A network device includes a memory and a memory management circuit. The memory is to store a shared buffer. The memory management circuit is to estimate respective bandwidth measures for one or more queues used in processing packets in the network device, and to allocate and deallocate segments of the shared buffer to at least one of the queues based on the bandwidth measures.

A network device may receive packets and may calculate, during a time interval, an arrival rate and a departure rate, of the packets, at one of multiple virtual output queues. The network device may calculate a current oversubscription factor based on the arrival rate and the departure rate, and may calculate a target oversubscription factor based on an average of previous oversubscription factors associated with the multiple virtual output queues. The network device may determine whether a difference exists between the target oversubscription factor and the current oversubscription factor and may calculate, when the difference exists, a scale factor based on the current oversubscription factor and the target oversubscription factor. The network device may calculate new scheduling weights based on prior scheduling weights and the scale factor, and may process packets received by the multiple virtual output queues based on the new scheduling weights.

A network device may receive packets and may calculate, during a time interval, an arrival rate and a departure rate, of the packets, at one of multiple virtual output queues. The network device may calculate a current oversubscription factor based on the arrival rate and the departure rate, and may calculate a target oversubscription factor based on an average of previous oversubscription factors associated with the multiple virtual output queues. The network device may determine whether a difference exists between the target oversubscription factor and the current oversubscription factor and may calculate, when the difference exists, a scale factor based on the current oversubscription factor and the target oversubscription factor. The network device may calculate new scheduling weights based on prior scheduling weights and the scale factor, and may process packets received by the multiple virtual output queues based on the new scheduling weights.

Scheduling traffic of a communication session between an application on WiFi and another device, by: receiving traffic of a first session; determining that the traffic belongs to a first classification; determining that a time allocated to the first classification times a second classification airtime is less than or equal to a time allocated to the second classification times a first classification airtime; selecting a counter associated with the first session as being the largest of multiple counters each associated with a different communication session of multiple communication sessions (which include the first session); determining that the traffic of the first session is to be scheduled for transmission to the application over WiFi in response to the counter being determined to be the largest of the plurality of counters; and based on the traffic being determined to be scheduled, increasing the airtime associated with the first classification and decreases the counter.

A method for dynamically allocating buffer credits between a system and a storage area network (SAN). The method includes one or more computer processors determining a forecast of a change related to a pattern of network traffic that originates from a computing system that links to a storage area network (SAN) via a network connection. The method further includes determining whether the forecast change related to the pattern of network traffic dictates a change to a buffer credit allocation associated with the network connection. The method further includes responding to determining that the forecast change related to the pattern of network traffic dictates the buffer credit allocation change by determining a value for the buffer credit allocation associated with the change. The method further includes transmitting a request to a switch of the SAN to modify a buffer credit allocation value corresponding to a port of the switch linked to the network connection.

Scheduling traffic of a communication session between an application on WiFi and another device, by: receiving traffic of a first session; determining that the traffic belongs to a first classification; determining that a time allocated to the first classification times a second classification airtime is less than or equal to a time allocated to the second classification times a first classification airtime; selecting a counter associated with the first session as being the largest of multiple counters each associated with a different communication session of multiple communication sessions (which include the first session); determining that the traffic of the first session is to be scheduled for transmission to the application over WiFi in response to the counter being determined to be the largest of the plurality of counters; and based on the traffic being determined to be scheduled, increasing the airtime associated with the first classification and decreases the counter.

Core network slices that belong to a given operator community are efficiently tracked at the network control/user plane functions level, with rich data analytics in real-time based on their geographic instantiations. In one aspect, an enhanced vendor agnostic orchestration mechanism is utilized to connect a unified management layer with an integrated slice-components data analytics engine (SDAE), a slice performance engine (SPE), and a network slice selection function (NSSF) in a closed-loop feedback system with the serving network functions of one or more core network slices. The tight-knit orchestration mechanism provides economies of scale to mobile carriers in optimal deployment and utilization of their critical core network resources while serving their customers with superior quality.

Core network slices that belong to a given operator community are efficiently tracked at the network control/user plane functions level, with rich data analytics in real-time based on their geographic instantiations. In one aspect, an enhanced vendor agnostic orchestration mechanism is utilized to connect a unified management layer with an integrated slice-components data analytics engine (SDAE), a slice performance engine (SPE), and a network slice selection function (NSSF) in a closed-loop feedback system with the serving network functions of one or more core network slices. The tight-knit orchestration mechanism provides economies of scale to mobile carriers in optimal deployment and utilization of their critical core network resources while serving their customers with superior quality.

Embodiments of the present disclosure relates to communication technology, and in particular to a method and an apparatus for scheduling traffic of a node, an electronic device and a storage medium. The method for scheduling traffic of a node comprises: obtaining a target node and each visitor of the target node, where traffic of the target node is to be scheduled out; obtaining a corresponding node accessed by each visitor when using a server address with a different path attribute, where each corresponding node pre-broadcasts a server address with at least two different path attributes for the visitor to access; obtaining a scheduled-in node from the corresponding node of each visitor; and scheduling the traffic of the target node to the scheduled-in node.

Systems, apparatuses, and methods for implementing a configurable packet arbiter with minimum progress guarantees are described. An arbiter includes at least control logic, a plurality of counters, and a tunables matrix. The tunables matrix stores values for a plurality of configurable parameters for the various transaction sources of the arbiter. These parameter values determine the settings that the arbiter uses for performing arbitration. One of the parameters is a minimum progress guarantee value that specifies how many times each source should be picked per interval. The minimum progress guarantee helps to reduce arbitration-related jitter. Also, the arbiter includes a grant counter for each source. After the minimum progress guarantees are satisfied, the arbiter selects the source with the lowest grant counter among the sources with packets eligible for arbitration. Then, the arbiter increments the grant counter of the winning source by a grant increment amount specific to the source.