Elasticsearch Delete-By-Query

Elasticsearch Delete-By-Query

Last Update: July 2020

Have you tried our Elasticsearch Check-Up ? (no installation required) get personalized recommendations that can prevent errors and improve your search and indexing speed

Delete-By-Query in Elasticsearch


Delete-by-query is an Elasticsearch API, which was introduced in version 5.0 and provides functionality to delete all the documents based on the matching query. In lower versions, users had to install the Delete-By-Query plugin and use the DELETE /_query endpoint for this same use case.

What it is used for

This API is used for deleting all the documents from indices based on the matching query. Once the query is executed, Elasticsearch runs the process in the background to delete all the matching documents and you don’t have to wait for the process to be completed.

Examples

Delete all the documents of an index without deleting the mapping and settings:

POST /my_index/_delete_by_query?conflicts=proceed&pretty
{
  "query": {
    "match_all": {}
  }
}

The conflict parameter in the request is used to proceed with the request even in the case of version conflicts for some documents. The default conflict behavior is to abort the request altogether.

Notes
  • A long-running delete_by_query can be terminated using _task API.
  • Inside the query body, you can use the same syntax for queries that are available under the _search API.

Common Problems

Elasticsearch takes a snapshot of the index when you hit delete by query request and uses the _version of the documents to process the request. If a document gets updated in the meantime, it will result in  a version conflict error and the delete operation will fail.


About Opster

Opster takes a different approach to Elasticsearch operation - Opster pro-actively troubleshoots, optimizes, automates and assists in what's needed to run Elasticsearch smoothly in production.


« Back to Index