contain.
View all the Data Sources available: