| Advances in virtualization technology have focused mainly on strengthening the isolation barrier between virtual machines (VMs) that are co-resident within a single physical machine. At the same time, a large category of communication intensive distributed applications and software components exist, such as web services, high performance grid applications, transaction processing, and graphics rendering, that often wish to communicate across this isolation barrier with other endpoints on co-resident VMs. This report presents our research that aims to optimize the communication performance between applications on co-located virtual machines. These efforts can be classified under two broad categories: (a) to improve inter-VM communication throughput of applications. (b) improvements to CPU scheduling algorithms at the hypervisor-level that address the latency requirements of inter-VM communication.;This dissertation improves both of the two aspects. XenLoop project introduces a shared-memory approach to bypass the traditional network communication data-path and improves communication throughput. AICT project and Group Scheduling algorithm optimize Xen scheduling algorithm for inter-VM communications while not sacrificing fairness. Evaluation results show that the new mechanisms are effective in improving inter-VM communication performance. |
