Amazon Elasticsearch Service

What do you like best?

One of the best aspects of Elasticsearch is the flexibility in searching using dates. Elasticsearch supports many different ways of searching for date specific information, which is much easier than using SQL in a typical relational database. It also has an inbuilt "weighting" mechanism for returning results. This allows a more nuanced result list than just sorting by one of the returned data values.

What do you dislike?

The only complaint I have with Elasticsearch is the learning curve. There is really no comparison to writing SQL with how you generate Elasticsearch queries, so it requires some learning and training to figure out how to work it. There is also some complexity in determining how to set up your indexes, which can have an impact on your search results and capabilities. Honestly, it's no different than the complexities of someone new learning how to query a relational database, but it is a completely different skillset.

What problems are you solving with the product? What benefits have you realized?

We are using Elasticsearch to quickly search for users in our system based on a multitude of criteria. We are also using it to generate data for our dashboard charts. This is where Elasticsearch's ability to break down date filters has really helped us.