DocumentCode :
154185
Title :
NEO: A Nonblocking Hybrid Switch Architecture for Large Scale Data Centers
Author :
Zhemin Zhang ; Yuanyuan Yang
Author_Institution :
Dept. of Electr. & Comput. Eng., Stony Brook Univ., Stony Brook, NY, USA
fYear :
2014
fDate :
9-12 Sept. 2014
Firstpage :
510
Lastpage :
519
Abstract :
As the scale of data centers and cloud computing applications increases, data center networks play a critical role in meeting the huge communication bandwidth requirement of such applications. The scalability of conventional electronic data center networks is limited by wiring complexity and reaching distance of links under fixed power budget. To overcome this problem, in this paper we propose a nonblocking hybrid switch architecture, called NEO (Nonblocking Electronic and Optical), which is able to provide nonblocking interconnections for as many as 1,000,000 servers in a data center. NEO maintains electronic interconnections for intra-pod networks, while providing interpod interconnections by optical core switches, which not only increases the scalability of the switch architecture, but also lowers the switch cost and power consumption compared to other existing optical switch architectures. We also design a packet scheduler for NEO, which adopts a credit flow control mechanism and a parallel scheduling algorithm to avoid packet loss, and provide low communication latency. Our simulation results demonstrate that NEO achieves very low average packet delay compared to other existing optical switching architectures under various traffic patterns.
Keywords :
bandwidth allocation; computer centres; packet switching; scheduling; telecommunication switching; NEO; communication bandwidth requirement; credit flow control mechanism; data centers; nonblocking electronic and optical; nonblocking hybrid switch architecture; packet scheduler; parallel scheduling algorithm; Optical buffering; Optical interconnections; Optical switches; Ports (Computers); Scheduling algorithms; Servers; Arrayed waveguide switch router; Data center networks; Hybrid switch architecture; Optical core switch; Parallel scheduling algorithm; Tunable wavelength converter; WDM;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel Processing (ICPP), 2014 43rd International Conference on
Conference_Location :
Minneapolis MN
ISSN :
0190-3918
Type :
conf
DOI :
10.1109/ICPP.2014.60
Filename :
6957260
Link To Document :
بازگشت