For example, I have an Elasticsearch cluster with a single user and two different applications that want to retrieve all documents from the index to process them. Both applications are using search with scroll mechanism and same user credentials.
Is there a chance that documents in the index will be divided between two applications, because generated scroll IDs contains same info about user which perform requests?
I've read this documentation page and as I understand, the behavior I mentioned above is a normal situation?
No, that is not possible.
The scroll id describes the state of a single request, if two requests are sent they will have separate scroll ids, even if they are for the same user.
That is a section titled Resource sharing check for users and API keys it means that if your two applications share a userid, nothing within security will stop them from accessing the scrolls of the other application. Elasticsearch will treat the 2 applications as the same user, and allow them to share resources.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.