I work with Elasticsearch as a developer on a daily bases for textual search. It is the greatest tool I've used in years. Really like it.

It is just the best back end for search engines in the market. A NoSQL database that is trustworthy. Also, it is open source. Incredibly easy to use.

Mostly for developers: other people would have a hard time with it.

I am extremely satisfied with Elasticsearch, its speed & querying capabilities among lots of data amaze me!

The fact that Elasticsearch offers really efficient & quick querying of data without a compromise on the different range of queries it can support is really awesome, also Elasticsearch can rank matching documents based on matching criteria which is also very useful. Elasticsearch also handles distributed queries very efficiently.

Reindexing documents can be a real pain if there is a major change in the schema, otherwise no complaints.

Our overall experience with Swiftype was really negative. A big waste of our time and money. After waiting for them to follow up on our last conference call, we switched to Google Custom Search Engine. It turned out to be far less expensive than we thought. Swiftype is a bad investment if you have lots of files to be indexed but far fewer monthly searches. Google Custom Search now works well for us -- at a tiny fraction of Swiftype's annual cost.

1. It was fairly easy to set up. 2. It seemed to provide more accurate search results over time. 3. Customer service concerning technical questions was good.

1. Reports provided are too limited in time span covered to be helpful. 2. Company said number of documents we wanted indexed exceeded our quota and would result in higher cost at next renewal. This conflicted with our own records, so we repeatedly asked for a list of the pages being indexed. We got lots of emails and telephone conferences and promises, but we never got the list. 3. Swiftype seemed to have something to hide. We would talk with them about above indexing issue, and they said they would get back to us soon. But they didn't. They seemed to forget about us. I guess they were hoping we would just go away. 4. The last month or more, most of our searches were spam bot searches for non-existent domain names. This made our reports really worthless. (We have not had this problem anywhere else on our sites.)

Elasticsearch helps in performing optimised searching over the vast amount of data present in your system. The configuration for Elasticsearch can have some learning curve, but once setup completely, the indexed searches are really fast as compared to other basic solutions and helps users in searching your system from vast amount of data. Its ideal for Big data and can be used in other e-commerce solutions also. Text searching is really fast and helps your system in maintaining performance and scalability over search operations.

Elasticsearch is a great solution for implementing optimised searches over the data you have lying in your databases. The setup and configuration is a bit tricky, but once you have configured the data points and the indexes, the search is lighting fast. You can go for the complete text search or for the autocomplete bases suggestions which can help your users to quickly search for the right things in your system.

The setup of Elasticsearch does have a learning curve for the developers, but the documentation is really helpful in getting you over the curve. Setting up indexes or data points for search can be tricky. Elasticsearch needs to setup a more proactive customer support so that developers who are facing deadlines, can easily sort their issues with the help of tech support.

It's been great!

As a backend developer, I know for a fact that I can rely upon elasticsearch for all the times when I have to support search based on string across the database. You throw some set of documents to elasticseatch, and you are good to go to search anything amongst the provided documents. They have some supercool features to support string based search for say support for distributed systems, support for aggregations and analysers & folding, etc. The documentation out there makes it lot easier to get going with it :)

The memory optimisation related issues could be alarming sometimes if not taken care of properly. You may end up having some weird system crashes.

We get to solve our search scalability problems on our on-premise product with the ease and consistency of Elasticsearch.

* Easy to use, especially when compared with alternatives s.a. Solr. * Free software. * Elastic has a great customer support. * It is very powerful, supporting heavy loads. * Well-documented.

* Demands a specific knowledge of the theory of search. * While way higher level than most services, it is not as high level as, let us say, a Google search. * Often used as a NoSQL in inappropriate settings.

Easy to get started, fast with basically no optimization. It can handle massive amounts of data without breaking a sweat.

Getting data INTO Elasticsearch can be tricky, depending on the source. Version 7 fixes a lot of these issues, however. Licensing is less than ideal, too: many useful features are locked behind a paid support license, and the open source version is receiving fewer and fewer new features.

If you worked in a big online store on the analytics side you can really understand the hard work that this software is doing, and the value that it adds to your conversion rate. It has great potential and used in combination with Kibana it proves a very good choice for fast and great search results.

Depending on your users conversion rate and volume it can become a high cost operation, if you have low margins.

Elastic-search is a distributed, Restful search and analytics engine.

Elastic-search has an important security layer to separate access to data and dashboards.

A bit more of a learning curve for complex searches, indexing more complex things.

Developing a powerful and a fast search engine, dealing with content coming from various data sources, inside our website.

ElasticSearch, as of Solr, is the right tool if you want to create a powerful search engine inside an application or a website. Indexing documents can be done through a REST API, as for querying data. Most of the modern frameworks and content management systems can ship with modules that interact with ElasticSearch to index and retrieve documents. For most complex systems, you can implement synonyms, auto-completion, auto-correction (using fuzzy search), and also configure how results should be sorted using a full text search. Searching with complicated parameters is really fast using ElasticSearch, and you can also retrieve facets to display to restrict results when needed.

Learning ElasticSearch seems to be a bit more difficult than learning it's competitor, Solr. In my experience, Solr has a better support for Drupal CMS which we use a lot in my company. But our last project using ElasticSearch was a success, thanks to all it's powerful features.

Elastic search has completely changed how we can fetch and visualize our customer's data. With the advanced text indexing, we can easily build a text search engine which gives accurate results. The powerful query format allows us to construct complex queries from deep filter trees and aggregate data in any way we want. Most importantly, all of this is extemely fast - giving results within milliseconds when working with complex queries on milions of documents. If performance ever becomes a problem, Elastic search is easily horizontally scalable and therefore future-proof.

One critical thing I'm missing is a proper way of handling migrations. If you for example need to change the format of indexed documents - and therefore the mapping, you need to develop your own way of programatically handling the reindexing - or do it manually.

Setting up a search platform using Elasticsearch was super fun and easy. I was able to get some great search query response times and development efforts were minimal compared to dealing with Lucene directly.

It's a great tool to setup a search platform almost of any size in a company. I have worked with Apache Lucene and Elasticsearch just saves all that trouble of dealing with the indices all the time in Lucene and rather provides seamless integration through APIs for search queries. It was really easy to setup the cluster and shards and manage the stack. The APIs offer a wide range of search features and saves a lot of effort coding hundreds of line for every small advanced search filtering function.

The documentation of the APIs could be improved. It was a bit overwhelming at times to lookup the correct functionality we wanted to actually implement. The tutorials and courses are quite expensive.

Easier to horizontally scale Distributed by design Excellent full-text search Elastic Search is schema free'instead, it accepts JSON documents, as well as tries to detect the data structure, index the data, and make it searchable. Elastic Search is document-oriented. It stores real world complex entities as structured JSON documents and indexes all fields by default, with a higher performance result. Elastic Search is API driven; actions can be performed using a simple Restful API. Elastic Search records any changes made in transactions logs on multiple nodes in the cluster to minimize the chance of data loss.

Elastic-search query DSL is less common and less flexible than Postgres SQL . Everything is indexed by default, which creates an index overhead. in order to understand how to properly query your data and how it is stored, you have less control over consistency.

Find some other product

easy to set up and configure .

Unfortunately, and this is VERY SERIOUS, we have noticed that the search results returned by Switftype are incorrect. Some of the results that it SHOULD BE including for a given user keyword request are NOT included and some of the results that it IS including DO NOT match the keyword request. We have worked with their technical support to no avail. Unless you test your search results carefully, you wont notice this very serious flaw in the product. Given that creating the correct set of search results is the basic function of this kind of product, this flaw means that Swiftype is unusable. We are now in the process of converting to another product.

By using RoR as a platform we use wrappers developed by Elasticsearch. This is a great, well-documented solution for a quick start, a great help to the developers.

As a web developer, I have been looking for opportunities to make our apps better, smarter and more developed. Elasticsearch made a revolution in our business by providing complex filtration systems and search, which weren't easy to implement previously: too much code and server consumption. Our products got better, we make good money and save our time.

I still haven't figure the implementation of replication and clusterization of Elasticsearch, but maybe it is a lack of information.

How it is designed to make search fast with the replicated indexes. It is also well designed to handle complex structured data and queries.

There is not as much admin interface to help manage the indexes. At times the cluster turns yellow due to shards being re-allocated.

It's simply the industry standard for search. The extensive functionalities make complex problems look easy.

Documentation could be a bit more helpful, especially in dealing with breaking changes.

A bit hard at first, specially tweaking the queries & indexes to get just what we needed from it. But after that, it has been a very reliable tool that we use everyday.

ElasticSearch provides some of the mos accurate & fast search results of the market. We moved from a Sphinx installation into ES and have never looked back. Multiple indexes, lots of querying options & great support from tooling.

Can be a bit hard to get into at first. However, the product is very well documented, so you just need to invest some time in their docs.

Its decent search to NOSQL DB and I find it good for content management

it has scalable and reliability issues. I would not recommend this for your SOR

We used Google Custom Search and were really disappointed when Google dropped paid support. We changed to Swiftype and are very happy. It has given us the similar functionality with improved UI. The UI for optimizing the search is really easy to use yet powerful.

Unlike Google, Swiftype does not have indexes ready to go. As a result, when you add a new site to the search set, it takes some time to index the site. Its understandable but using Google was really simple.

The ease of search makes it possible for clients to get the answers they need quickly.

I didn't find any downsides to the software.

The flexibility provided for the configuration of the indexes and the searching criteria makes it suitable any kind of data and search needs. The query building flow is also easy to understand.

A server crash could result in a total loss of data, so it can’t be used to store important data.

Is great, easy implementation and configuration, just the requirements are a bit high.

It was easy to setup installing it with AWS or as a server. also has a lot of implementations options with Rails and other platforms.

It requires a lot of hardware >1GB of memory since is based on Java.

It is very easy to integrate with javascript projects because it stores data in json format. Querying is easy with lot of selectors and modifiers available.

Reindexing is very time taking due to which there is data mismatch sometimes.

Easy to setup, user-friendly and powerful app, also you can manage multiple websites with. Good support, integration and a flexible management dashboard to tweak all in one place.

It lacks some styling options. The prices between the 3 price plans are a bit exaggerated, so you need to choose wisely and analyze the needs for the mid-long term.