Couchbase Reduces Latency by 80% for Distributed Database Solutions Using AWS Local Zones

Founded in 2011, Couchbase’s mission is to empower enterprises to develop, deploy, and run mission-critical applications anywhere by delivering a high-performance, flexible, and scalable modern database. Its NoSQL database solutions combine the processing of key-value information in memory with the distributed storage of JSON document data without schemas. With five corporate offices across three countries, Couchbase serves enterprises of all sizes, including Emirates Airline, Tommy Hilfiger, Wells Fargo, LinkedIn, DirectTV, and Marriott Hotels.

Because hyperfast responsiveness is integral to Couchbase’s value proposition, the company relies on edge computing to deliver its solution to customers. That type of computing delivers data processing, analysis, and storage close to its customers’ endpoints. However, to support edge computing, its customers had to run their workloads in either independent or colocation data centers near their data sources. “One of the hardest things about taking advantage of edge computing is that it’s highly distributed,” says Wayne Carter, vice president of engineering at Couchbase. “Prior to the release of AWS Local Zones, the way you did edge computing was to stand up a data center in a metropolitan area. To do that in one metropolitan area is hard enough. Do that across 20 or 30 areas, or do that across multiple continents, and the operational complexity was just so high.”

In 2020, Couchbase announced that it would offer its cloud-based NoSQL database in the AWS Marketplace, which provides customers with the ability to find, test, buy, and deploy software that runs on AWS. Later, AWS began developing Local Zones, and Couchbase decided to become an early adopter of this new solution. “What excites us most about AWS Local Zones is how easy it makes it to get the advantages of edge computing,” says Carter. “Massively high speeds are in reach of any developer.”

Before recommending Local Zones to its customers, Couchbase conducted a series of latency tests. Engineers deployed a simple web application backed by the Couchbase database to an AWS Availability Zone and a Local Zone, then accessed the app from both zones in parallel. Engineers measured and compared the latency on the network and on the database during a full round trip.

The company measured three factors: the total time on the server to process the request, database time on server, and total round-trip time from client to server to client. Throughout the process of testing Local Zones, Couchbase received support from the AWS team. “Our customers care about speed and reliability,” says Carter. “They look to us to advise them on solutions, so we were working with the AWS team to make sure that we were using AWS Local Zones correctly.”

During this process, Couchbase built on its use of Amazon Elastic Compute Cloud (Amazon EC2), which provides secure and resizable compute capacity for virtually any workload. In addition to deploying applications using AWS Availability Zones, the company can deploy Amazon EC2 instances in Local Zones, helping its customers run their database solutions closer to their endpoints and enhancing the performance of their business applications. After running hundreds of iterations of its latency test, Couchbase experienced an average of 80 percent reduction in latency once it switched to Local Zones for certain use cases. “AWS Local Zones integrate with ease into our customers’ use of AWS,” Carter says. “They’re able to gain those higher speeds without having to take on additional operational complexity.”

Now, Couchbase can deliver its NoSQL database solutions without requiring its customers to set up and manage on-premises data centers. Using Local Zones, Couchbase’s customers can run all of their infrastructure in the cloud, including edge workloads, which makes managing their compute resources simpler. Moreover, because AWS offers pay-as-you-go pricing, Couchbase’s customers can increase their cost savings by moving away from a hybrid infrastructure model. In a recent benchmarking survey by Altoros, Couchbase outperforms competitors by nearly 400 percent in terms of throughput and costs up to 90 percent less to operate.

As a result of helping its customers optimize their resources, the company has lowered the barrier to entry for adopting its database solutions. Local Zones help Couchbase’s broader business strategy of facilitating edge computing for its customers, including large enterprises and small startups. “We’re seeing an increase in customers using edge computing because it’s so much easier,” says Carter. “They don’t have to manage the largely distributed data centers that they would have to do if this service weren’t available.”

Couchbase is looking to grow its use of Local Zones globally as it proceeds with an international expansion that will offer its low-latency database services to an even broader range of customers. Additionally, the company is exploring other AWS services to improve database speed, enhance performance, and reduce operational costs. “We really like the breadth of edge computing solutions that come from AWS,” says Carter. “We are excited to try more AWS services and see how we can further decrease compute costs and move to even smaller form-factor edge computing devices.”