Amazon Web Services (AWS) has announced that Amazon Aurora, the MySQL-compatible database engine, is now available for customers in three AWS regions - U.S. East, U.S. West, and European Union.
The service was first announced by Andy Jassy, senior Vice President, AWS in November 2014 at AWS re:Invent Conference held in Las Vegas. At that time, the database engine was made available only to selected companies for beta-testing. After the nine-month-long beta-testing, AWS has now officially launched the new database service on AWS cloud that will target AWS customers with large database workloads. According to AWS, the service will expand to new regions over time.
Amazon has been offering its Relational Database Service (RDS), a managed MySQL database service, since 2009. The RDS allows Amazon customers to provision a reliable and scalable database server with a point and click in just a few minutes. Over the past five years, AWS has added support for Oracle, PostgreSQL, and Microsoft SQL Server in RDS. Amazon Aurora is the newest addition in the RDS portfolio. AWS says the current customers of Amazon RDS for MySQL can migrate to Amazon Aurora easily using Amazon Aurora's new features. It can be done with "a few clicks from the RDS console."
Main Features of Amazon Aurora
Aurora is a cloud-native MySQL-compatible relational database engine which provides the speed and availability of best commercial databases along with the cost-effectiveness of open source databases. Amazon claims Aurora will provide the scalability and reliability of the commercial databases at a price which is one-tenth of the price of those databases. It can handle more queries simultaneously (both write and read) than other products. Amazon also claims the new service is able to deliver five times the price-to-performance of a traditional relational database running on the same class of hardware.
"Amazon Aurora provides up to five times better performance than MySQL at a price point one-tenth that of a commercial database while delivering similar performance and availability." says the AWS website.
Each Aurora instance can deliver over 500,000 select queries and 100,000 updates per second. The service continually backs up data to Amazon Simple Storage Service (S3). In event of current database failure, data is restored within a minute using the back-up. The service is wire compatible with MySQL 5.6 using InnoDB storage engine. This feature enables users to easily migrate from MySQL through the standard utilities such as mysqldump.
The storage scheme of Aurora uses cluster volumes, where each cluster volume is a single, virtual volume containing copies of data from various availability zones in a single region. Each cluster volume is also backed by SSD storage. This architecture of Aurora ensures backing up data to S3.
For redundancy, Aurora supports a maximum of 15 read replicas across availability zones in the same region. For users to achieve cross-region replication, it is required that users maintain a separate MySQL database instance outside RDS. A RDS-based MySQL database instance and Aurora database clusters must be in the same region to replicate.
Amazon says migrating to Aurora has considerably improved the performance of applications of many MySQL customers, along with the benefits of cost effectiveness and better availability. Substantial improvements are usually observed in Read replica lag and latency. Moreover, Aurora could be a more cost effective alternative than MySQL, thanks to its "ability to use read replica for fail-over and pay only for the actual storage used by the database." The product site promises 99.99% availability of the service.
According to Amazon's documentation, Aurora doesn't yet support the feature of automated cross-region replication through the RDS console. The feature of integration with the AWS Key Management System is also not available currently. Amazon Aurora is also the only database in RDS without support for the AWS Key Management System or encryption at rest.
"Today's commercial-grade databases are expensive, proprietary, high lock-in, and come with punitive licensing terms that these database providers are comfortable employing. It's why we rarely meet enterprises who aren't looking to escape from their commercial-grade database solution. Now, with Amazon Aurora, companies can get at least the same availability, durability, and security as commercial-grade databases for one-tenth of the cost." explained Raju Gulabani, Vice President, Amazon Database Services.
Customers will pay by the hour for the primary database instance and any replicas. It will cost $0.29 per hour for database instances of 2 vCPUs and 15.5 GB memory and $4.64 per hour for 32 vCPUs and 244 GB memory.
Customers will also pay $0.10 per GB per month for storage, and will get a total of six copies of data, with two copies in each availability zone. For I/O requests, customers will pay $0.20 for every million I/O requests made by the database.