Multicast Optimization And Recovery In Multihoming Environment

Reliability of multicasting is increasingly becoming an important issue as the number of end users continues to grow, their demand for reliable service increases. This thesis proposes a novel algorithm for creating a recovery model while optimizing both inter and intra domain bandwidth. This is achieved by creating a centralized rendezvous point within the intra domain topology. The rendezvous point will create a static multicast tree and it will avoid link congestion during inter-domain link failure. This algorithm also reduces link congestion surrounding the border routers. This is achieved by shifting the root of the multicast tree from the border router to the rendezvous point. This rendezvous point is then selected based on an optimization algorithm to reduce bandwidth congestion. A Steiner tree was used to optimize the intra domain links. The simulation results indicate up to 30% increase over conventional optimization algorithms which do not consider a rendezvous point model.

Multicast Optimization And Recovery In Multihoming Environment

Reliability of multicasting is increasingly becoming an important issue as the number of end users continues to grow, their demand for reliable service increases. This thesis proposes a novel algorithm for creating a recovery model while optimizing both inter and intra domain bandwidth. This is achieved by creating a centralized rendezvous point within the intra domain topology. The rendezvous point will create a static multicast tree and it will avoid link congestion during inter-domain link failure. This algorithm also reduces link congestion surrounding the border routers. This is achieved by shifting the root of the multicast tree from the border router to the rendezvous point. This rendezvous point is then selected based on an optimization algorithm to reduce bandwidth congestion. A Steiner tree was used to optimize the intra domain links. The simulation results indicate up to 30% increase over conventional optimization algorithms which do not consider a rendezvous point model.

Surviving switch failures in cloud datacenters

Switch failures can hamper access to client services, cause link congestion and blackhole network traffic. In this study, we examine the nature of switch failures in the datacenters of a large commercial cloud provider through the lens of survival theory. We study a cohort of over 180,000 switches with a variety of hardware and software configurations and find that datacenter switches have a 98% likelihood of functioning uninterrupted for over 3 months since deployment in production. However, there is significant heterogeneity in switch survival rates with respect to their hardware and software: the switches of one vendor are twice as likely to fail compared to the others. We attribute the majority of switch failures to hardware impairments and unplanned power losses. We find that the in-house switch operating system, SONiC, boosts the survival likelihood of switches in datacenters by 1% by eliminating switch failures caused by software bugs in vendor switch OSes.