Introduction

This article is part of a series of pills for Product Managers who want to get a feel for technology, to be able to test it, install it and demonstrate its capabilities. At taniwa we believe it’s bold to present yourself as a PM or Project Leader in a knowledge domain where you’re ignorant.

Knowledge is power

Think of ElasticSearch (ES) whenever your product needs:

1. Indexed searches, that is, searching for text in different parts of a document intelligently.
2. Groupings by document types. For example, an Ikea catalog and its facets (“Sport”, “Color”, “Size”).
3. Speed in both searches and search bar suggestions.
4. Centralizing lots of information so you can search it in real time. Application logs no matter how large.

ES is an old product, based on Apache Lucene, but it has managed to stay in the fight by doing a few things very well that many companies want in their products:

• Text searches, in any language and applying very flexible text-mining algorithms.
• Robust. Setting up an ES cluster that really works is easy (unlike others).
• Fast. Lightning-fast parallel insertions and blazing searches across millions of documents.
• REST interface and JSON results. From day one, which put it ahead of its time (not anymore).

In addition to its search and indexing engine, it has other value-laden elements such as:

• Logstash: Document insertions from log files, for example.
• Kibana: Web console to exploit and visualize data.

They’re usually installed together and that’s why we talk about an ELK solution.

We won’t include in this article how to install ES but you have many options that in order of preference for us are:

1. ES Cloud, which allows you to deploy it on AWS, Azure or Google Cloud machines among others.
2. Download a Docker locally and start playing.
3. Create your own cluster. Not so preferred unless necessary.

Amazon has its own similar service called OpenSearch with its theft story that you can read here Amazon: NOT OK - why we had to change Elastic’s licensing

Documents and Mappings

A document is a structure of fields each with its data type. For example, a blog post would have: