TCP Vegas: new techniques for congestion detection and avoidance
SIGCOMM '94 Proceedings of the conference on Communications architectures, protocols and applications
Simulation-based comparisons of Tahoe, Reno and SACK TCP
ACM SIGCOMM Computer Communication Review
Decay-usage scheduling in multiprocessors
ACM Transactions on Computer Systems (TOCS)
Achieving Service Rate Objectives with Decay Usage Scheduling
IEEE Transactions on Software Engineering
A Case For Grid Computing On Virtual Machines
ICDCS '03 Proceedings of the 23rd International Conference on Distributed Computing Systems
Measuring and Understanding User Comfort With Resource Borrowing
HPDC '04 Proceedings of the 13th IEEE International Symposium on High Performance Distributed Computing
Towards virtual networks for virtual machine grid computing
VM'04 Proceedings of the 3rd conference on Virtual Machine Research And Technology Symposium - Volume 3
Linux Kernel Development
JSSPP'04 Proceedings of the 10th international conference on Job Scheduling Strategies for Parallel Processing
VSched: Mixing Batch And Interactive Virtual Machines Using Periodic Real-time Scheduling
SC '05 Proceedings of the 2005 ACM/IEEE conference on Supercomputing
The user in experimental computer systems research
Proceedings of the 2007 workshop on Experimental computer science
The user in experimental computer systems research
ecs'07 Experimental computer science on Experimental computer science
Server virtualization in autonomic management of heterogeneous workloads
ACM SIGOPS Operating Systems Review
FEAS: a full-time event aware scheduler for improving responsiveness of virtual machines
ACSC '12 Proceedings of the Thirty-fifth Australasian Computer Science Conference - Volume 122
| Hi-index | 0.00 |
We are developing a distributed computing system, Virtuoso, which presents virtual machines (VMs) as its fundamental abstraction to end users. Long-running noninteractive VMs may coexist on the same host used to support VMs being used by highly interactive users. We must simultaneously provide high average computation rates to the non-interactive VMs while keeping the users of the interactive VMs happy. We report here an initial work on using direct user feedback to achieve this balance. The user is provided with a (physical or logical) button that can be pressed when he feels his machine is responding inadequately. In response, the scheduler boosts the priority of his VMs relative to the other VMs in the system. The priority then declines with time. The goal of the control algorithm driven by this mechanism is to maintain a targeted average time between button presses while simultaneously delivering a high compute rate to the other VMs.