Elasticsearch (and ELK)

Elasticsearch is similar to Solr, as it exposes the capacities of Lucene on the web, in open source.
Elasticsearch is tailored for Big Data. Created by Shay Bannon, Elasticsearch is managed the Elastic company. Datafari uses the ELK stack for its analytics functionnalities.

From its inception, Elasticsearch wanted an easy-to-scale system, with a REST approach. Elasticsearch is also releasing special functionnalities dedicated to logs analysis, through Kibana and logstash.

IIt is a stable backbone, able to handle scalabality throughmachine clustering, easy to manage, and supporting REST calls.


You can get more technical information on the website of Elasticsearch. Contrarily to Solr, Elasticsearch is not a project from the Apache foundation, but uses the same licence. It is available under the Apache v2 licence. France Labs proposes its expertise to install, configure, extend and maintain Elasticsearch on your systems.

Elasticsearch is the open source search engine that became popular thanks to the big data needs. It proposes advanced functionnalities, it is highly configurable, and it can compete with proprietary solutions. A search engine, it is a technical building block, able to digest big data, and to make it available to users in a clever way, in a few milliseconds.

Historically, Elasticsearch is an evolution from the Apache Lucene project. The latter is the heart of the search engine, but is just a subset of the functionnalities achievable through Elasticsearch. Furthermore, Lucene is a java API, it needs to be integrated at the source code level, whereas Elasticsearch is standalone.

ElasticSearch is becoming the de facto search tool for web startups, and benefits from a wide users community.

Use cases for ElasticSearch

It is not obvious to imagine all the possible uses of Elasticsearch, that is why we expose here some of them.

  • Big Data : Elasticsearch is completary to Hadoop, can synchronise to it in order to mine the data. Its integration with Hadoop is not as finegraine as Solr, but its capabilities and its ease of use contributed to its wide usage. Hadoop provides the basic building blocks for big data: storage and analysis. But the "front end", the one that allows to dig into the data, is not part of Hadoop mission.
    Elasticsearch then comes into play to allow its users to search into the data stored in Hadoop. It is one thing to store petabytes of data, it is another to make it relevant for your business.
    Elasticsearch also comes into play as a front-end to use the results of the batch processes done in Hadoop.
    Use case: Netflix uses Elasticsearch to analyse data stored in its Hadoop, to analyse logs and errors.
  • Social network sites: Elasticsearch is well appreciated by web startup, thanks to its simplicity for scaling up on clusters of machine, reliably. For instance, Xing uses it for its near realtime search among its 14 million members. ElasticSearch accompanies web startup in their data growth, being able to handle up to billions of data in less than a second.