This guest blog post is by Elad Rave, co-founder and CEO of Teridion. We thank Teridion for being a sponsor.
Imagine this: You’re stuck in traffic, but you have a navigation app that tells you which routes are congested and which routes are free. Great, right? Not necessarily. This information isn’t very helpful if you’re boxed in by other cars and can’t actually go in another direction.
Your Internet traffic faces a similar problem. Over the last decade, the monitoring industry and its many Real User Monitoring (RUM) and synthetic monitoring tools have grown by leaps and bounds, but at the end of the day, the tool you select can only provide visibility, not control.
Some may have a limited ability to direct traffic based on DNS entries, or permit you migrate across CDNs. Some might optimize traffic within the cloud, but not between the cloud and the user.
However, none provide a globally scalable end-to-end solution that optimizes end-user performance in real time. Visibility without control leaves you on the outside looking in.
Teridion takes a completely different approach by providing visibility into the performance of routes across cloud service providers, and by taking advantage of cloud-based networking to send your traffic across the best-performing link in real time.
This is particularly useful given the increasing number of applications and content hosted in public clouds, be it AWS, IBM Cloud, Digital Ocean, or dozens of regional providers. Thus, it makes perfect sense to leverage the cloud for networking along with compute and storage.
In the figure below, ISP routes include those that are operating normally, as well as some that may be congested or even down. Measurement agents from Teridion are deployed in Cloud Service Provider (CSP) data centers to calculate performance, reflective of the underlying ISP topology. Data from all agents is correlated and stored in a database.
Now, when your traffic needs to flow from one location to another, the management system will draw on this database to determine where to place the virtual routers. The virtual routers will set up the path using overlays to direct the traffic across the most optimal route.
Consider measurement agents in one or more provider data centers in Sydney, Tokyo, San Jose, Seattle, and Chicago. These agents are not limited to a single provider. For example, if two providers are available in San Jose and Tokyo, each will link.
Measurement agents constantly measure latency, throughput, and packet loss, forming a heat map of cloud performance that is in fact a reflection of the underlying ISP performance. If routes between Seattle and Chicago are congested, the map will reflect this. Depending upon the application, the path selected could be optimized for latency, throughput, and geography (i.e., don’t transit a specific country).
Now, if you are in Sydney and need to access an application in Chicago, instead of hopping between one ISP and another, with very little control, you have an option. You can now jump onto the cloud in Australia, and take the fast lane to your destination, all the while knowing that it is the optimal path. For the first time, instead of a bystander to Internet performance, looking in through the window, you now have control.
There are, of course challenges, including building an accurate heat map and calculating the optimal paths. Ensuring sufficient network performance across cloud compute resources that sometimes have very different characteristics can also pose challenges. And of course you must account for variations in the cloud providers’ global performance and interconnections.
Nonetheless, Teridion’s approach is the most viable option available today, offering application developers a chance to realize their vision, unfettered by limitations imposed by the Internet.
To request a customized performance analysis, before and after Teridion, or to arrange for a free trial, contact firstname.lastname@example.org.