Performance evaluation of OpenDaylight SDN controller

Enourmous amount of data has resulted into large data centres. Virtual machines and virtual networks are an integral part of large data centres. As a result, software defined network controllers have emerged as viable solution to manage such networks. The performance analysis of network controllers...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Proceedings - International Conference on Parallel and Distributed Systems s. 671 - 676
Hlavní autori: Khattak, Zuhran Khan, Awais, Muhammad, Iqbal, Adnan
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 01.12.2014
Predmet:
Benchmark testing
Benchmarking
Cbench
Java
OpenDaylight
OpenFlow
Protocols
Software
Software Defined Networking
Switches
Throughput
ISSN:1521-9097
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Enourmous amount of data has resulted into large data centres. Virtual machines and virtual networks are an integral part of large data centres. As a result, software defined network controllers have emerged as viable solution to manage such networks. The performance analysis of network controllers is generally carried out through benchmarking. Although several benchmarking studies exist, recently launched OpenDaylight SDN Controller is not considered in any benchmarking study yet. In this paper, we present initial results of benchmarking of Open-Daylight SDN Controller with well known Floodlight controller. We present latency and throughput results of OpenDaylight SDN Controller and Floodlight under different scenarios. We find that OpenDaylight SDN Controller has low average responses as compared to Floodlight. We also note that the standard benchmarking tool - Cbench - has no support for real traffic patterns in a data centre, since data centre traffic is considerably complex. In addition to benchmarking of OpenDaylight SDN Controller, we propose modifications in Cbench to accommodate models of real traffic in data centres. We also discuss our initial implementation.
ISSN:1521-9097
DOI:10.1109/PADSW.2014.7097868