Riverbed & Pivotal: Big Data Power for the Microservices World


More and more IT shops are turning to containerized microservices to help build flexible and portable applications. But with formerly monolithic apps being broken down into numerous interlinked components spread across local servers and cloud platforms, application monitoring for these architectures has become increasingly complex — and many tools simply aren't up to the task.

Figure 1. Cloud architectures are becoming increasingly complex

That's where SteelCentral AppInternals comes in. Thanks to our partnership with Pivotal, you can deploy AppInternals natively within the Cloud Foundry platform, so you can take advantage of this containerized microservices architecture without sacrificing visibility into application performance. AppInternals is uniquely suited for this environment:

  • AppInternals can scale to gather and retain all the data you need to keep tabs on your applications.
  • AppInternals is simple to deploy into a containerized environment and its licensing model is easy to understand.

"We have a distinct advantage," says Riverbed Director of Product Management Peco Karayanev, "because Riverbed has a big data platform that can do this end-to-end monitoring in ways other vendors can't." Let's dive in and see how this partnership works.

No data left behind

In a containerized architecture, applications are broken apart into individual services that can be portable across servers or networks. That makes for great flexibility, but it also adds complexity; an application that might have once only had five dependencies could, in a microservices world, be rebuilt with twenty or more, any one of which could break or become a performance bottleneck.

Your APM tool needs to be able to trace across all those tiers — and a single analysis console needs to be able to process and stitch together all of those distributed traces. But the truth is that many tools simply don't have the processing capabilities to keep up, relying instead on sampled transactions and incomplete traces. AppInternals gathers all the data you need and won't cut corners even for applications with tens of thousands of components, thanks to its hyperscale clustered architecture. AppInternals collects fine-grained systems data, such as CPU and memory metrics, at second by second intervals. If a shared resource dependency impacts a transaction’s response time, we can catch it.

And with the big data technology underlying AppInternals, you can be sure that the data you need is retained for later analysis. "If you're trying to diagnose a problem," explains Karayanev, "you don't want to just learn that at  9:00 a.m. yesterday you had an error message. You need to know what the message specifically said, and what the impacts were. In these environments, containers can come alive and then disappear as needed. If you’re not using a big data approach to track the forensics, you’re losing the audit trail of these problems."

The power of simplicity

AppInternals has a suite of capabilities that you might otherwise need to cobble together a stack of tools from different vendors to replicate. Because it's a single, standardized product, you don't need to worry about training staff on a customized suite of monitoring programs. And thanks to our partnership with Pivotal, AppInternals runs natively inside the Pivotal environment — in essence, it becomes a part of the Pivotal stack when you deploy it.

Figure 2. AppInternals runs natively inside the Pivotal environment


Riverbed's licensing terms also make it easy to use AppInternals and Pivotal together. Unlike other vendors that charge extra to use tools in containerized environments, Riverbed's licensing is based on what you use, not what you install. "The license consumption is a fraction of what a legacy application would use," says Karayanev.

An end-to-end solution

The bottom line when it comes to containerized architectures is that you can’t find the needle in the haystack if you’re only grabbing a handful of hay at a time. The scale at which you need to monitor today’s applications is a big data challenge and requires big data technology like SteelCentral AppInternals to address.

And, of course, your microservices-based applications don't exist in a vacuum. They connect across WANs and LANs to end-user devices, and AppInternals' integration with other SteelCentral products means that you'll be able to monitor your whole network and hunt down bottlenecks wherever they lie — whether that's on a containerized app in the cloud or an end user's cell phone. Only Riverbed can offer those capabilities.

If you're already using Pivotal Cloud Foundry and want to test out AppInternals, sign up for our 14-day free trial. You can also find out more about AppInternals or read about our partnership with Pivotal on our website.