Risky Business: Implementing a Redundant Networking and Multi-CDN Monitoring Strategy
In this article, we’ll recap the key takeaways from Catchpoint's recent webinar with AWS on the importance of implementing a comprehensive redundant networking and multi-CDN strategy.
Last month, we partnered with AWS to put together a webinar on the importance of implementing a comprehensive redundant networking and multi-CDN monitoring strategy. You can replay the event in full here.
In this article, we’ll recap the key takeaways covered by the panel of experts who included Leo Vasiliou, Director of Product Marketing at Catchpoint, and Steve Campbell, our Chief Strategy Officer. Joining from AWS were Christopher Isenburg, Specialty Sales, Edge Services and EJ Chen, Solutions Architect and Edge Specialist.
What is digital interaction?
The panel kicked things off by setting up a base understanding of the concept of digital interactions.
Digital interactions: any interaction between a client (human or robot) and an organization that is only possible due to digital technologies.
Digital experiences are the summation of these interactions, and can be perceived as fast, slow, or unavailable.
The importance of quality individual digital interactions is increasing over time, according to Steve. These interactions aren’t simply trivial or entertaining. They make up a growing portion of our daily lives, from paying bills online to working at a a fully remote job.
The long-term success of SaaS and XaaS (anything as a service) organizations depends on maintaining consistent, reliable, available, and fast service delivery chains. Otherwise, users accustomed to high-quality experiences will opt for the next available service that can more quickly meet their needs.
The case for digital modernization
The adoption or conversion of services from brick-and-mortar to online is speeding up. Users expect everything to be accomplishable online and they want it to be both fast and easy. This push for digitization has also impacted the underlying approach to architecting these services, requiring IT professionals to rethink how they support their service delivery chain. At the same time, the delivery chain has become almost completely decentralized, existing outside and beyond their control.
“The Internet is your network,” Leo said. “The cloud is your data center. CDNs are your cabinets. Slow is the new down -- reachability is the new availability.”
Multi-provider usage (Catchpoint 2021 SRE Report)
What does this mean in actuality? Let’s take a simple example of a used car marketplace.
As a user, you have the option to log in using an email address or a social media account like Facebook or Gmail. While browsing used cars, there’s a microservice contacting the Google Maps API to help you visually sort listings in your area. There’s also likely a microservice that will contact appraisal services like Kelly Blue Book in order to provide you with the fair market value of the car. Go even further and you’ll realize this marketplace is relying on multiple CDNs to serve high-quality pictures of these cars, and cloud solutions like AWS to handle database and storage needs.
The user’s digital experience is seamless and comprehensive, but the underlying technology making it so is completely decentralized and reliant on multiple vendors, solutions, and APIs to be successful. Any one of those could degrade, fail, or completely change with little –to no notice.
Would a user continue to browse used car listings if the images failed to load? It’s not likely. They’d simply search for another solution.
Despite the obvious risks of offloading so much functionality to third-party providers, the alternative is far worse: delivering lackluster services that can’t match the expectations of today’s users. Organizations must modernize their digital presence and rely on a multitude of third-party providers to create competitive services that users find valuable.
How monitoring has evolved into observability
Modern digital experiences aren’t meticulously controlled by one organization’s internal team with internal software, servers, and databases--they’re everywhere. Therefore, “everywhere” is exactly where you should be monitoring.
But how do you monitor from everywhere? You can’t deploy agents for lambda functions. The available surfaces for such tools are quickly eroding. Instead, the monitoring landscape is shifting into one focused on observability.
As our Chief Marketing Officer, Nik Koutsoukos, recently phrased it, “The monitoring tools of yesterday typically focused on one or more domains to identify typical and known problems… On the other hand, observability takes a more holistic approach, examining multiple outputs of the system to infer system health.”
Are your services observably performant?
Are all of your dependency services observably performant? In order to know this, you have to be prepared to capture performance metrics from multiple vantage points across your entire network which, as we’ve learned earlier, is the entire Internet.
Application delivery chain (Catchpoint)
There are a lot of upsides to embracing the shift towards observability, which include:
- Improved resiliency
- Improved scalability
- Improved performance
- Improved reachability
You can read more about improving these four key areas in other Catchpoint blog posts.
How to use observability to support a multi-CDN monitoring strategy
EJ Chen showcased several great examples of how you can use CDN metrics (for example, cache statistics from Amazon CloudFront) in addition to observability data from Catchpoint. This data can be used to investigate issues like higher than expected response times for retrieving your cached CDN content.
Since Catchpoint can capture key details provided within Amazon CloudFront’s HTTP response headers, you can use that information to run performance tests. You can also map test nodes to responses, which adds more dimensionality to your log data, allowing you to reveal performance trends tied to various CDN locations.
Here’s an example: let’s say you are using a service that provides edge locations in Mexico City and New York. Upon investigating your test node logs in Catchpoint, you notice that content delivered to your audience near Mexico City has a slow response time. After drilling into the data, you discover that your content is being served from the New York edge location.
In this scenario, it’s likely that you have a configuration issue that needs to be resolved so that your audience near Mexico City can be served by their closest edge location. This will also improve your CDN costs.
EJ also outlined additional cases for utilizing Catchpoint with AWS, so make sure you watch the webinar all the way through! (Hint: You can find those use cases at the 19:35 minute mark.)
Support your network and services by observing performance from every market
When setting up a multi-CDN strategy, it is essential to support your network and services by observing their performance from every market that your users access. That’s one of the reasons Catchpoint is perfectly poised to lead the digital experience observability market - we provide the largest Global Observability Network in the industry. This means you can set up test observers across more locations--both domestic and international--than with any other observability solution.
Want to learn more?
For more information:
- Sign up for a Guided Tour.
- Visit the Catchpoint Store on AWS Marketplace.
- Read more about how Catchpoint can help you set up a multi-CDN support strategy.