CTO Jonah Kowall introduces Kentik’s AIOps solution for network management, explaining the need for evolution and innovation in the network performance monitoring and diagnostics market and providing an overview of the components of Kentik’s revolutionary platform.
Kentik has been many things to many people. Our company was created with a vision of solving any problem for any user wishing to analyze and mine their network traffic data. Our platform has been proven to scale to support the largest networks in the world—while providing rapid query and responses—without rolling up data, thanks to our data architecture. Our sophisticated users, with complex needs and advanced imaginations, have been able to answer any question in near real-time, across any environment, without needing to plan in advance.
That’s what makes Kentik unique in the network monitoring and analytics market.
The ongoing challenge is that today’s network professional is being overwhelmed by complexity and diversity, making it difficult to know what questions to ask. The trends towards multi-vendor, overlay-driven networking and the addition of new data centers in the public cloud are requiring network professionals to be generalists—needing to know a little about a lot of different networking and application delivery technologies. At the same time, these networks are complex, so it’s also essential for network teams to have depth and expertise at a level that’s typically scarce within any given organization.
This is the reason Kentik has been working hard through 2019 to build features that bring forward both the questions and answers that the modern network professional needs.
AIOps from Kentik
Today, Kentik introduces the first AIOps platform specifically for network professionals. With this platform, Kentik is providing insights, visualizations, and greater capabilities to take action.
The focus areas for Kentik are intuitive onboarding and workflows to make managing networks easier. We accomplish this both by surfacing interesting or anomalous data, and by making it much easier to implement integrations with your favorite automation technologies.
By collecting and correlating large volumes of network traffic data, associated metadata, and other third-party data in real-time, Kentik is able to turn these disparate data sources into instant insights about business-impacting network conditions—insights which were not possible to glean before.
An open, API-driven system with many pre-built integrations is necessary for the multi-vendor network infrastructures today. Kentik has built support for many leading technologies (e.g., Streaming Telemetry and Flowspec), while also expanding our platform to support SNMP and Syslog. This means additional devices can quickly be added and supported across disparate data sources. We’ve also recently added support for Palo Alto Network devices, Cisco ASA, Cisco NBAR and AVC, along with Cisco SD-WAN and Meraki MX product lines.
Our integrations leverage our ability to ingest and enrich data in real time across multiple data sources. Our scalable data pipelines are called K/Ingest and K/Enrich and, with these platform services, we can correlate traffic data with threat feeds, custom DNS information, user information, and other elements not found in the network data stream itself. This correlation can be done both at query time or at ingestion time—depending on the user requirements for real-time alerting and actions based on the newly combined data.
One of the top initiatives in any network professional’s short list is automation and taking action on both data and insights. With K/Automate we have built-in support for understanding traffic changes and being able to automatically take actions such as blackhole traffic by, leveraging Flowspec, automatically open tickets in ServiceNow, and even notify your teams via PagerDuty or OpsGenie with a click of a button.
K/Advise provides the ability to surface interesting anomalies in context. This means when there are large deviations or other anomalies, the user is presented with that data, front and center.
K/Advise enables actionable data to be presented to the user without any configuration or other adjustments after the data is ingested to the platform.
The network monitoring market is full of stagnation, even to the point where Gartner is no longer publishing the “Network Performance Monitoring and Diagnostics (NPMD)” Magic Quadrant beyond 2019. This lack of innovation is readily apparent. Consider, for example, how on-premises appliances are still the norm.
Gartner’s initial goal of adding the requirement of “diagnostics” into the network performance monitoring Magic Quadrant was to show that tools had to evolve from being reactive, packet-based technologies into proactive, machine learning (ML) driven systems.
As the first AIOps platform for the network professional, Kentik has done just that. Expect to see much more as we continue rolling out capabilities in the months ahead!