Data center network multi-path dynamic load balancing method based on SDN

Abstract

A data center network multi-path dynamic load balancing method based on an SDN comprises the following steps: monitoring a network state, and obtaining real-time state information of equipment; dividing large and small streams; for the small flow, carrying out routing in an ECMP mode, and carrying out hash operation based on the 10-tuple value of the flow to select a forwarding path; for the large flow, adopting a large flow routing algorithm; calculating the average link utilization rate and the link load variance, and re-routing the flow occupying the maximum bandwidth on the link with the maximum link utilization rate to the path with the maximum bottleneck bandwidth. According to the method, an initial routing algorithm and a rerouting algorithm are combined, and the load balancing effect is enhanced; the low time delay requirement of the small flow is met, and the high throughput requirement of the large flow is met; a dynamic routing algorithm based on a probability selection algorithm is adopted for the large flow, and the problem that the large flow is distributed to the same path due to the fact that path information is not updated in time is avoided; the upper limit of the number of times of rerouting in a single period is set, unnecessary rerouting caused by network fluctuation is avoided, and the overhead of a controller is reduced.

Description

technical field [0001] The invention belongs to the field of data center networks, in particular to an SDN-based multi-path dynamic load balancing method for data center networks. Background technique [0002] With the rapid development of Internet technology and the rapid growth of network traffic, network operators continue to increase the deployment density of servers and storage devices, data center network nodes and links are growing exponentially, and data centers have gradually become a gathering place for network traffic. The continuous increase of data traffic in the data center and the different requirements of different types of traffic on the link and the quality of service have put forward higher requirements on the data center network, and most of the existing routing algorithms do not comprehensively consider the real-time status of the link and Various traffic characteristics, thus causing some links in the network to be overloaded while other links are still...

AI Technical Summary

Problems solved by technology

However, the DLB algorithm used in this patent for the new flow has the characteristics of selecting the next hop with a greedy strategy. The urgency of scheduling large flows on the network can alleviate congestion to a certain extent, but the unrestricted DR algorithm may cause a lot of additional overhead; at the same time, the demand for low latency of small flows is not considered

Images