Ca, Inc.

David Sanchez Charles

Victor Muntés-Mulero

Marc Sole Simo

Jaume Ferrarons Llagostera

Victor Muntes Mulero

Alberto Huelamo Segura

David Solans Noguero

Erhan Giral

Smrati Gupta

Marta Arias

Giorgio Stampa

An autonomous controller for SDN, virtual, and/or physical networks can be used to optimize a network automatically and determine new optimizations as a network scales. The controller trains models that can determine in real-time the optimal path for the flow of data from node A to B in an arbitrary network. The controller processes a network topology to determine relative importance of nodes in the network. The controller reduces a search space for a machine learning model by selecting pivotal nodes based on the determined relative importance. When a demand to transfer traffic between two hosts is detected, the controller utilizes an AI model to determine one or more of the pivotal nodes to be used in routing the traffic between the two hosts. The controller determines a path between the two hosts which comprises the selected pivotal nodes and deploys a routing configuration for the path to the network.

Efficient machine learning for network optimization

A multivariate clustering-based anomaly detector can generate an event for consumption by an APM manager that indicates detection of an anomaly based on multivariate clustering analysis after topology-based feature selection. The anomaly detector accumulates time-series data across a series of time instants to form a multivariate time-series data slice or multivariate data slice. The anomaly detector then performs multivariate clustering analysis with the multivariate data slice. The anomaly detector determines whether a multivariate data slice is within a cluster of multivariate data slices. If the multivariate data slice is within the cluster and the cluster is a known anomaly cluster, then the anomaly detector generates an anomaly detection event indicating detection of the known anomaly. The anomaly detector can also determine that a multivariate data slice is within an unknown cluster and generate an event indicating detection of an unknown anomaly.

Multivariate clustering-based anomaly detection

A multivariate path-based anomaly detection and prediction service ('anomaly detector') can generate a prediction event for consumption by the APM manager that indicates a likelihood of an anomaly occurring based on path analysis of multivariate values after topology-based feature selection. To predict that a set of metrics will travel to a cluster that represents anomalous application behavior, the anomaly detector analyzes a set of multivariate date slices that are not within a cluster to determine whether dimensionally reduced representations of the set of multivariate data slices fit a path as described by a function.

Multivariate path-based anomaly prediction

In a datacenter setting, a summary of differences and similarities between two or more states of the same or similar systems are predicted. Initially, a Long Short-Term Memory (LSTM) neural network is trained with to predict a summary describing the state change between at least two states of the datacenter. Given a set of training data (at least two datacenter states that are annotated with a state change description), the LSTM neural network learns which similarities and differences between the datacenter states correspond to the annotations. Accordingly, given a set of test data comprising at least two states of a datacenter represented by context graphs that indicate a plurality of relationships among a plurality of nodes corresponding to components of a datacenter, the LSTM neural network is able to determine a state change description that summarizes the differences and similarities between the at least two states of the datacenter.

Descriptive datacenter state comparison

In a datacenter setting, annotations or descriptions of relevant parts or subgraphs corresponding to components in the datacenter are predicted. Given a set of training data (library of subgraphs seen in the past labeled with a textual description explaining why were they considered relevant enough to be placed in the historical database), the recurrent neural network (RNN) learns how to combine the different textual annotations coming from each relevant region into a single annotation that describes the whole system. Accordingly, given a set of input or test data (datacenter state modeled a context graph that is not annotated), the system determines which regions of the input graph are more relevant, and for each of these regions, the RNN predicts an annotation even in a previously unseen or different datacenter infrastructure.

Automatic datacenter state summarization

Incomplete state information for nodes of a datacenter is completed utilizing historical state information. A context graph is received having a plurality of nodes that correspond to components of the datacenter, each node including properties that correspond to the represented component. It is determined that at least one of the properties for a node is incomplete. A context hash is derived from the context graph and compared to a plurality of subgraph hashes. Based upon the comparison, one or more subgraphs are identified that are similar to the context hash. Where a single similar subgraph is identified, the property information for the single subgraph is imputed to the context graph to complete the incomplete information. Where more than one similar subgraph is identified, the property information for the identified similar subgraphs is aggregated and the aggregated property information is imputed to the context graph to complete the incomplete information.

Growing community of inventors

Barcelona, Spain

David Sanchez Charles