DigitalCube Improves Performance for Thousands of Customers Using Amazon CloudFront

Web hosting company DigitalCube wanted to accelerate its cadence for issuing regular updates to enhance performance and security for its customers’ websites. However, with each update, the company’s small developer team was spending hours individually deploying updates to thousands of distributions, leaving it limited time to work on new products and features.

DigitalCube needed a faster, simpler update solution, and it found one on Amazon Web Services (AWS). The company started using a new feature from Amazon CloudFront—a content delivery network service built for high performance, security, and developer convenience. Using new feature CloudFront Functions—which users can use to write lightweight functions in JavaScript for high-scale, latency-sensitive content delivery network customizations at AWS Edge locations—DigitalCube can scale quickly and perform thousands of updates simultaneously. For more compute-intensive edge workloads, the company continues to use Lambda@Edge, a feature of Amazon CloudFront that lets users run serverless code at AWS Regions, which improves performance and reduces latency. By using CloudFront Functions as well as Lambda@Edge, DigitalCube discovered that it could save not only time but also money while improving performance for its customers.

Based in Japan, DigitalCube is a web hosting provider for WordPress websites at scale. Businesses use the company’s tools to build secure and scalable websites with no coding and fully managed maintenance. DigitalCube was an early adopter of AWS. It has used AWS services since 2010, and it built one of its primary products—a static site generator and serverless hosting solution called Shifter—using AWS.

Software updates are an essential part of DigitalCube’s business. The company issues regular updates to its software and the code underlying its customers’ websites to improve performance and strengthen security. However, deploying updates took up a significant amount of time that the company’s developer team wanted to use for other projects. The team was using Lambda@Edge, which required it to deploy updates for distributions individually. “We have to update over 4,000 distributions,” says Seiji Akatsuka, customer success manager at DigitalCube. “With so many distributions, we needed a more lightweight solution that was better suited to our use case.”

As the company and its use cases evolved, it needed a solution that was better suited to deploying updates at scale. After reading about the feature in an AWS blog post, the company decided to migrate its update workflow to CloudFront Functions. DigitalCube began the migration in May 2021 and completed it in under 2 months, helping the company quickly become faster and more efficient.

Previously, DigitalCube spent several hours completing each of its update deployments. Now, the company can deploy updates across thousands of distributions quickly. “Using CloudFront Functions, we’ve reduced update deployment times from 4 hours to 2 minutes,” says Okamoto Hidetaka, engineer at DigitalCube. CloudFront Functions is well suited to this large-scale, low-latency use case because the feature can run code geographically close to the source of a request coming in or a response going out. It performs this logic at edge locations, running code closer to users to deliver higher performance. Meanwhile, DigitalCube still uses Lambda@Edge for certain use cases and specific customer requirements. Because the company can use these two services in combination with ease, it can achieve better flexibility as well as substantial time savings. By migrating to CloudFront Functions, DigitalCube has reduced its overall Amazon CloudFront costs by 60 percent.

In addition, faster updates and less maintenance work result in better quality throughout DigitalCube’s offerings. The company can quality test more simply on AWS because it can run tests that affect thousands of distributions without having to scale internally or introduce downtime. As a result, DigitalCube achieves high visibility across its infrastructure. “High visibility is a major factor for us,” says Daniel Olson, chief operating officer of DigitalCube. “Being able to fix things at scale more quickly is a huge asset.” For example, if a customer requests a change or an engineer notices a problem, DigitalCube can identify and resolve it much more quickly. The speed at which the company can identify and diagnose issues at scale and deploy updates to fix them makes running its continuous integration and continuous delivery pipeline simpler. “Previously, we couldn’t do anything else while we were performing updates,” says Hidetaka. “Now we can spend the 3 hours we saved on other projects.”

Further, the company’s improved efficiency has produced a better environment in which its team can innovate. “So much of our sunk cost used to be a result of time spent on maintenance,” says Olson. “The developer experience is much better now that we’re using CloudFront Functions.” These benefits are invaluable for a small team working to support thousands of customers. “Using AWS is a great choice because it provides a range of valuable services,” says Hidetaka. “We can focus on providing better, faster, more secure websites for our customers.”

By migrating to CloudFront Functions, DigitalCube has accelerated its update workflow, reduced costs, and improved the performance of its customers’ websites. The migration has also provided the company’s developers with an opportunity to work on important projects and innovate.

Having seen the value of learning about and adopting additional AWS services, DigitalCube plans to continue growing and evolving on AWS. The company is currently working on implementing new product features for its web hosting solution Shifter using AWS. “Adopting Amazon CloudFront was a giant leap forward for all the work that we do at DigitalCube,” says Olson. “We’ve hit our stride with it, and there’s still so much potential to do even more.”