What’s New for Search in DSE 6.7
date: December 18, 2018
DataStax Enterprise (DSE) 6.7 is built on the best distribution of Apache Cassandra™ and DSE 6.7 Search is built with a production-certified version of Apache Solr™ 6.
We are continuing with the themes of improving performance and stability as well as eliminating complexity for users from the previous release.
Here’s a quick tour of some of the enhancements and features found in DSE 6.7 Search.
New Shard Selection Algorithm
A profound feature that will benefit all DSE Search users is the introduction of a new search shard selection algorithm. One of the huge benefits of DSE search, in general, is the ability to leverage search indexes to execute vastly more flexible queries than native CQL, such as full table scans and query tables by non-primary key values. The algorithm that determines the search shards or token ranges to query from the replicas is what enables this capability.
This new algorithm has been introduced to greatly reduce the amount of shard or token filtering the coordinating node must do. This not only greatly reduces the amount of resources required to perform operations, but this reduction in work translates to performance gains since this is the main engine behind distributed search queries. This new algorithm also doesn’t impose the suggested 8 tokens when configuring DSE Search for virtual nodes.
Almost all modern or cloud applications have geospatial requirements, and it has been reported that roughly 80% of enterprises manage location data. DSE 6.7 elevates the native geospatial functionality while making it easier to configure, execute, and manage geospatial queries and data types.
Java Topology Suite
With the recent license changes from the Java Topology Suite, the Java library standard for creating and manipulating vector geometry, DataStax is now able to directly package this library directly with DSE, eliminating the need to manually extend DSE’s capabilities.
With geospatial functionality becoming more standard, working with the right datatypes is as important as the platform’s query capabilities. DSE can automatically create a search index by inferring the data types in the CQL schema and generating a corresponding Solr schema. DSE 6.7 also handles the now native spatial data types LineStringType, PointType, and PolygonTypes automatically, reducing the steps a user must go through to enable spatial analysis on search applications.
Extra Support for Facet Heatmap
Finally for our geospatial users, DSE 6.7 has added support for Solr’s facet heatmap functionality through our CQL solr_query API allowing users to leverage this powerful feature from Solr to generate heatmaps from location data through the native DSE protocol.
Download DSE 6.7 today.