Adding server power only improves site server performance to a point: Once server power is maxed out, the physical distance between a site’s host server and the visitor becomes a key component in how long a page takes to load. Content Delivery Networks (CDNs) are a widely used solution to improve load times and performance on websites by decreasing the physical distance data travels.
Not every CDN is created equal, however. Your business may opt to measure latency and acceleration to determine which CDN, if any, best suits the audience.
A server’s “megabits per second” performance only tells part of the tale when it comes to load speed. Bandwidth measures how much data moves at once, while latency measures how long that data takes to move from the source to the destination. Ookla Speedtest explains the situation with a pipe metaphor: latency measures how much time it takes for water to enter a pipe and reach the end of that pipe, while bandwidth measures the pipe’s diameter.
Moving large amounts of data, such as app updates, isn’t time-sensitive (in terms of when it starts)–so latency is not an issue here. However, latency is an extremely important performance metric for things like loading web pages, which should at most take only a few seconds.
How CDNs Work
CDNs utilize a network of servers across multiple geographical locations that mirror website content from the original source. When website visitors access a web page, their devices can receive the information from a physically closer server, reducing the time it takes for site data to reach the viewer. It’s a lot like buying milk from the corner store down the street from your house instead of driving out to a rural farm to get it.
However, CDNs don’t improve performance for everyone. For example: If the person trying to buy the metaphorical milk lived closer to the dairy farm, going to the store would take longer. Additionally, if the dairy farm and corner store were equidistant from the milk shopper, they would not see a performance boost. CDNs can also help with capacity and bandwidth management.
Calculating just how much a CDN improves a site’s performance in one location is straightforward: Measure how long a page takes to load before and after utilizing the CDN. In practice, however, testing how well a site and CDN are performing across different regions is tricky without designated testing stations. The acceleration test is often accomplished by measuring how long it takes to download files of various sizes in each tested geographical location from the mirror server and the host server.
According to cloud services provider Radware, picking the most effective CDN often requires some market research to identify where a site’s users are located. For example, a site that’s hosted out of Boston and does most of its traffic in the Northeastern and Western United States would benefit more from a CDN that improves load times in Los Angeles, Portland, and Seattle than one that boosts load times in Boston, Beijing, and London.
Acceleration testing data can help businesses make smart decisions when it comes to CDNs. Find out more about how Apica can support your CDN testing process from more than 83 countries and 2,600 monitoring nodes across the globe on our website.