High Capacity Caching Servers
By deploying extremely powerful machines with large amounts of cache we are able to insure that requests are served quickly directly from the edge. All of our cache servers have several terabytes of cache capacity spread out across several tiers in each server: RAM, SSD, and Spinning Disk.
Tiered Caching Architecture
No matter how large you make your edge caching servers, there is just no way to cache everything on them. Using a multi-tiered architecture allows us to have even more data cached, even if it isn’t directly at the edge of the network. The tiered cache allows for longer tail content to live closer to the end users and eliminates long or possibly slow trips back to the origin for content that might have fallen out of edge cache.
Instant Content Refreshes
Rather than purging large objects from cache when their TTL expires, we will check back with the Origin Servers with a simple query to find out if the object we have is still valid to keep in cache, thus resetting the TTL. This saves bandwidth at the Origin Servers and allows for much faster cache fills. If the origin server goes online, we can configure out platform to respond with stale content so users will still have a like-normal experience.
Most network communication protocols were developed many decades ago when networks were a lot less reliable. Because of that they, by default, are tuned to be reliable rather than fast. As networks have become more dependable it has become common place to take out some of the overly chatty checks and balances and re-tune the settings to focus on speed. We have done that pretty extensively and we have the highest throughput rates in Asia to show for it!
End-to-End Control and Visibility
We provide full control over all of your content and configurations on our platform. We offer a self service portal with a simple and clean user interface to make configuration changes and purge stale content. If a user portal isn’t your thing, we offer APIs to handle all of those task programatically. In addition to being able to work with configurations and content, we provide detailed statistics about your traffic in real time via our APIs. Everything from throughput, to hits per second, and everything in between.
Many case studies have found that page load times directly impact a companies bottom line. As users have more buying options and competition is greater, users will have even higher expectations.
Amazon found in a performance case study that 1 second slower page loads times would cost them roughly 1.6 billion dollars in sales. Test our next-generation CDN platform and see how much faster your site can load with a free trial.
Google found that as little as four tenths of a second was enough to drop search traffic by 20%. This of course has a major impact on their advertising revenue. Drop us an email and lets discover how we can accelerate your business.