AWS Updates Aurora and DynamoDB, Unveils Neptune Graph Database
Among the raft of new product announcements made by Amazon Web Services (AWS) during Wednesday morning's re:Invent keynote were significant updates to its database portfolio.
AWS CEO Andy Jassy introduced previews of two new capabilities in the Amazon Aurora relational database service, which first launched in 2014 and has become the fastest-growing service in AWS history.
The first is Aurora Multi-Master, which Jassy described as the industry's "first relational database service with scale-out across multiple datacenters." Aurora Multi-Master lets users create "multiple write master nodes" of their databases across several availability zones, insulating them from failures when one availability zone is disrupted.
Aurora Multi-Master is currently in preview with support for single-region master duplications, Jassy said. Support for multiple regions is expected in 2018.
Also in preview is Aurora Serverless, touted by Jassy as an on-demand, auto-scaling, serverless database service. Aurora Serverless is essentially a new version of the original Aurora service that has all the same features but is able to adjust much more dynamically according to a customer's usage rate. There's no need to provision instances upfront, according to Jassy. Aurora Serverless automatically scales up and down as workload capacities change, and powers down resources when they're not in use. Users are billed by the second.
According to a blog post Wednesday by AWS evangelist Jeff Barr, Aurora Serverless is expected to become generally available with MySQL support in the first half of 2018, followed by PostgreSQL support in the second half. Those interested in accessing the Aurora Serverless preview can sign up here.
Jassy also unveiled two new features for its DynamoDB NoSQL database service: Global Tables and Backup and Restore.
Now generally available, the Global Tables feature makes DynamoDB "the first fully managed, multi-master, multi-region database in the world," according to Jassy. While DynamoDB already replicates users' tables over multiple availability zones, Global Tables extends that replication across multiple regions, boosting customers' fault tolerance and expanding the global reach of their applications.
According to a separate blog post by Barr, the Global Tables feature is currently available out of AWS regions in Northern Virginia, Oregon, Ohio, Ireland and Frankfurt, with more coming next year.
Meanwhile, the new Backup and Restore feature lets users easily perform backups of hundreds of terabytes of data on demand. The backup component of this feature is now generally available in the Northern Virginia, Ohio, Oregon and Ireland regions, while restore capabilities will come in early 2018. That includes the ability to do "point-in-time" restores to any second within a 35-day window.
A new managed graph database service called Neptune is now in limited preview from AWS, addressing the challenges inherent in building apps with highly connected data.
"The core of Amazon Neptune is a purpose-built, high-performance graph database engine optimized for storing billions of relationships and querying the graph with milliseconds of latency," said Randall Hunt, senior technical evangelist at AWS, in a blog post Wednesday detailing the new product.
Neptune supports the Property Graph and the Resource Description Framework (RDF) graph models, as well as the Apache TinkerPop Gremlin and RDF SPARQL query languages. It's also fully manged, with AWS doing the back-end maintenance, management and patching tasks for customers.
More from AWS re:Invent 2017:
Gladys Rama is the senior site producer for RCPmag.com and senior editor of AWSInsider.net.