Traffic load balancing method and device

Abstract

The invention provides a traffic load balancing method and a traffic load balancing device. The method comprises that: stacking equipment in a convergence layer monitors the traffic of an aggregation link between the stacking equipment and a core layer; the stacking equipment finds a data stream causing the unbalance of the traffic from the data streams transmitted by the aggregation link when monitoring the traffic of the aggregation link is unbalanced; and the stacking equipment determines an input interface from which the found data streams enters the stacking equipment, reallocates a new medium access control (MAC) address to the input interface, and transmits the new MAC address allocated to the input interface to a terminal server, so that the terminal server transmits the data stream by utilizing the new MAC address as a destination MAC address to convert the traffic unbalance into traffic balance. By the method and the device, the load balance among member links in the aggregation link between the convergence layer and the core layer can be realized.

Description

technical field [0001] The invention relates to communication technology, in particular to a flow load balancing method and device. Background technique [0002] With the increasing dependence of enterprises on information access, the impact of data centers on the daily business of enterprises is also increasing. Once the enterprise's data center fails, it will have a great impact on the normal operation of the enterprise's daily business and bring huge losses to the enterprise. In general, when the data center fails, the loss of the enterprise is divided into the following aspects: the impact on the daily work of the enterprise such as the inability of employees to work normally, direct property loss, such as loss of orders, compensation for loss of corporate partners, etc., corporate reputation Losses such as losing some customers. Therefore, reliability has become an important aspect to measure the quality of a data center. [0003] At present, in order to improve the ...

AI Technical Summary

Problems solved by technology

[0014] However, because different terminal servers and different business applications correspond to different data flows, and even the flow of different data flows is very different, which leads to unbalanced flow on each member link in the above-mentioned aggregated link.

When the unbalanced traffic is transmitted to the stacking device in the aggregation layer, because the member devices of the stacking device have the above-mentioned local forwarding priority characteristics, that is, when the member devices of the stacking device in the aggregation layer receive data flows, they preferentially pass through the aggregation link The member link corresponding to the outbound interface of the stack device is forwarded, which will cause the aggregated link between the stack device and the core layer, such as image 3 The traffic of the member links in the aggregated link L11 shown is unbalanced. For example, there is an Figure 4 A member link shown is link L between device C and device A CA Packet loss due to full load and overload, other member links such as link L between device D and device B DB In the case of light load, the load balancing of aggregated links from the aggregation layer to the core layer becomes empty talk, which seriously affects the quality of user service and reduces the forwarding performance of the entire system

Images