CloudScope: Diagnosing and Managing Performance Interference in Multi-tenant Clouds

Abstract

Abstract-Virtual machine consolidation is attractive in cloud computing platforms for several reasons including reduced infrastructure costs, lower energy consumption and ease of management. However, the interference between co-resident workloads caused by virtualization can violate the service level objectives (SLOs) that the cloud platform guarantees. Existing solutions to minimize interference between virtual machines (VMs) are mostly based on comprehensive micro-benchmarks or online training which makes them computationally intensive.In this paper, we present CloudScope, a system for diagnosing interference for multi-tenant cloud systems in a lightweight way. CloudScope employs a discrete-time Markov Chain model for the online prediction of performance interference of co-resident VMs. It uses the results to optimally (re)assign VMs to physical machines and to optimize the hypervisor configuration, e.g. the CPU share it can use, for different workloads. We have implemented CloudScope on top of the Xen hypervisor and conducted experiments using a set of CPU, disk, and network intensive workloads and a real system (MapReduce). Our results show that CloudScope interference prediction achieves an average error of 9%. The interference-aware scheduler improves VM performance by up to 10% compared to the default scheduler. In addition, the hypervisor reconfiguration can improve network throughput by up to 30%.

Self-Consumption Optimization of Renewable Energy Production in Distributed Clouds

“…However, we do not consider over-commitment -i.e. servers must always have enough resources to run all their allocated VMs -as they can cause performance interference leading to SLA violations [25].…”

vmBBProfiler: a black-box profiling approach to quantify sensitivity of virtual machines to shared cloud resources

“…Performing a large number of experiments to build a rich library, they could predict performance degradation of an unknown application/VM through weight-averaging its similarity with the ones from this library. Chen et al [20] used a few Hypervisor metrics to model/calculate slowdown of VMs and migrate them according to these slowdown values.…”

RA2: Predicting Simulation Execution Time for Cloud-Based Design Space Explorations

“…A DSE simulation would typically be running on a VM which shares the underlying physical resource with many other co-located VMs running potentially different applications. Despite recent advances in resource isolation in contemporary VM hypervisors, performance interference from noisy neighbors could still affect the accuracy of many performance predictions [5].…”

“…A DSE simulation would be running on a virtual machine (VM) which shares the underlying physical resource with other co-located VMs via a layer called the hypervisor [4]. Despite recent advances regarding resource isolation in contemporary hypervisors, performance interference from noisy neighbors could pose a significant problem for accurate execution time predictions [5]. Last but not least, popular cloud platforms today offer a wide range of resource types which come with varying levels of performance.…”