Skip to Main Content
User-perceived performance continues to be the most important QoS indicator in cloud-based data centers today. Effective allocation of virtual machines (VMs) to handle both CPU intensive and I/O intensive workloads is a critical management capability in virtualized clouds. Although a fair amount of research have dedicated to measuring and scheduling jobs among VMs, there still lacks of in-depth understanding of interference factors that impact the efficiency and effectiveness of resource multiplexing and scheduling. In this paper, we present experimental study on performance interference in parallel processing of CPU-bound and network-bound workloads on Xen Virtual Machine Monitor (VMM). Based on our study, we conclude five key findings that are critical for cloud service providers and consumers. First, co-locating network-intensive workloads in isolated VMs incurs high overheads for extensive context switches and events in Dom0 and VMM. Second, co-locating CPU-intensive workloads in isolated VMs incurs CPU contention due to fast I/O processing in I/O channel. Third, running CPU-intensive and network-intensive workloads in conjunction delivers higher aggregate performance. Fourth, performance of network-intensive workload is insensitive to CPU assignment among VMs, whereas adaptive CPU assignment is critical to CPU-intensive workload. Last, limitation on grant table is a potential bottleneck in the current Xen system.