8.3.0 release notes

Upgrading to Enterprise Search 8.3.0? See Upgrading and migrating.

Operators can now filter on sites in SharePoint Online by using the indexing rule path_template. This change will break existing path_template rules. Operators must prefix all existing path_template rules with /sites/{site-name}. See indexing rules.

Authenticate using JSON Web Tokens (JWT). Learn more in the App Search docs and Workplace Search docs. (beta)
Use the web crawler to extract content from downloadable binary files, such as PDF and DOCX files. Enterprise Search uses the ent_search_crawler Elasticsearch ingest pipeline to power the web crawler’s binary content extraction. Learn how to install the Ingest Attachment Processor plugin for your Elasticsearch cluster and how to enable binary content extraction. (technical preview)
Use the new Confluence Cloud connector package to deploy and run a Confluence Cloud connector on your own infrastructure. (technical preview)
Use the new Custom connector package to deploy and run a custom connector on your own infrastructure. (technical preview)
Workplace Search connectors now route all indexed documents through the Elasticsearch ingest pipeline named ent_search_connectors. Edit this ingest pipeline to customize how data is transformed before indexing.

The web crawler now uses UTF-8 encoding by default for responses without a specified charset, instead of ISO-8859-1. Adjust the new crawler.http.default_encoding configuration to override this behavior.
Addressed a potential SSRF vulnerability for external connector packages by disallowing redirects away from the configured host. Disallowed localhost external URLs for connector packages, unless explicitly configured.
GitHub sync failed when there was any empty repository. Solved so sync continues normally.

App Search users with dev and editor roles do not have access to the web crawler.