Posted On: Nov 16, 2022
Amazon OpenSearch Service now supports OpenSearch and OpenSearch Dashboards version 2.3. With this version, Amazon OpenSearch Service adds several features such as new algorithms to the machine learning (ML) commons library, improvements to aggregations, improvements to map visualizations, alerting, anomaly detection, and more.
OpenSearch 2.3 uses Lucene 9.1, and the move to the latest version of Lucene offers a number of exciting advancements, and will continue to offer more value in future releases. Some of the improvements in OpenSearch Service that come with OpenSearch 2.3 (includes features that were launched as part OpenSearch versions 2.0, 2.1, and 2.2) include new ML algorithms that have been added to ML Commons such as Linear regression, Logistic Regression, Localization, and RCFSummarize, support for Lucene-based k-NN search, and improved Dashboards user experience for anomaly detection. In addition, this release adds supports for document-level alerting, support for multi-terms aggregation as part of OpenSearch core, adds the ability to upload custom GeoJSONs for region map visualizations and better zoom levels (14x), support for searching by relevance using SQL and PPL, centralized management of notifications on Dashboards, amongst several other improvements. For a complete list of new features and improvements, please see the release notes for OpenSearch versions 2.0, 2.1, 2.2, and 2.3.
This release also introduces inclusive terminology (such as cluster manager instead of master, allowlist instead of blacklist) throughout OpenSearch. For information on upgrading to OpenSearch 2.3, please see documentation. For a list of breaking changes with the new major OpenSearch version, please see documentation.
OpenSearch 2.3 is now available for Amazon OpenSearch Service domains in the Amazon Web Services China (Beijing) region, operated by Sinnet, and the Amazon Web Services China (Ningxia) region, operated by NWCD.
To learn more about Amazon OpenSearch Service, please visit the product page.