Piggybacking control and data plane telemetry for machine learning-based tunnel failure prediction
AssigneeCisco Technology, Inc.
InventorsJean-Philippe Vasseur, Grégory Mermoud, Vinay Kumar Kolar
Life of this patent
Estimating the term…
Abstract
In one embodiment, a device identifies one or more telemetry data variables for use to predict failure of a tunnel in a software-defined wide area network (SD-WAN). The device sends a Bidirectional Forwarding Detection (BFD)-based telemetry request towards a tail-end router of the tunnel that requests the one or more telemetry data variables. The device receives the requested one or more telemetry data variables. The device uses the received one or more telemetry data variables as input to a machine learning-based model, to predict a failure of the tunnel.
Claims19
A method comprising:
- identifying, by a head-end router of a tunnel in a software-defined wide area network (SD-WAN), one or more telemetry data variables for use to predict failure of the tunnel, wherein the one or more telemetry data variables comprise one or more of:
- a queue statistic, memory usage, or processor load of a tail-end router of the tunnel;
- sending, by the head-end router, a Bidirectional Forwarding Detection (BFD)-based telemetry request towards the tail-end router;
- receiving, at the head-end router, the requested one or more telemetry data variables;
- and
- using, by the head-end router, the received one or more telemetry data variables as input to a machine learning-based model executing locally on the head-end router, to predict a failure of the tunnel, wherein the machine learning-based model is trained to predict failure of a given tunnel in the SD-WAN.
An apparatus, comprising:
- one or more network interfaces to communicate with a software-defined wide area network (SD-WAN), wherein the apparatus is a head-end router of a tunnel in the SD-WAN;
- a processor coupled to the network interfaces and configured to execute one or more processes;
- and
- a memory configured to store a process executable by the processor, the process when executed configured to:
- identify one or more telemetry data variables for use to predict failure of the tunnel, wherein the one or more telemetry data variables comprise one or more of:
- a queue statistic, memory usage, or processor load of a tail-end router of the tunnel;
- send a Bidirectional Forwarding Detection (BFD)-based telemetry request towards the tail-end router;
- receive the requested one or more telemetry data variables;
- and
- use the received one or more telemetry data variables as input to a machine learning-based model executing locally on the head-end router, to predict a failure of the tunnel,
- wherein the machine learning-based model is trained to predict failure of a given tunnel in the SD-WAN.
A tangible, non-transitory, computer-readable medium storing program instructions that cause a device to execute a process comprising:
- identifying, by a head-end router of a tunnel in a software-defined wide area network (SD-WAN), one or more telemetry data variables for use to predict failure of the tunnel, wherein the one or more telemetry data variables comprise one or more of:
- a queue statistic, memory usage, or processor load of a tail-end router of the tunnel;
- sending, by the head-end router, a Bidirectional Forwarding Detection (BFD)-based telemetry request towards the tail-end router;
- receiving, at the head-end router, the requested one or more telemetry data variables;
- and
- using, by the head-end router, the received one or more telemetry data variables as input to a machine learning-based model executing locally on the head-end router, to predict a failure of the tunnel, wherein the machine learning-based model is trained to predict failure of a given tunnel in the SD-WAN.
Dependencies are inferred from the claim text.