IPDR/SP
- Media Piracy Detection
- Based on correlative usage characteristics, detects rebroadcasting of linear TV/PPV TV content like live sports.
- DOCSIS Device Clone Detection
- Provide a network-wide view of MAC addresses. Simple queries can detect devices with duplicate MAC addresses to identify theft of services.
- Capacity Planning
- Provide a comprehensive view of the individual usage characteristics on the various network partitions. By aggregating these individual usage points into larger segments, users can build a model to understand CAGR by product and prioritize and optimize capital and operational expenditures.
- Top Talkers
- Enables highlighting of the biggest network hogs, enabling the operator to apply various mitigation strategies, e.g. move to DOCSIS 3.1 or upsell to a higher bandwidth package.
- Congestion Monitoring
- Enable monitoring by service group and get utilization metrics that highlight how micro-bursts of congestion may be affecting users QoE.
NetFlow/IPFIX
- Offload Router-Based Processing
- Enable higher sampling rates (1:2000 vs 1:10000) and consistent exporting of NetFlow/IPFIX data. Routers will preserve their primary function under heavy load and stop sending out NetFlow/IPFIX in extreme conditions when flow data is needed most.
- Replication target 1: Arbor Networks DDoS Scrubbing appliances.
- Replication target 2: DeepField user application discovery service.
- Per-target filtering removes known/static flows from being replicated. This enables substantial cost savings by not sending e.g. IPTV flows (obviously not an attack) to DDoS mitigation.
- Create an Avro data feed to enable various consumers within the org to subscribe to various feeds from the VelociData.
- Enable higher sampling rates (1:2000 vs 1:10000) and consistent exporting of NetFlow/IPFIX data. Routers will preserve their primary function under heavy load and stop sending out NetFlow/IPFIX in extreme conditions when flow data is needed most.
Telemetry
- Deploy more classes of service and need a reliable way to monitor how full the router buffers are on each class and eventually dynamically balance the buffer sizes. Model-Driven Telemetry and Juniper Telemetry Interface were the only way to accomplish this goal. VelociData is collecting from thousands of routers from multiple router vendors and providing a normalized, consistent real-time dashboard view into the buffer capacity for all the routers and alerting when anomalistic conditions exist.
BGP
- Collect monitoring statistics of routing changes and memory usage to support routers that are BGP peers.
IPv4-IPv6
- Reduce cost by side-stepping purchasing more precious IPv4 space through the use of MAP-T protocol to ferry IPv4 Data over IPv6 address space. Telemetry from Arista is used to stand up a dashboard that monitors the number of successful mappings and the number that are unable to be processed in the Arista device. These are pushed to another system that can handle the (hopefully) low-volume exceptional cases. Alerts are set up in the case a sudden larger fraction is offloaded to the other device.