Intelligent and autonomous SDN applications need to monitor the network state in order to take appropriate actions. In this letter, we compare the impact of active and passive network state collection methods on an SDN load-balancing application running at the controller. We do this comparison through: 1) the results of a mathematical model evaluation we derive for the SDN load-balancer, and 2) the results of a series of elaborate experiments we ran on our emulation setup. The results show that in case of low-variation traffic, the load-balancer with passive state collection performed better than the active one, which was confirmed by both model and experimental evaluation. However, the load-balancer with the active state collection was more resilient to the nature of the traffic load.
[1]
Anja Feldmann,et al.
Logically centralized?: state distribution trade-offs in software defined networks
,
2012,
HotSDN '12.
[2]
Harsha V. Madhyastha,et al.
FlowSense: Monitoring Network Utilization with Zero Measurement Cost
,
2013,
PAM.
[3]
H. Jonathan Chao,et al.
Improving the performance of load balancing in software-defined networks through load variance-based synchronization
,
2014,
Comput. Networks.
[4]
Scott Shenker,et al.
CAP for networks
,
2013,
HotSDN '13.
[5]
Jessica K. Hodgins,et al.
Temporal notions of synchronization and consistency in Beehive
,
1997,
SPAA '97.