7.13.0 release notesedit

Upgrading to Enterprise Search 7.13.0? See Upgrading & migrating.

Enterprise Searchedit

New featuresedit

  • Adds Docker images and Linux archive for AArch64.
  • Adds the configuration setting kibana.external_host.

    • Optionally set this to the base URL of the Kibana instance you are using with Enterprise Search.

Breaking changesedit

  • Removes the warm phase in the default Enterprise Search ILM policies.

App Searchedit

New featuresedit

  • Adds a precision parameter to the search and search settings API operations, allowing precision tuning.

    • See Precision tuning in the App Search documentation.
    • This functionality is in beta.
  • Adds automatic web crawling.

    • Crawl automatically at a regular interval. Configure the interval through a UI or API.
  • Persists the web crawler URL queue and seen URLs list to allow crawls to resume after a restart or crash.

    • This change also increases the potential scale of crawls.
    • Increases the default unique URL limit to 100000.
  • Adds web crawler support for custom fields via meta tags and data attributes.

  • Adds a UI to provide sitemaps for web crawls.

    • See Sitemap in the App Search documentation.
  • Supports using * as a wildcard character within crawl rules.
  • Deletes documents disallowed by crawl rules and robots.txt directives when recrawling after rules have changed.
  • Allows configuring the web crawler user agent string.

  • Adds an API operation that returns the web crawler user agent string.

  • Allows crawling of private IP addresses and localhost, via configuration.

  • Allows configuring the fields the web crawler uses to determine duplicate content.

  • Displays document counts per-domain within the web crawler UI.
  • Adds web crawler support for the HTML base element.
  • Logs each redirect the web crawler follows as a url-fetch event.
  • Accepts Elasticsearch tokens for API access.

  • Adds default App Search role mappings for Elastic Stack roles.

Bug fixesedit

  • Ensures the web crawler honors crawl rules when following redirects.

    • The web crawler will also follow only those redirects that stay within the domains configured for each crawl.
  • Removes unnecessary attempts to authenticate App Search API requests against the Elasticsearch SAML realm when SAML auth is configured.

    • This change should reduce the number of 400 errors in Elasticsearch logs.

Known issuesedit

  • When fetching a robots.txt file, the web crawler will not follow redirects.

Workplace Searchedit

New featuresedit

Bug fixesedit

  • Speeds up full syncs by skipping unmodified documents.
  • Reduces the number of calls made to Salesforce when indexing.
  • Improves memory usage during indexing.

    • This fixes some issues where Workplace Search was running out of memory while it indexed.
    • The change also improves the use of threads on larger systems with more memory. In these cases, Workplace Search will use the host CPU more effectively and index more quickly.