Embodiments of the invention relate to methods and systems for processing network traffic data units. The methods and systems may include functionality for generating a first network traffic data unit comprising an indication that the first network traffic unit should be delayed for a time before re-transmission to a client device; and transmitting the first network traffic data unit to a first edge device.

Embodiments of the invention relate to methods and systems for processing network traffic data units. The methods and systems may include functionality for generating a first network traffic data unit comprising an indication that the first network traffic unit should be delayed for a time before re-transmission to a client device; and transmitting the first network traffic data unit to a first edge device.

A clock determining method includes: when both a second network device and a first network device are synchronous with a reference clock, simulating, by using delay information between the second network device and the first network device and clock frequency information of the second network device, a second virtual clock synchronized with a first virtual clock, where the first virtual clock is used to simulate a clock of the first network device. A clock of the second network device can thus be simulated to perform a subsequent operation by using the simulated clock. For example, the simulated clock may be used to estimate precision time protocol (PTP) message synchronization performance of the second network device. Therefore, the PTP message synchronization performance of the second network device may be pre-determined before a global navigation satellite system (GNSS) fails, to guide network operation and maintenance activities.

Systems, devices, and methods of measuring directional latency and congestion in a communication network. A Uni-Directional Latency Determination Unit is connected in a communication network, located between an end-user device and a server. It monitors packets transported between the end-user device and the server, and it estimates a uni-directional latency of packet transport from the end-user device to the server or from the server to the end-user device. It utilizes a Transmission Control Protocol (TCP) Header and Timestamp Analyzer, to perform an analysis of data contained in timestamps of TCP packet headers of transported packets; and particularly, it analyzes data contained in a TSval field of such TCP packet headers. Congestion mitigation operations are accordingly deployed or activated.

A method performed by a central server node in a distributed machine learning environment is provided. The method includes: managing distributed machine learning for a plurality of local client nodes, such that a first set of the plurality of local client nodes are assigned to assist training of a first central model and a second set of the plurality of local client nodes are assigned to assist training of a second central model; obtaining information regarding network conditions for the plurality of local client nodes; clustering the plurality of local client nodes into one or more clusters based at least in part on the information regarding network conditions; re-assigning a local client node in the first set to the second set based on the clustering; and sending to the local client node a message including model weights for the second central model.

This application provides a network performance monitoring method, a network device, and a storage medium, and belongs to the field of network technologies. In this application, a forwarding plane samples network performance data based on a fine-grained time periodicity, and records a quantity of network performance exceptions; and a control plane generates, based on a coarse-grained time periodicity, an alarm when the quantity of network performance exceptions recorded by the forwarding plane is greater than a threshold. On a basis of meeting a fine-grained requirement on network performance monitoring, because the control plane does not need to report all the collected network performance data, a volume of data that needs to be reported by the control plane is greatly reduced. This resolves a problem of overload of a main control CPU that is caused by massive data reporting, and reduces dependency of the network performance monitoring on performance of the main control CPU of a device. This further resolves a problem that a large quantity of bandwidth resources are occupied due to the massive data reporting, reduces dependency of the network performance monitoring on the bandwidth resources, and helps meet a requirement for deploying a large quantity of performance monitoring nodes in a live network.

This application provides a network performance monitoring method, a network device, and a storage medium, and belongs to the field of network technologies. In this application, a forwarding plane samples network performance data based on a fine-grained time periodicity, and records a quantity of network performance exceptions; and a control plane generates, based on a coarse-grained time periodicity, an alarm when the quantity of network performance exceptions recorded by the forwarding plane is greater than a threshold. On a basis of meeting a fine-grained requirement on network performance monitoring, because the control plane does not need to report all the collected network performance data, a volume of data that needs to be reported by the control plane is greatly reduced. This resolves a problem of overload of a main control CPU that is caused by massive data reporting, and reduces dependency of the network performance monitoring on performance of the main control CPU of a device. This further resolves a problem that a large quantity of bandwidth resources are occupied due to the massive data reporting, reduces dependency of the network performance monitoring on the bandwidth resources, and helps meet a requirement for deploying a large quantity of performance monitoring nodes in a live network.

Methods, apparatus, systems, and articles of manufacture to monitor mobile Internet usage are disclosed. Example apparatus disclosed herein to monitor application usage associated with a device means for differentiating between foreground requests and background requests included in a burst of logged requests. Disclosed example apparatus further include means for calculating a presentation duration for a first application executed by the device based on at least a subset of the foreground requests identified in the burst of logged requests.

Methods, apparatus, systems, and articles of manufacture to monitor mobile Internet usage are disclosed. Example apparatus disclosed herein to monitor application usage associated with a device means for differentiating between foreground requests and background requests included in a burst of logged requests. Disclosed example apparatus further include means for calculating a presentation duration for a first application executed by the device based on at least a subset of the foreground requests identified in the burst of logged requests.

A network device includes processing circuitry and a plurality of ports. The ports connect to a communication network. The processing circuitry is configured to receive, via an input port, data packets and probe packets that are addressed to a common output port, to store the data packets in a first queue and the probe packets in a second queue, both the first queue and the second queue are served by the output port, to produce telemetry data indicative of a state of the network device, based on a processing path that the data packets traverse within the network device, to schedule transmission of the data packets from the first queue at a first priority, and schedule transmission of the probe packets from the second queue at a second priority higher than the first priority, and to modify the scheduled probe packets so as to carry the telemetry data.