re:Invent 2017 – AWS Updates Aurora, DynamoDB and Unveils Neptune Graph Databank Among the raft new product announcements by Amazon Web Services (AWS), Wednesday morning’s reInvent keynote was a significant update to its database portfolio. Andy Jassy, CEO of Amazon Aurora AWS, presented previews of two new capabilities within the Amazon Aurora relational data service. This service was launched in 2014 and has been the fastest-growing in AWS history. First is Aurora Multi-Master. Jassy called it the industry’s first relational database service that can scale out across multiple datacenters. Aurora Multi-Master allows users to create multiple write master nodes of their databases across multiple availability zones. This protects them from any failures if one zone is down. Jassy stated that Aurora Multi-Master is currently in preview and supports single-region master duplicates. 2018 will see support for multiple regions. Jassy also has Aurora Serverless in preview. This is a serverless database service that provides on-demand, autoscaling support for multiple regions. Aurora Serverless, which is basically a new version the original Aurora service, has all the same features and can adjust dynamically to customer usage rates. Jassy says that there is no need to provision instances in advance. Aurora Serverless scales automatically as workload capacities change and shuts down resources when not in use. The second billing cycle is used to bill users. A blog post Wednesday by Jeff Barr, AWS evangelist, stated that Aurora Serverless will be generally available with MySQL support in half of 2018, and then PostgreSQL support the second half. Register here to access the Aurora Serverless preview. DynamoDB Jassy also revealed two new features for DynamoDB NoSQL’s DynamoDB NoSQL Database service: Global Tables, Backup and Restore and According to Jassy, DynamoDB’s Global Tables feature is now generally available. It makes DynamoDB the “first fully managed, multimaster, multi-region database in all of the world.” DynamoDB already replicates tables across multiple availability zones. Global Tables extends that replication across multiple locations, increasing fault tolerance and expanding global reach for customers’ applications. Barr also posted a blog post stating that the Global Tables feature is available in Northern Virginia, Oregon and Ohio, Ireland, and Frankfurt. More regions will be added next year. The new Backup and Restore feature allows users to perform easy backups of hundreds terabytes on demand. This feature is currently available in Northern Virginia, Ohio and Oregon, while restore capabilities are expected to be available in early 2018. This includes the ability to perform “point-in time” restores to any second within a 30-day window. Neptune A new managed graph service called Neptune, which addresses the challenges of building apps with highly connected data, is now available from AWS. Randall Hunt, senior technical advocate at AWS, stated that Amazon Neptune’s core is a high-performance, purpose-built graph database engine that can store billions of relationships and query the graph with milliseconds delay. He also described the new product in a blog post. Neptune supports the Property Graph, the Resource Description Framework (RDF), graph models, and the Apache TinkerPop Gremlin as well as RDF SPARQL and RDF SPARQL queries languages. It is fully managed by AWS, who perform back-end maintenance, management, and patching tasks on behalf of customers. More information from AWS reInvent 2017: