Kuijia: Traffic rescaling in software-defined data center WANs

Abstract

Network faults like link or switch failures can cause heavy congestion and packet loss. Trafc engineering systems need a lot of time to detect and react to such faults, which results in signifcant recovery times. Recent work either preinstalls a lot of backup paths in the switches to ensure fast rerouting or proactively prereserves bandwidth to achieve fault resiliency. Our idea agilely reacts to failures in the data plane while eliminating the preinstallation of backup paths. We propose Kuijia, a robust trafc engineering system for data center WANs, which relies on a novel failover mechanism in the data plane called rate rescaling. Te victim fows on failed tunnels are rescaled to the remaining tunnels and enter lower priority queues to avoid performance impairment of aboriginal fows. Real system experiments show that Kuijia is efective in handling network faults and signifcantly outperforms the conventional rescaling method.