Method supporting hierarchical queue scheduling and flow shaping and apparatus thereof

Abstract

The invention discloses a method supporting hierarchical queue scheduling and flow shaping. The method comprises the following steps that: a message is received and an internal priority is mapped for the message; the message is sent into a buffer queue and a buffer queue WRED algorithm is utilized to carry out distinguishment on a responsive flow and a non-responsive flow; with regard to the responsive flow and the non-responsive flow, different drop thresholds are configured, so that an object that the non-responsive flow is dropped in advanced and the responsive flow is protected on the network congestion condition is achieved; the message is sent into the buffer queue and flow shaping is carried out by a three-layer flow shaper; and after service group scheduling and port level scheduling are carried out on the message in the buffer queue, the message is sent out from the queue; and editing is carried out on the message and the edited message is sent out from a destination port. According to the realization method for hierarchical flow shaping and queue scheduling in the invention, the structure is flexible and the hardware cost is low; besides, an access-gathering-core hierarchical network topology is reflected really and an actual demand on H-QoS deployment by the network is met.

Description

technical field [0001] The invention relates to network communication technology, in particular to a method and a device for queue scheduling and traffic shaping in a traffic management engine in a network switching device. Background technique [0002] The switch chip is divided into an uplink packet processing engine, a traffic management engine and a downlink packet processing engine. In the upstream direction, the processing engine of the message maps the internal priority of the switching chip according to the external priority field (such as 802.1p, DSCP, etc.) carried in the message header, and searches the forwarding table to obtain the destination port number of the next hop. After receiving the packet, the traffic management engine calculates the buffer queue index number according to the destination port number and internal priority. Packets need to pass through the queue drop management algorithm before they are allowed to enter the cache queue and wait for sche...

AI Technical Summary

Problems solved by technology

The shortcomings of the network switching chip in the prior art are: (1) it will produce chaotic bandwidth preemption results, the bandwidth of a certain user may be preempted by high priority services of other users, and the overall bandwidth of each user cannot be guaranteed; (2) ) cannot truly reflect the access-aggregation-core three-layer network topology in IP networking

The switching equipment at the aggregation layer requires the centralized configuration of H-QoS functions, and maps the various communication equipment and service types connected to the nodes at different levels of the aggregation point for hierarchical QoS management; (3) User traffic shapers and service queues There is no linkage mechanism between them, and the unused bandwidth of a service queue cannot be used by other service queues

Images