Category: Elasticsearch

Venkata Gummadi works on vector search response times:

As data engineers, we are tasked with implementing these sophisticated solutions, ensuring organizations can derive actionable insights from vast datasets. This article explores the intricacies of vector search using Elasticsearch, focusing on effective techniques and best practices to optimize performance. By examining case studies on image retrieval for personalized marketing and text analysis for customer sentiment clustering, we demonstrate how optimizing vector search can lead to improved customer interactions and significant business growth.

Read on for a vector search primer and some guidance of how you can improve the performance of vector search queries. I’d expect that much of this can also apply to Azure AI Search and Amazon OpenSearch.

Leave a Comment

The Big Data in Real World team is done with this index:

Simple problem with a simple solution. In this post we will see how to delete an index in Elasticsearch.

Read on for the command to delete an index, and a demonstration of it in action.

Comments closed

The Big Data in Real World team needs an alias:

An alias as the name suggests is an alias or another name to the index in Elasticsearch. It is quite useful when you want to refer to an index by another name. So instead of performing an reindex to rename or cloning an index you can create an alias to the index.

Click through for the script to create an alias, how you might use one, and the right way to delete one without removing the underlying article.

Comments closed

The Hadoop in Real World team want to delete multiple documents in Elasticsearch matching some condition:

Deleting a single document is pretty straightforward in Elasticsearch. We can simply issue a DELETE on the document id and the document will be deleted from the index.

Click through for that example, as well as the multi-document delete (and a check to ensure that it worked).

Comments closed

The Hadoop in Real World team show how to migrate specific documents when building a new index:

As shown in the other post, we still use a reindex by specifying the source and destination but this time we also specify a query in source along with the term which indicates that the documents with state = ‘ny’ will be filtered from the source.

So only documents with state ny will be copied to the new index account_v3 with this reindex operation.

Click through for an example of how this works.

Comments closed

The Hadoop in Real World team moves data around in Elasticsearch:

In this post we will describe how to copy an index and its contents to a new index in Elasticsearch.

We currently have an index named account. We are going to copy the account index and its content to another index named account_v2 using the reindex API.

Click through to see how.

Comments closed