Jonathan Lacefield

What’s New in DataStax Enterprise Graph 5.1

By Jonathan LacefieldApril 28, 2017

We are pleased to introduce you to the great new features that we are rolling out with DataStax Enterprise Graph (DSE Graph) 5.1.  

DSE Graph is tightly integrated with the full DataStax Enterprise platform providing a unique experience where graph, search, operational analytics, management and developer tools are provided through an always-on, linearly scalable data platform. DataStax is proud to continue to push the graph database market through the following, additional graph innovations being released in DSE 5.1.

These include:

This is the second major release for DSE Graph and we’re excited to see the product mature as quickly as it has. DSE Graph is quickly gaining adoption and leveraging your feedback to make it the best choice for large scale, graph centric challenges.

Now, let’s review some of the great features introduced in DSE Graph 5.1.  

DSE Search Enhancements in DSE Graph

DSE Search provides the DSE platform with always-on, fast, and full featured search capabilities.  DSE Graph 5.0 leveraged DSE Search backed indexes to provide distributed search capabilities to power fast, global, graph traversals.  DSE Graph 5.1 extends that foundation by providing the following additional capabilities:

Support for fuzzy searches and phrase searches

DSE Graph 5.1 provides additional predicates for those of you looking to solve fuzzy and phrase search challenges while traversing DSE Graph.

For Example, to find “David” and its one-letter misspellings you can create a DSE Search index on the first_name field of the person vertex label and simply run:

g.V().hasLabel(‘person’).has(‘first_name’, fuzzy(‘David’, 1))

Simplified the indexing mechanics of textual data with DSE Graph

With DSE Graph 5.0, textual graph data that was indexed with a DSE Search index defaulted to only tokenized search functionality, which doesn’t enable searches using full word equality semantics.  In DSE Graph 5.0, you had to make an explicit choice for graph users who wanted the ability to search using non tokenized search functionality (i.e. equality semantics) by declaring an index to be an asString() type.  We learned from your feedback that, while this explicit designation is an important feature to advanced users, most graph users simply want to declare that a graph property should be indexed by DSE Search and receive the search benefits of both tokenized and non-tokenized indexes.  

DSE Graph 5.1 provides a default mechanism for indexing textual data for this use case.  A user can simply declare that a textual graph property should be indexed as a search index and there’s no need to supply the asText() or asString() keywords.  You will receive the full benefits of tokenized and non-tokenized indexes.  The asString() and asText() index keywords still exists if you only want to explicitly configure the index of textual graph data to enable either tokenized or non-tokenized functionality.

Advanced configuration of DSE Search backed graph indexes

One of DSE Graph’s design goals is to simplify the user experience so you can leverage the power of DSE’s tightly integrated graph solution without the need to configure individual components directly.  DSE Graph achieves one aspect of this simplified user experience goal by automatically configuring all of the DSE Search configurations for search backed graph indexes.  While DSE Graph 5.0 provided a good solution for most graph users, advanced users of DSE Graph provided feedback that they would like the ability of making fine grained tuning configurations of specific DSE Search graph indexes.

DSE Graph 5.1 provides this capability by ensuring that your configuration changes of DSE Search’s key indexing configurations are preserved and honored by DSE Graph.  This feature provides you with the ability to tune index resources, performance, analysis chains, and other DSE Search specific aspects of DSE Graph.

GeoSpatial Graph Enhancements

In the initial release, DSE Graph 5.0 provided an introductory set of feature support for geospatial graph data.  As you started working with DSE Graph, we received a lot of great feedback regarding desired functionality for geospatial support in DSE Graph.  DSE Graph 5.1 now provides a robust set of features that supply a comprehensive solution for geospatial, graph data. These enhancements include:

Enhanced functionality for searching geospatial data

DSE Graph 5.1 has enhanced the Geo.inside() function supplied in the 5.0 release of DSE Graph to include the ability to specify a unit of measure, such as MILES, KILOMETERS, METERS, DEGREES, to retrieve more fine-grained search results.  The Geo.inside method has also been enhanced to support searching by linestrings, polygons, and points.  Note, polygon searching requires the addition of the JTS library to DSE Search.

Improved accuracy of geospatial data

DSE Graph 5.1 has enhanced the accuracy and protection capabilities of geospatial data by providing additional controls in the withBounds function to ensure that spatial data isn’t searchable outside of the defined spatial area, whether geo or cartesian based.

DSE Graph Performance and Stability Enhancements

DSE Graph 5.1 provides numerous enhancements that reflect the platform’s maturation.  There are too many to include in this blog.  We recommend checking out the release notes for a full listing. We’ll highlight just a few items here:

  • Significant enhancements to error messaging and client side feedback
  • Providing snapshotting controls to explicitly tell DSE Graph OLAP to persist a backup to memory, disk, or both
  • Support for all DSE authenticators in DSE Graph
  • Support for a Date data type

DataStax Graph Loader Updates

The DSE Graph Loader is a utility designed to enable fast ingestion of bulk data into DSE Graph.  In DSE 5.1, we’ve focused a lot of effort on improving the performance, usability, and maturity of the loader.  We recommend checking out the release notes for a full listing of the DSE Graph Loader enhancements and highlight a few items below:

  • The DSE 5.1 Graph Loader has improved its intelligent response to outage scenarios so that it can automatically respond to downed DSE Graph nodes and/or data centers and transparently take corrective action.  This enables DSE Graph Loader to continue loading data even if there is failure in the underlying DSE Graph data platform.  
  • DSE Graph Loader supplies a fast path when loading data using custom ids that provides a 30% increase in loading speed
  • DSE Graph Loader has improved several of the default values for loading bulk data which helps you achieve fast loading performance in many situation without having to research advanced configuration topics.
  • DSE Graph Loader’s metrics and logging have been enhanced to ensure that you have a clear and easy understanding of the end results of loading bulk data.

Inclusion of TinkerPop 3.2.3

DSE Graph leverages the standard in property graph query languages, Apache TinkerPop™ Gremlin as its core API.  DataStax is not only committed to providing the best choice for distributed graph databases, but in conjunction with the release of DSE 5.1, DataStax has also invested heavily in advancing the Apache TinkerPop tool suite, leading to a lot of great enhancements for graph querying.

Several highlights of the enhancements to TinkerPop that are driving faster, more robust graph querying as well as a better developer experience within DSE Graph include:

  • Support of a more efficient serialization format for graph traversals, GraphSON 2.0
  • Improved ability to interrupt a graph traversal in the case where an inadvertent traversal was executed
  • Enhanced configuration support for the Gremlin Console
  • Enhanced, more accurate and clear error messaging
  • Performance and stability enhancements to the Groovy Gremlin scripting engine that improve overall traversal execution predictability and speed

DSE Graph – Just Getting Started

It’s exciting to see how much DSE Graph has matured in only one release.  The innovations and enhancements in not only the core DSE Graph engine, but also the DSE platform has created a remarkable data platform that is the best choice for people who are delivering large scale applications that have customer-facing application characteristics (Contextual, Always-On, Real-Time Responsiveness, Distributed, Scalable) and data that contains highly complex or connected relationships.

You can download DSE 5.1 now and read through our updated docs for more information.

To learn more about the new capabilities in DSE 5.1 release, please check out the following blogs:





SHARE THIS PAGE
SUBSCRIBE

Comments

Your email address will not be published. Required fields are marked *

Tel. +1 (408) 933-3120 sales@datastax.com Offices France Germany

DataStax Enterprise is powered by the best distribution of Apache Cassandra™.

© 2017 DataStax, All Rights Reserved. DataStax, Titan, and TitanDB are registered trademark of DataStax, Inc. and its subsidiaries in the United States and/or other countries.
Apache Cassandra, Apache, Tomcat, Lucene, Solr, Hadoop, Spark, TinkerPop, and Cassandra are trademarks of the Apache Software Foundation or its subsidiaries in Canada, the United States and/or other countries.