Juniper Networks, Inc.

Shivaprasad Gali

Sri Sampath Mallipudi

Chandrasekhar A

Patricio Giecco

Byeongseob Ki

Sanjeev Kumar Mishra

Hsiuyen Tsai

Sabyasachi Mukhopadhyay

Madhura Mysore Nagaprakash

Techniques are described for predicting future behavior of links in a network and generating dynamic thresholds for link metrics for use in path selection. In one example, a computing system receives historical values of a link metric for links of a network. The computing system executes a machine learning system which processes the historical values of the link metric to generate: (1) a predicted future value of the link metric for each link; and (2) a threshold for the link metric indicating whether the predicted future value for each link is anomalous. The computing system computes a path based on the predicted future values of the link metric and the threshold for the link metric. The computing system provisions the computed path, thereby enabling a network device to forward network traffic along the computed path.

Link behavior prediction for use in path selection

In general, techniques are described for retrieving operational command response text from network devices. A collector network device comprising an interface and a processor may be configured to perform the techniques. The interface may receive, via a messaging bus between the network management system and a webserver, a first command to request management data stored by a managed network device, and send, in response to the first command, a second command to direct the managed network device to output the management data. The interface may also receive, from the managed network device, the management data. The processor may generate, from the management data, a plurality of partial responses that each includes a portion of the management data, where the interface may next send, via the messaging bus and to the webserver, each of the plurality of partial responses as a separate message.

Retrieving operational command response text from network devices

Devices, systems, and techniques for global network state management are described herein. In one example, the present disclosure describes an example method performed by a software-defined network (SDN) controller that manages a network of a plurality of devices. The method include maintaining, by processing circuitry of the SDN controller, global network state information, configuring, by the processing circuitry, a notification service with a subscription for updates to a portion of the global network state information, wherein the notification service is configured to arrange the updates to the portion of the global network state information into a plurality of events, and distributing, by the notification service and via a plurality of interfaces, the plurality of events as event notifications to a plurality of subscriber micro-services of the SDN controller.

Global network state management

A first plurality of network configuration controllers of a controller may distribute, using a consistent hashing algorithm, a plurality of connection sessions with a plurality of network devices among the plurality of network configuration controllers. The controller may monitor a number of connection sessions maintained by each of the first plurality of network configuration controllers. The controller may add, based on monitoring the number of connection sessions maintained by each of the first plurality of network configuration controllers, an additional network configuration controller to the first plurality of network configuration controllers to form a second plurality of network configuration controllers. The second plurality of network configuration controllers may, in response to adding the additional network configuration controller to the first plurality of network configuration controllers, re-distribute, using the consistent hashing algorithm, one or more connection sessions of the plurality of connection sessions among the second plurality of network configuration controllers.

Network controller horizontal scaling for network device configuration sessions management

Techniques are described for in-service configuration data migration for distributed micro service-based applications. In one example, a network device comprises a plurality of legacy data repositories comprising configuration data in key-value pair format that specifies a plurality of parameters and corresponding values for operation of the network device, and a hierarchical configuration data model having a plurality of nodes arranged in a hierarchical organization having a root node and a plurality of leaf nodes. Each of the nodes of the configuration data model is configured to store a set of configuration data parameters for the network device. One or more of the nodes includes a plurality of external references to respective parameters of the plurality of parameters stored within the plurality of legacy data repositories. Process circuitry is configured to perform a migration of the configuration data from the legacy data repositories to the hierarchical data model.

Growing community of inventors

Bengaluru, India

Shivaprasad Gali