Amazon Web Services’ RDS for Aurora is transformational for the dbPaaS market. Aurora’s availability, scalability, performance and low cost should make it a primary DBMS consideration for CIOs, DBAs and information managers.
On 27 July 2015, Amazon Web Services (AWS) put Amazon RDS for Aurora (Aurora), its MySQL equivalent RDBMS platform as a service (dbPaaS), into full public availability as a managed RDBMS engine for Amazon Relational Database Service (Amazon RDS). Although accessible from anywhere in the world, it is available in the US East (Northern Virginia), US West (Oregon) and Europe (Ireland) regions, with other regions to follow. Aurora is fully compatible with Oracle MySQL 5.6 with InnoDB and is the first RDS RDBMS engine built by AWS. Aurora was originally announced at AWS re:Invent 2014. More than one thousand AWS customers participated in a nine-month preview, and many have applications in production.
Aurora is engineered from the ground up to integrate with AWS infrastructure, not to run on top of it, as other RDS services do. This allows Aurora to take full advantage of the S3 storage system, AWS three-data-center redundancy and many other AWS services, including the management console and conversion utilities. The result is a higher-performing RDBMS with better availability, due to its use of S3 with its "11 nines" of availability (i.e., 99.999999999%).
Aurora has arrived with support from many third-party vendors, such as Tableau, Looker and Talend, with more to follow. It has full compatibility with RDS for MySQL (as proven by most of the early adopters of the preview) and also supports MariaDB connectors. AWS has also validated application compatibility with MySQL, so applications and third-party software that currently leverage a MySQL database will run on Aurora virtually unchanged. AWS offers a simple, single-step migration utility (through the management console) to convert from MySQL on RDS to Aurora. Aurora's tight integration with the cloud infrastructure of AWS gives it high performance, fast recovery, scalability to 64TB of storage and high availability — all at a lower price point than other RDS products.
The downside of Aurora is minimal. There are some features, such as encryption at rest, coming in the near future. Additionally, moving from other RDBMS engines other than MySQL (such as Microsoft SQL Server, Oracle and PostgreSQL) will require more time and effort to convert to Aurora.
Its preview showed promise for large, mission-critical, scalable transaction systems. In addition to pushing hard against competitive offerings (from both AWS and others, such as Microsoft Azure and Oracle Cloud Database), Aurora will become the AWS RDS standard and will attract many applications looking to leverage cloud infrastructure. For these reasons, we believe that Aurora will be game-changing in the DBMS market not only as an alternative for current dbPaaS implementations (effectively killing RDS for MySQL) but as a serious alternative for many on-premises applications previously not considered for cloud implementation.
Enterprises using RDS for MySQL should seriously consider a switch to Aurora, for both new and existing applications, leveraging the performance, availability, scalability and lower cost.
CIOs, database administrators (DBAs) and information managers should consider dbPaaS as an alternative to current infrastructure. This is now possible with products such as Aurora, even when the applications require high levels of availability, scalability and resiliency.
Enterprises interested in Aurora should watch for additional third-party support and continuous new functionality as Amazon adds regular incremental upgrades.