Cisco’s announcement of Tetration Analytics means that IT and network leaders can no longer ignore the need for Big Data intelligence. In this post, Kentik CEO Avi Freedman explains how that’s good for Kentik, which has been pushing hard for this particular industry transition. Cisco’s appliance-based approach is distinct from the SaaS option that most customers choose for Kentik Detect, but Tetration’s focus on Big Data analytics provides important validation that Kentik is on the right track.
Tetration Announcement Validates Big Data Direction
I’d like to welcome Cisco to the 2016 analytics party. Why? Because while Cisco didn’t start this party, they are a big name on the guest list and their presence means that IT and network leaders can no longer ignore the need for Big Data intelligence. And that is good for those of us who have been tooting our horns for this particular industry transition.
For those of you who missed the news, Cisco just announced Tetration Analytics, a full rack appliance meant to collect sensor data from data center infrastructure and analyze it with Big Data and machine learning power. Cisco claims that customers will be able to analyze billions of data records in a second.
The Tetration Analytics announcement is a watershed moment in the embrace of Big Data analytics by the IT and network industry. It signals the product mainstreaming of what has largely been the domain of home-grown or purely services-driven open source software projects. Cisco’s seal of approval indicates that Big Data has finally arrived for operational network and infrastructure management, an area long overlooked by the mainstream Big Data analytics sector (not least because reliance on Hadoop put the required performance beyond reach).
Compelling Business Case
By offering a multi-million dollar rack of integrated hardware and software for data center analytics, Cisco has recognized the demand for a more out-of-the-box solution. The company has clearly calculated that it can offer a compelling business case for high-performance analytics to organizations for whom the DIY or services-based model would have been at least as expensive — and significantly riskier.
Architected properly, one platform can satisfy operational, business, performance, and security intelligence.
Of course, Tetration Analytics isn’t just about standalone analytics. It exists in the context of a multi-billion dollar business segment where Cisco has been investing heavily, including Insieme, AVC, and its take on converged server and data center networking architectures. We believe that architected properly, one platform can satisfy operational, business, performance, and security intelligence in a scalable and open way. So it will be intriguing to see where Cisco takes Tetration.
Will Tetration go to the WAN? To the Internet Edge? Given ACI, AVC, and the on-chip performance instrumentation that Cisco has built, will they extend to NPM and APM functionality? We see happy customers of Cariden and Tail-f products, which Cisco has kept open. It will be telling to see how multi-vendor, open, and integrate-able Tetration will become. Or will it remain within a walled garden of Cisco-only infrastructure?
Bullish on Big Data
Whatever Cisco eventually does with Tetration, we’re very bullish on this development. In part that’s because we’re huge believers in Big Data analytics; it’s in Kentik’s DNA. For years, so many of our peers in the networking industry complained that the visibility provided by single-box or federated appliances was far too slow, shallow, and costly to give them the operational insight they need. We founded Kentik to meet the demand for highly granular intelligence using Internet-scale network data delivered at operational speed.
Cisco’s approach provides important validation that we’re on the right track.
As we grow, we see dozens of web companies, streaming music providers, digital enterprises, service providers, and many other organizations using Kentik Detect. And we see how excited they become when they can store and query tens-to-hundreds of billions of instantly accessible flow data records and perform ad hoc, multi-dimensional analytics with sub-second response times. Cisco’s approach, which addresses similar requirements for feeds, speeds, and analytics, provides us with an important additional validation that we’re on the right track.
Another reason that we welcome Cisco’s announcement is because we’re complementary. In an era where IT is obsessed with public and hybrid clouds, it’s impossible to miss the fact that Cisco has taken a hardware-centric, private cloud approach. While we offer on-premises deployments, the vast majority of our customers opt to consume Kentik Detect as a SaaS offering. They love the fact that they can go from setup to full utilization of a super-powerful network intelligence platform in fifteen minutes — without having to deploy any hardware or software. That said, we’ve already seen demand for architectures like Tetration: strictly in-perimeter and with tightly-coupled analytics.
So… Big Data network analytics is real. The party is on. Cisco’s announcement has drawn attention from millions of folks to the very point we’ve been making all along: no organization should settle for decade-old 1RU technology when the power of Big Data analytics — Internet scale at operational speed — is now available for IT operations. Are you ready to join the 2016 analytics party?