We sit down with Jonathan Lacefield to discuss the latest trends in graph processing, from DataStax perspective? Talk through how the challenges being associated with graph are evolving, current tips/tricks, tools being seen, etc.
0:15 - Welcoming Jonathan back to the show and a quick review of his background in product management
1:56 - Jonathan gives us a quick overview of graph database history at DataStax - Aurelius, Titan, Apache TinkerPop, Gremlin, and key figures like Marco Rodriguez and Matthias Broecheler
5:48 - How using Cassandra as the pluggable storage layer for Titan led to the creation of DataStax Enterprise Graph
8:40 - The natural relationship between graph processing and analytics/graph analytics. Traditionally this has meant extracting operational data to a separate database for graph analytic processing, but DSE Graph supports both operational and analytic queries, and the trend is toward abstracting the operational/analytic choice from the application developer
12:35 - DSE Graph has been in 3 releases: 5.0, 5.1, and 6.0. The trend has been toward improving scalability and ease of use. We’re working toward a unified multi-model architecture in which all data can be accessed CQL, Gremlin and Spark SQL.
17:28 - Problems being solved with graph databases - Customer 360, Authorization, Entity Resolution
22:22 - Future directions for DSE Graph - continued support for multi-model, hybrid cloud, writing data once and accessing through the right API for the job - CQL, SQL, Gremlin
25:33 - Privacy is an emerging use case for graph, especially the ability to give users more selective control over what personal data is maintained
27:37 - Graph query languages like Gremlin, Cypher require a paradigm shift for developers - domain specific languages provide a way to minimize this learning curve
33:20 - DSE 6 is the foundation for a lot of great things for DSE Graph, and we’d really like your feedback on how to make it even better