SteelCentral AppInternals: APM Tough Enough For Cloud-Native Containerized Apps
As enterprises build their applications and processes for the cloud, they're coming to rely more and more on containerized microservices: formerly monolithic apps broken up into multiple self-contained single-purpose services that can run across distributed environments. With microservices, you can develop and deploy applications in an agile fashion and without using too many resources—but such systems are complex to monitor and understand.
The good news: SteelCentral AppInternals can provide the visibility you need to get the full benefits for a containerized architecture.
- It's hyperdistributed, so it works at any scale
- It's high definition, collecting one-second metrics to keep pace with dynamic and transient environments
- It's multi-stack and works with a variety of platforms and container software
Playing well with others
AppInternals is agnostic when it comes to most of the platforms and container orchestration tools it can work with, and we’ve built native support where it was needed. So no matter what you've got in your stack, chances are we can work with it.
Compatible platforms include:
• Pivotal Cloud Foundry
• AWS EC2
"The Essential Guide to Monitoring Containers and Microservices," a new e-book from Riverbed, goes into more depth on how AppInternals provides visibility into even the most complex microservices architectures. We'll dive into the high points here to give you the big picture, and offer a glimpse into how AppInternals delivers.
Microservices: Flexibility at a price
Microservices allow companies to be more agile in delivering new features, because individual application functions can be developed independently. This makes it faster to QA and deploy new features, and improves reliability because each microservice functions independently rather than being dependent on other parts of the application. And microservices frequently run in containers—isolated user-space instances that work like lighter-weight versions of VMs—allowing them to be distributed across cloud and hybrid cloud architectures and make more efficient use of system resources.
But all those benefits come at a price. In cloud-native application architectures, transactions are served up by a complex web of microservices. And the environment a containerized application runs in is itself constantly changing. Containers live fast and die young—a good proportion of instances run for less than ten seconds.
All that has an impact on application performance. You might have a shared resource dependency. You might have too much CPU allocated, or not enough. But most monitoring solutions just aren't up to the task of gathering, analyzing, and presenting the system data you need second by second; often, they rely on sampled or triggered traces, which provide an incomplete picture. Too often, customers find themselves using multiple monitoring technologies, collecting so many different types of data that it becomes difficult to pull everything into one transaction-centric view for troubleshooting. And that just doesn’t scale.
AppInternals stands alone
SteelCentral AppInternals is the only APM solution that's up to the challenge, providing unparalleled data quality at any scale. AppInternals can trace every transaction across every application tier, stitching them all together across tens of thousands of components with a single analysis server. You have access to data that's complete, that's presented in context, and that's correlated so you can trace individual transactions across a complex distributed infrastructure.
AppInternals can achieve this because it eschews sampling. Instead, it captures complete data thanks to Riverbed's proprietary big data technology—and does it while adding practically negligible performance overhead.
Get the ebook
"The Essential Guide to Monitoring Containers and Microservices" gives you a deeper dive into how Riverbed AppInternals makes it possible to monitor application performance within a microservices environment without compromises. In particular, you'll learn how AppInternals meets the five requirements for next-generation APM tools:
Registration is free, so download the e-book today
AppInternals in action
AppInternals' easy-to-deploy, non-intrusive instrumentation monitors both apps and the environment they run in, as shown in Figure 1.
Figure 1. Deploying AppInternals
Once deployed, SteelCentral provides automatic discovery and mapping of apps, containers, and their relationships, along with integrated analytics. Figure 2 illustrates an application map in the AppInternals UI. As you can see in the lower part of the window, these visualizations and mappings are container-aware.
Figure 2. AppInternals application map
With these visualizations, you can determine which user owns each transaction and how those transactions traverse your back-end infrastructure. That's functionality that only Riverbed can provide.
Find out more
For more details on all of this, register to download our ebook, "The Essential Guide to Monitoring Containers and Microservices"—it's free to download. You should also check out more information about Riverbed's partnership with Pivotal to integrate AppInternals with their Cloud Foundry environment.