Measuring Performance Gains from AWS Cloudfront - Testing from the Edge

by Interzoid Team

---

Posted on May 28th, 2020

---

---

AWS Cloudfront is one Edge Computing solution delivering significant performance gains through global caching networks. But how does one measure results?

Edge Computing is an implementation of distributed computing whereby data and computation occurs closer to the location it is needed, rather than in one globally-accessed, centralized location. In practice, this means Web page content, digital objects, and other files are replicated and stored on multiple servers around the world, enabling a client to access the nearest server network to obtain and view or interact with the assets over the Web. The content is only swapped out when it changes, greatly reducing the geographic distance data must travel from and to a browser or device upon request. This significantly increases speed on behalf of a user, including viewing content from Web applications, mobile devices, and IOT devices.

This time delay reduction is accomplished with a geographic distribution of endpoints that access local servers to provide content. The decreased latency is not limited to locally cached pages and files. The distributed endpoints typically also provide access to high-speed backbone networks, reducing network jumps when accessing data or functionality that does exist elsewhere on the globe, such as real-time account access or retrieving real-time data. The performance increases here can also be substantial.

According to Grandview Research, Edge Computing spending was $3.5 billion USD in 2019, and will increase to over $43 billion in 2027 (37% annual growth).

AWS Cloudfront is one Edge Computing solution, where S3 (Cloud storage) objects, including an entire Web site, can be cached on local servers around the world. It also enables leveraging the AWS global backbone network, efficiently routing content to an edge location that can best serve Web content.

Cloudfront is a service that is fairly easy to turn on, only requiring some configuration parameters to begin serving content through the AWS edge content delivery network (CDN).

Click here to see the locations of Cloudfront edge server locations on six continents.

Once an Edge Computing solution is in place, however, how does one test for the performance increases presumably delivered? One solution would be to physically travel around the world to test performance gains from different international cities, but that would be time-consuming and expensive, and obviously not continuous.

Another more efficient idea would be to leverage a performance testing product that itself is a global network of servers, loading and measuring the performance of Web content retrieval from different geographic locations.

Interzoid's Global Performance API requires parameters only for the Web address of the content (such as an .html, .css, or .js file, or even other APIs), and then also the testing location. That's it. A roundtrip performance measurement is then obtained for loading the content from the specified geographic location. Combining multiple measuring locations and recording measurements at certain time intervals can provide valuable performance insight for your organization. Here are some example measurements:

As one example, one can get a sense of who has Edge Computing implemented in the the world of digital media. Take a look at these testing results for various news outlets and their global access performance rates.

Sample API call:

Here is the technical detail page for the Global Performance API.

It is also available in AWS Marketplace, enabling usage through your existing AWS account.