Virtual local area network (VLAN)-based transparent interconnection of lots of links (TRILL) traffic priority scheduling method

Abstract

The invention discloses a virtual local area network (VLAN)-based transparent interconnection of lots of links (TRILL) traffic priority scheduling method. The topology of a TRILL network is cognized by a link state protocol, RBs determine VLAN-based TRILL traffic processing priorities among the RBs by negotiation according to extended field information in a received handshake protocol message, different service is provided for different users by a shared link on the basis of not extending hardware resources to maximally protect and utilize the conventional resources, different links are not required to be provided for different clients, and the service corresponding to corresponding clients is realized on each link to effectively ensure the quality of service of TRILL traffic.

Description

technical field [0001] The present invention relates to the communication field, in particular to a VLAN (Virtual Local Area Network)-based TRILL (Transparent Interconnection of Lots of Links) flow priority scheduling method. Background technique [0002] Most of today's data center networks are built following the standard hierarchical concept, which is divided into access layer and aggregation / core layer. , STP (spanning tree protocol) is run on all Layer 2 links. When more than one path is reachable between any two points, STP will block the redundant path to ensure that only one path is reachable between two points. , so as to prevent the occurrence of loops. This mode has been adopted on a large scale for a long time in the past, because it is very simple to deploy, access layer devices do not require complex configuration, and most network policies can be distributed to the entire network as long as they are deployed centrally at the aggregation layer. But as the siz...

AI Technical Summary

Problems solved by technology

[0003] In the future, the horizontal traffic inside the data center will become larger and larger, and STP still needs to be run between the newly added device and the original device. If only one link between the two servers is feasible, the rest of the 10 Gigabit switch ports will be blocked. Not only is it a great waste of investment, but it also cannot support the rapid expansion of services; secondly, when the number of cross-links increases, the design of the Layer 2 network will become very complicated, which link should be reserved and which link should be blocked , Where is the layer-3 gateway located? There will be a lot of problems like this, which will lose the advantage of simple layer-2 network configuration; finally, the traditional layer-2 MAC address has no hierarchical concept, which can easily lead to edge The MAC address space of the device is exhausted, especially in a virtualized data center, where the number of MAC addresses of virtual machines may be in the thousands

Images