LAS VEGAS — Amazon’s AWS Re:invent conference could be described as overwhelming. An audience of 53,000 people is swarming across a venue that spans seven different hotels. And Amazon Web Services (AWS) CEO Andy Jassy gave a two-and-a-half-hour keynote this morning where he announced about 20 new products and services, including everything from a new blockchain managed service to (bizarrely) a shoebox-sized machine learning car that’s available for $399. Developers were racing the car at one of the hotels immediately following the keynote. Oh, and Jassy’s speech was punctuated with a house band that sang songs alluding to his next announcement.
One of the songs the band sang was “Blackbird” by the Beatles. Can you guess what these lyrics were alluding to? “Blackbird singing in the dead of night/Take these broken wings and learn to fly/All your life/You were only waiting for this moment to arise.”
You got it: AWS’ plans to move off of Oracle databases by the end of 2019.
Jassy laid out AWS’ updated database portfolio, saying “I can assure you enterprises are singing.” He said, “The old world has been a miserable world in the last couple of decades for enterprises” that have had to work with “old guard databases like Oracle.”
He complained that companies such as Oracle and Microsoft were “abusive” to customers, constantly auditing and fining them for compliance breaches. “Overnight, if Oracle decides it wants to double the cost, that’s what they do. Microsoft decides something that’s good for them but not good for you. People are sick of it.”
In response, AWS has beefed up its own database portfolio. It announced new features to its relational database Amazon Aurora and its key-value database Amazon DynamoDB. In addition, it announced two new purpose-built databases.
Amazon Aurora, the fastest-growing service in AWS history, is a MySQL and PostgreSQL-compatible relational database built for the cloud and used by tens of thousands of customers around the world. The new Amazon Aurora Global Database offers customers the ability to update a database in a single region and have it automatically replicated to other AWS regions for higher availability and disaster recovery.
Amazon DynamoDB is a fully managed, key-value database service used by more than a hundred thousand AWS customers to deliver single-digit millisecond latency for some of the world’s largest applications. Amazon DynamoDB’s new on-demand feature provides read/write capacity provisioning, which removes the need for capacity planning, allowing customers to only pay for the read/write requests they consume.
AWS also announced two new purpose-built database services. The first is Amazon Timestream, a fully managed time series database for IoT and operational applications.
Time-series data means that it derives insights from enormous amounts of data that changes over time. Common examples include DevOps data that measures change in infrastructure metrics over time and IoT sensor data that measures changes in sensor readings over time. AWS’ new Amazon Timestream can process trillions of events per day with up to one thousand times faster query performance than a general-purpose relational database.
And AWS announced Quantum Ledger Database (QLDB), a cryptographically verifiable ledger database service, available in preview.
And the company also announced a managed blockchain service.
Jassy said AWS executives had spoken with hundreds of customers about blockchain to understand what problems they wanted to solve with it. “We don’t build things for optics,” he said. “We only build things to solve real problems.”
They learned that customers were trying to solve two different types of problems. First, they want a ledger with a central authority of trust to serve as an immutable transactional log. One of the major use cases for this ledger is to track an enterprise’s supply chain. “You need a very performable ledger,” said Jassy, in a meeting with journalists after the keynote. “It’s not something you can do easily in relational databases. We built something that turned out to be QLDB. Tons of companies that want that cryptographically verifiable ledger want that.”
Amazon QLDB is a new class of database that provides an immutable and cryptographically verifiable ledger that customers can use to build applications that act as a system of record. Amazon QLDB removes the need to build complex audit functionality into a relational database or rely on the ledger capabilities of a blockchain framework.
The second problem that enterprise customers were trying to solve was to transact together as peers, but within a decentralized framework, rather than a centralized one. “In those cases, you want those blockchain frameworks,” said Jassy.
Amazon Managed Blockchain is a new managed blockchain service for customers to create and manage blockchain networks that can scale to support thousands of applications running millions of transactions. Amazon Managed Blockchain supports two popular open source blockchain frameworks, Ethereum and Hyperledger Fabric.
“That’s what we’re starting with,” said Jassy. “With those two purpose-built tools, we think we will solve the lion’s share of what people are trying to solve.”
*11/29/18 Update: After publication of this story, Herain Oberoi, AWS’ general manager of databases, analytics, and blockchain, told SDxCentral that by the end of 2018, 88 percent of Amazon’s workloads would be migrated off Oracle databases to Aurora or DynamoDB, and 97 percent of Amazon critical services will be migrated off Oracle databases by the end of this year.