Amazon uses NoSQL databases to run their web stores, and they tend to be quite secretive about what technology their using. Earlier, they used Oracle for their data but eventually switched to their own databases.
Amazon doesn’t openly disclose their database systems, but the Amazon jobs page for their “Fast Data Technologies” team offers a clue:
“Amazon’s Fast Data Technologies team builds platforms to share, store, move, and process data for our retail and digital businesses. We own one of the largest NoSQL databases globally, handling trillions of requests daily. We develop top-notch solutions using AWS technologies, and create our own when necessary. We help Amazon’s internal businesses manage, optimize, and act on their vast data efficiently.”
Contents
They Move Away From Oracle
According to this CNBC news, Amazon moved away from Oracle for more scalable data structures. Although Amazon planned to use their own services, they didn’t specify the database. They also transitioned their data warehouse to their petabyte-scale solution, Amazon Redshift.
The current CEO of Amazon, Andy Jassy, tweeted that after moving from Oracle, they’d mainly use DynamoDB and Aurora.
Though the article is a few years old, Amazon may have changed their technical infrastructure since then.
DynamoDB by Amazon
Amazon likely uses DynamoDB, a high-performance NoSQL database from AWS. This fully managed cloud database supports document and key-value data models. DynamoDB offers high performance, scalability, and availability for applications needing fast access to large data volumes.
One of the primary benefits of using DynamoDB is its scalability. It handles large amounts of data, meeting the demands of massive applications. It also offers high performance and availability, ensuring applications run smoothly, even under heavy traffic.
Amazon Aurora
CNBC also reported on Amazon Aurora as a notable database service.
Amazon Aurora is a powerful relational database service available through Amazon Web Services. Compatible with MySQL and PostgreSQL, it was announced in 2014 and has quickly become a popular choice for cloud databases.
According to the Amazon AWS website, Aurora is three times faster than PostgreSQL and five times faster than MySQL. This makes it an excellent choice for a commercial database due to its reliability, availability, performance, and cost-effectiveness.
Why does Amazon use Aurora?
Amazon chose Aurora because it offers the same functionality as MySQL and PostgreSQL with added high availability and durability. Aurora is a flexible option that can be scaled up or down quickly according to business needs.
An important feature of Amazon Aurora is its 99.99% availability. Data is replicated six times across three Availability Zones, with continuous backups to Amazon S3. This ensures your data remains safe and accessible, no matter the circumstances.
Aurora doesn’t just provide high availability and durability; it also ensures top-notch security. With Amazon VPC, you can set up a private network for your database, isolating and securing your data. Additionally, Aurora encrypts data at rest using keys you control through AWS Key Management Service.
- High availability & durability
- Scalable & flexible
- Strong security
This could all change very quickly
Amazon, like other tech giants, constantly evolves. I’m not entirely sure which specific NoSQL implementation Amazon uses for their web stores. Even if I knew, this could change anytime!
The main thing I know for certain is Amazon is using a highly scalable architecture for their operations. It’s definitely not a simple database structure!