Kentik Data Engine
Architected for tomorrow
Kentik Detect™ is powered by the Kentik Data Engine™ (KDE), a clustered, high-availability datastore that’s purpose-built for the unique performance requirements of network monitoring. KDE handles streaming ingest at Internet scale and delivers speedy responses to ad-hoc, multi-dimensional queries across trillions of data points. Fast, multi-tenant, open, and secure, KDE is bottleneck-free and massively extensible, making Kentik Detect ready today for the visibility needs of tomorrow.
The big data difference
Network utilization keeps growing, but visibility hasn’t kept pace. Why? Because as traffic explodes, so does network data. To get actionable intelligence from that data, you need near-instant ingest and near-infinite capacity. Existing visibility tools can’t do that. Designed when computing and storage were scarce, they’ve got built-in bottlenecks that compromise both monitoring and forensics. That’s what makes the Kentik Data Engine different: it’s a purpose-built big data platform that’s optimized for network visibility.
Learn more about big data as a value multiplier for network visibility.
The engine at work
Kentik Data Engine ingests continuous, Terabit-scale network data from routers, switches, and host agents on server infrastructure. Available within moments, this traffic and performance data is fused into a single time series for a unified view. KDE is multi-tenant, enabling response within seconds to simultaneous queries from multiple users, and it stores unsummarized data for months, allowing detailed forensic look-back. In short, Kentik Data Engine provides the speed, breadth, and detail that’s lacking in existing visibility tools, whether appliance-based or open-source.
Read how KDE is designed for optimal multi-tenant performance.
Open and secure
Kentik Detect provides open access to your KDE data via PostgreSQL clients and REST APIs. Cloud-scale performance means that you’re never limited in how you can much data you can access. That makes it easy to integrate with customer portals, internal monitoring systems, and automated DDoS mitigation systems. But Kentik Detect is also highly secure. Data can be encrypted in transit, and each customer’s instance of KDE runs completely segregated on the back end, with no access path from any customer to the data of another.
Questions about the details? Check our Knowledge Base article on Kentik Security.
Flexible and extensible
Appliance-based monitoring is typically static, lacking not only capacity but flexibility. But Kentik Detect gives you a constantly expanding range of network data fields, plus the ability to apply tags and also custom columns that are populated at ingest. Used in ad-hoc analyses and multi-dimensional anomaly detection, custom columns enable you to tailor Kentik Detect to your own business and network context, which frees you to optimize the network experience for your users and customers.