Overcoming the Cloud Migration Blues

log blueCloud.jpg
David Pitlik

Like a thundering herd of wildebeest in the Serengeti, the great corporate cloud migration is irreversibly underway. According to 451 Research, 90 percent of companies are already in the cloud, and 60 percent of workloads will be running on a hosted cloud service by year’s end—a 15 percent increase from 2018.

Migrating to the cloud offers greater agility, enabling companies to create new services, roll out new applications faster, and introduce new ways to communicate with customers. Of course, like nearly anything else worth doing, migrating and operating in the cloud is fraught with considerable challenges. According to Ray Krug, solutions architect at NETSCOUT, gaining visibility across on premises and cloud environments is one area that has been historically difficult.

IT is tasked with migrating applications to the cloud, which necessitates understanding all of the components and their interdependencies, he said in a recent conversation with the CUBE’s Dave Vellante. “An application may be talking to a database or an ancillary application downstream,” he said, and unless you understand and fully account for those dependencies, you might fail to migrate certain critical components, “which will then have a ripple effect that impacts other business processes.” Complicating matters, he said, IT often can’t just go to the documentation team to uncover all the dependencies, because either the developers of the application are no longer with the company or the documentation itself simply doesn’t exist.

The Secret to Migration Success: Turning Wire Data into Actionable Smart Data

The typical approach to understanding the performance of components is to use rudimentary network data to look at how applications communicate with one another. for example, IT might look at various logs or infrastructure information, such as CPU utilization, performance information. The shortcoming with this approach is that IT can’t see the entire picture, including all dependencies.

“It’s a lot of bespoke data,” said Krug, “and trying to bring that together and come up with a conclusion that this is all the components, and this is how it’s performing is tricky.”

Instead, Krug explained that wire, or packet, data is the key to fully understanding how everything is communicates. “The challenge with wire data is that it’s hard to interpret,” he said. “NETSCOUT has solved that problem by utilizing patented Adaptive Service Intelligence technology, which translates wire data into meaningful key performance metrics, also known as smart data. Our nGeniusONE analytics platform then takes the metadata and allows us to display service dependency maps.” Basically, he said, NETSCOUT provides the ability to look into all the nooks and crannies to discover the dependencies as well as the performance of applications and the network.

Starting with wire data has several major advantages. Regardless of whether you are using a hybrid on-premises/cloud or multicloud model, wire data is everywhere. IT can instrument into both the cloud and data center to gain invaluable visibility into traffic. ASI then takes the metadata and brings together a view of all service dependencies across the infrastructure, with uniform metrics.

“NETSCOUT is able to instrument inside the workloads to gain access to the traffic going in and out of virtual machines in different clouds, allowing us to translate it into key performance metrics,” concluded Krug. “Whether it’s workloads in the cloud, private cloud, or data center, we deliver visibility end to end. And that visibility is the key to migration success.”

Learn more about nGeniusONE

Watch the Ray Krug CUBE Interview below

Ray Krug from NETSCOUT Discusses Cloud Migrations

David Pitlik is a long-time technology and business writer and frequent contributor to NETSCOUT’s blog.