In its continuing quest for cloud dominance, Amazon has taken aim at the major players in the relational database space with a new product it calls Aurora.
Amazon Web Service head Andy Jassy told an audience at the company’s Re:Invent conference this morning in Las Vegas that the company set out to build a new database designed to operate in the company’s cloud.
Aurora is designed for high performance and reliability, with support for up to 6 million database inserts per minute, and 30 million selects per minute. Data stored in Aurora will be replicated 6 times across 3 availability regions, and the database will automatically back up information to the company’s S3 storage service. In addition, Aurora will self-heal around errors, and maintains its cache even through a server restart.
Users who have their existing data stored in MySQL can move back and forth between Aurora and a MySQL database using Amazon’s migration tool so that they can try out the service, and switch back to their old MySQL install if necessary.
It’s also cheap: running Aurora on a r3.large instance only costs 29 cents an hour. There aren’t any minimum costs for running Aurora, and customers will only get charged for the resources they use.
Today’s announcement is a strike against large relational database software providers like Microsoft, which has its own SQL Server product. SQL Server has been an important part of the company’s cloud strategy, especially as it pushes its own Azure cloud platform.
People interested in giving the new service a spin can sign up for a preview of Aurora here.