Enterprise Network Performance Analytics

by | Mar 28, 2017 | Blog

RASA Networks and Aruba/HPE

RASA

ArubaLogo

The words “Cloud-Based Analytics” and “Data Science” are, to most of us, buzzwords marketing folks throw around to add some sizzle to normally pretty boring stuff. But in the case of RASA Networks and what they are bringing to the table for Aruba/HPE… it is more than mere marketing bluster.

The founder of RASA Networks – a company Aruba/HPE purchased in May of 2016 – is Partho Mishra. Partho has a history with Marvell and Cisco before launching RASA. His idea was to capture the thought processes of a Wireless LAN Professional, someone who understands how Wi-Fi works, and take those sometimes complex and convoluted techniques from gathering information, processing that information, then making changes and seeing the results… all that slow and expensive processes many of us do for our clients and turn it using machine learning into an ‘cognitive system’ that could be used by enterprises to optimize their Wi-Fi systems.

To this end they worked with data scientists and Wi-Fi domain experts like us to gather and process ‘big data’ from sensors throughout the enterprise and can recreate that process of a consultant. But it grew much larger and offers a superior solution today. Instead of merely re-creating what an expert consultant might do, they’ve enhanced the system to now accomplish things not even the best experts could do manually!

They started with a sensor-based solution to gather data points to build a huge data set that could be employed to answer wide ranging questions. But since joining Aruba, they no longer need an sensor-based overlay platform; they gather data from existing Aruba access points and controllers. Management of Aruba has helped RASA (now called RASA Analytics) by allowing them to ask for just about any metric they might need to be output from existing Aruba access points. (this alone is quite a feat… getting details that might not be significant to actual throughput issues on an AP – and yet are very valuable for data analytics).

Now with an enterprise full of Aruba AP’s can answer questions that we might not have even had as individuals. This analytical engine can see things on a campus-wide scale to solve issues outside the scope of any one set of analysis tools. Since it is cloud-based, it can also look across campuses. As professionals, we pride ourselves with our tools. We can show up on site and use survey tools, packet analysis tools, and spectrum analysis tools – but those are only looking in our actual location… with RASA’s big data, they can see wider and farther than anything within an individual’s scope.

This is still new and an emerging science… but I can foresee results of this collected data and its associated insights being used in future versions of ARM, Clarity, Clientmatch, and Airwave to improve their functionality. Though in the short term, whichever enterprises can use RASA’s features will be able to quickly and efficiently optimize and tune their current Wireless LAN infrastructure.

I’m looking to watch to this space and see what Aruba and RASA can do together to make Cloud-based Analytics and Data Science part of how we manage WLANs in the future.


 

As part of ATM-17 Insiders, I was privileged to sit in on many Deep Dive sessions to better understand different up and coming technologies. This session on Aruba’s Machine Learning and Analytics was presented by Partho Mishra and Sujatha Mandava. Though not a product available today – this technology should be productized soon.