Cisco Technology, Inc.

Abhishek Kumar

Jean-Philippe Vasseur

Grégory Mermoud

Javier Cruz Mota

Erwan Barry Tarik Zerhouni

Moiz Raja

Harmanpreet Singh

Kamalkrishna Rameshan

Jerome Henry

Robert E Barton

Vinay Kumar Kolar

Santosh Ghanshyam Pandey

Vikram Kumaran

The present technology pertains to a system, method, and non-transitory computer-readable medium for evaluating the impact of network changes. The technology can detect a temporal event, wherein the temporal event is associated with a change in a network configuration, implementation, or utilization; define a first period prior to the temporal event and a second period posterior to the temporal event; and compare network data collected in the first period and network data collected in the second period.

Analyzing the impact of network events across time

The disclosed technology relates to a process of providing dynamic machine learning on premise model selection. In particular, a set of machine learned models are generated and provided to an on premise computing device. The machine learned models are generated using a cluster of customer data (e.g. telemetric data) stored on a computing network having different ranges of computational complexity. One of the machine learned models from the set of machine learned models will be selected based on the current available computational resources detected at the on premise computing device. Different machine learned models from the set of machine learned models can then be selected based on changes in the available computational resources and/or customer feedback.

Dynamic machine learning on premise model selection based on entity clustering and feedback

The disclosed technology relates to a process for metered training of fog nodes within the fog layer. The metered training allows the fog nodes to be continually trained within the fog layer without the need for the cloud. Furthermore, the metered training allows the fog node to operate normally as the training is performed only when spare resources are available at the fog node. The disclosed technology also relates to a process of sharing better trained machine learning models of a fog node with other similar fog nodes thereby speeding up the training process for other fog nodes within the fog layer.

Distributed learning model for fog computing

The present technology pertains to a system, method, and non-transitory computer-readable medium for evaluating the impact of network changes. The technology can detect a temporal event, wherein the temporal event is associated with a change in a network configuration, implementation, or utilization. The technology defines, based on a nature of the temporal event, a first period prior to the temporal event or a second period posterior to the temporal event. The technology compares network data collected in the first period and network data collected in the second period.

System and method of comparing time periods before and after a network temporal event

Technologies for dynamically generating topology and location based network insights are provided. In some examples, a method can include determining statistical changes in time series data including a series of data points associated with one or more conditions or parameters of a network; determining a period of time corresponding to one or more of the statistical changes in the time series data; obtaining telemetry data corresponding to a segment of the network and one or more time intervals, wherein a respective length of each time interval is based on a length of the period of time corresponding to the one or more of the statistical changes in the time series data; and generating, based on the telemetry data, insights about the segment of the network, the insights identifying a trend or statistical deviation in a behavior of the segment of the network during the one or more time intervals.

Technologies for dynamically generating network topology-based and location-based insights

In one embodiment, a network assurance system that monitors a network labels time periods with positive labels, based on the network assurance system detecting problems in the network during the time periods. The network assurance system assigns tags to discrete portions of a feature space of measurements from the monitored network, based on whether a particular range of values in the feature space has a threshold probability of occurring during a positively-labeled time period. The network assurance system determines a set of the assigned tags that frequently co-occur with the positively-labeled time periods in which problems are detected in the network. The network assurance system causes performance of a mitigation action in the network based on the set of assigned tags that frequently co-occur with the positively-labeled time periods.

Deriving highly interpretable cognitive patterns for network assurance

Systems, methods, and computer-readable media for an adaptive gossip protocol. A node in a cluster can detect a gossip protocol synchronization triggering event which can include an indication that the node has received data from a second node via a gossip protocol, an update to data maintained by nodes in the cluster, and/or an operation. In response to the triggering event, the node can determine a dynamic gossip interval for disseminating data from the node to other nodes via a gossip protocol, the dynamic gossip interval being based on a synchronization state associated with the cluster and/or one or more gossip protocol events associated with the cluster. Next, the node can select a third node in the cluster for disseminating the data from the node to the third node. The node can then transmit the data to the third node via the gossip protocol based on the dynamic gossip interval.

Adaptive gossip protocol

In one embodiment, a service in a network samples application traffic throughputs for a set of applications present in a network. The service generates a throughput model based on the sampled application throughputs for the set of applications. The service performs anomaly detection on wireless throughput measurements from the network by comparing the wireless throughput measurements to the generated throughput model. The service sends an anomaly detection notification based on a determination that the wireless throughput measurements from the network are anomalous.

Wireless throughput issue detection using coarsely sampled application activity

In one embodiment, a network assurance system discretizes parameter values of a plurality of time series of measurements obtained from a monitored network by assigning tags to the parameter values. The network assurance system detects occurrences of a particular type of failure event in the monitored network. The network assurance system identifies a set of the assigned tags that frequently co-occur with the occurrences of the particular type of failure event. The network assurance system determines, using a Bayesian framework, rankings for the tags in the identified set based on how well each of the tags acts as a predictor of the failure event. The network assurance system initiates performance of a corrective measure for the failure event based in part on the determined rankings for the tags in the identified set.

Analyzing common traits in a network assurance system

In one embodiment, a network assurance service that monitors a network detects anomalies in the network by applying one or more machine learning models to telemetry data from the network. The network assurance service ranks feedback from a plurality of anomaly rankers regarding relevancy or criticality of the detected anomalies. The network assurance service clusters the plurality of anomaly rankers into clusters of similar rankers, based on the received ranking feedback. The network assurance service uses the clusters of similar rankers to assign reliability scores to each of the anomaly rankers. The network assurance service selects, based on the reliability scores, a subset of the plurality of anomaly rankers to receive an anomaly detection alert regarding a particular detected anomaly to be ranked. The network assurance service provides the anomaly detection alert to the selected subset of the plurality of anomaly rankers for ranking.

Eliminating bad rankers and dynamically recruiting rankers in a network assurance system

Disclosed are systems, methods, and computer-readable storage media for load balance resources amongst computing nodes. Various embodiments of the present technology can be used to, prior to assigning a first data shard included in a first cluster of data shards as a leader of the first cluster of data shards, determine whether a first computing node storing the first data shard also stores a second data shard designated as a leader of a second cluster of data shards, yielding a determination, and when the determination indicates that the first computing node stores the second data shard designated as the leader of the second cluster of data shards, designate the first data shard as non-electable to be the leader of the first cluster of data shards and electing an alternate data shard included in the first cluster of data shards as the leader of the first cluster of data shards.

Growing community of inventors

Vancouver, Canada

Abhishek Kumar