Amazon Web Services Inc. (AWS), a company that reflects a general industry trend towards more specialized, purpose-built database, has launched Amazon Neptune. This high-performance graph database is delivered as a fully managed service.
AWS announced a Neptune preview last November. It stated that the core of Amazon Neptune was a purpose-built graph database engine engine optimized for storing billions and querying the graph with milliseconds delay.
Graph databases are being developed to provide better data analysis and lookups in an ever-connected world of ubiquitous networking and Internet of Things (IoT). Graph databases are not traditional SQL lookups, nor the newer NoSQL databases. Graph databases are focused on connectivity and storing data about network edges and vertices, or relationships. Graph databases can link data sets and allow them to be retrieved in one operation.
AWS stated that graph databases are useful for connected contextual and relationship-driven data. “Some examples of applications include social media networks, recommendation engine, driving directions, logistics and diagnostics, as well as fraud detection and genomic sequencing.
A relational database is required to create such applications. AWS stated that multiple tables with multiple foreign key would be required. This would result in complex joins and unwieldy queries.
Neptune supports the following graph database models: Apache TinkerPop3 style Property graphs queried using Gremlin; Resource Description Framework (RDF queried using SPARQL).
AWS stated that if you have applications that work with TinkerPop or SPARQL, you should be able start using Neptune by simply updating your endpoint.
Amazon Neptune is available in the US East (N. Virginia), US East(Ohio), US West, Oregon, and EU (Ireland).
Comments are closed