Abstract

The control packets in the data center networks (DCNs) have to contest with the data packets although they are usually much shorter in size and much more important in network management. Moreover, the uneven distribution of the packets may create potential traffic hotspots in the DCN which could degrade network performance drastically. To bridge these gaps, a layout-independent constructing algorithm and a scheduling method are put forward towards layout-independent wireless facility in data centers. First of all, a conflict aware spanning tree algorithm is developed to construct the wireless facility network (WFN). Secondly, a scheduling method which contains three steps, route calculation, traffic estimation, and flow scheduling, is presented. In the route calculation step, a route set between each node pair is calculated in advance for later usage. The scheduler estimates the traffic loads on the links on a regular basis in the traffic estimation step. Then, arrived data and control flows are scheduled according to multiple policies based on given route sets and scheduling objectives in the flow scheduling step. Finally, a series of experiments have been conducted on NS3 based on two typical data center layouts. Experimental results in both scenarios have validated our proposal’ effectiveness.

Highlights

  • With the rapid development of modern cloud-based network services, more and more data centers are built around the world since the data centers (DCs) are the best infrastructure to support these applications

  • To evaluate the performance of the wireless facility network (WFN) construction and its scheduling methods proposed in this paper, a series of simulation experiments are conducted based on extending the Flyways simulator developed on NS3 [7, 30]

  • To fulfill our simulation requirements, we have extended the simulator by WFN construction, routes calculation, and flow scheduling

Read more

Summary

Introduction

With the rapid development of modern cloud-based network services, more and more data centers are built around the world since the data centers (DCs) are the best infrastructure to support these applications. Traditional tree based data center network (DCN) structure faces a lot of challenges to carry out so huge network traffic, such as structure design, network management, and load balancing. Many novel or revolutionized network structures, such as FatTree [4] and Jellyfish [5], have been presented to replace or enhance the traditional tree based DCNs. it is hard to implement these architectures in practice since they have introduced too much overheads like cabling or multiports servers.

Objectives
Results
Conclusion
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call