
LiveAction this summer expanded its network management and analytics solutions via the acquisition of Savvius. In a recent interview with The New Intelligence (TMC’s (News - Alert) new artificial intelligence event brand), Darren Kimura and John Smith of LiveAction explained the thinking behind the combination and discussed how network management has changed and will continue to evolve.
Kimura said LiveAction’s core technologies focused on using things like NetFlow and SNMP from routers and other gear. But, with software-defined networking, he added, access to those devices is not as good as it used to be and API data off controllers is not real time anymore. So, the value of doing network monitoring based on those sources has diminished.
However, the Savvius solutions drills down to the packet level, he added, which is the source of truth. Combining all these data sources allows network specialists to get enriched analytics for better visualization and insights, said Kimura, LiveAction’s vice chairman and chief strategy officer.
A typical LiveAction user in a network operations center sees a screen with red, yellow, and green images, he said. If that user sees red in Mumbai, for example, he or should knows there’s a problem. That user can then click from the macro view of the Mumbai site to the branch view to see what’s happening – like if the network is dropping packets there. But that’s where the LiveAction detail stopped, Kimura said.
However, Savvius now lets users do things like double-click on red at the branch level to see exactly where that problem occurred. That provides them with more detailed information. The fact that users can do all of the above through one interface and one common workflow makes things easier.
LiveAction is also working on a solution for the service provider market that will be able to collect information about a large number of end user devices, anonymize it, and use that to create recommendations for other carriers.
Smith, LiveAction’s CTO, said that network management has changed significantly in the past few years, and it’s poised for even more important change.
Ten years ago, network management relied mostly on statistical analysis, said Smith. But, there’s no longer normal statistical distribution, he said, so doing mean average quantiles and things like that doesn't work anymore. So, it’s hard to understand patterns on a per site, per application basis.
Machine learning now allows network operations and planners and security people to see patterns at this scale, he added. Over time, LiveAction analytics could also potentially help provide recommendations based on policies in intent-based networks. But, taking that next step requires a new kind of thinking.
“Right now people are in the trust phase,” he explained. “They’ve built automated systems. But they don't quite trust them yet.”
Learn – or share – the benefits and challenges of employing artificial intelligence and machine learning for network management at the upcoming Adaptive & Intent-Based Networking expo, part of TMC’s The New Intelligence. The event is planned for Jan. 30 through Feb. 1 in Fort Lauderdale, Florida.
TMC is accepting session and speaker suggestions for this event through July 31. To submit your ideas, please visit our Call for Speakers form at this link.
Edited by Erik Linask