MongoDB Atlas delivers enterprise-grade performance at scale on Amazon EBS

With tens of thousands of customers relying on MongoDB Atlas, the company’s AI-ready cloud database platform, MongoDB needed its underlying storage to be able to scale dynamically. “As the world continues to evolve in the era of AI, applications and markets have become more dynamic, and the ability to elastically scale storage has become critical,” says Ralph Capasso, vice president of engineering for MongoDB Atlas Data Services.

Although MongoDB Atlas includes native storage auto-scaling capabilities, the company requires high-performance, durable storage that provides a global reach. Modern applications can grow from hundreds to millions of users overnight, and data needs to be virtually instantly available. Every minute of waiting can translate to lost productivity.

MongoDB Atlas also needs to be able to respond to demand spikes, which is why MongoDB worked with the AWS team to enhance Amazon EBS Elastic Volumes for more flexible modifications. Through this collaboration, MongoDB further strengthened Atlas’s ability to support growing and increasingly variable workloads without disruption on AWS.

MongoDB Atlas uses Amazon EBS as its cloud block storage foundation for AWS workloads. To help address scaling needs for its workloads, MongoDB adopted the Amazon EBS volume portfolio. Amazon EBS General Purpose Volumes balance price performance for the majority of workloads. “Amazon EBS gives our customers the ability to scale more rapidly and effectively,” says Capasso.

For performance-critical and input/output-intensive workloads, MongoDB relies on Amazon EBS io2 Block Express Volumes, which deliver up to 256,000 IOPS and 4,000 MB per second of throughput per volume. MongoDB combines these volumes with instance store and Amazon Elastic Compute Cloud (Amazon EC2) instances, which provide secure and resizable compute capacity to flexibly match the right storage tier to each customer’s performance requirements. “Every AWS database in our fleet right now is powered and backed by Amazon EBS, which is all about scaling rapidly. Using Amazon EBS gives us the ability to respond to our customers and deliver real value through our MongoDB Atlas offering. That flexibility is what helps us meet customers where they are,” says Capasso.

In addition to scale and performance, Atlas must run smoothly by adjusting to changing customer workloads. To provide customers with an improved ability to tune, MongoDB worked alongside AWS Enterprise Support—which helps businesses accelerate innovation, strengthen security, and streamline cloud operations—on key innovations. MongoDB shared requirements and feedback and helped the Amazon EBS team innovate to improve the volume modification experience for Elastic Volumes.

Customers who need to respond more quickly to demand spikes can now better keep up with demand by tuning volume performance more frequently to better match their workloads. Amazon EBS volumes can receive up to four Elastic Volumes modifications within a rolling 24-hour window.

MongoDB Atlas gives organizations the ability to ship modern applications faster. Using Amazon EBS, Atlas automatically matches the right storage to each customer, so they get the performance they need without managing infrastructure decisions.

Atlas does this on a global scale across 34 AWS Regions. Using Amazon EBS features, MongoDB Atlas delivers a flexible database service without trade-offs between performance, cost, and simplicity.

Working alongside AWS, MongoDB is well positioned to support modern workloads and businesses. MongoDB’s customers can respond to demand spikes in near real time, matching storage performance to their workloads. This translates to an improved experience for both MongoDB’s customers and their users.

“It was a neat process working alongside AWS to push the boundaries of what’s possible,” says Capasso. “It was a great demonstration of how every customer is important. It has made our products more flexible and dynamic, which means we can meet our customers where they are.”