Bridging the gap between software and hardware techniques for I/O virtualization
ATC'08 USENIX 2008 Annual Technical Conference on Annual Technical Conference
Getting 10 Gb/s from Xen: safe and fast device access from unprivileged domains
Euro-Par'07 Proceedings of the 2007 conference on Parallel processing
GViM: GPU-accelerated virtual machines
Proceedings of the 3rd ACM Workshop on System-level Virtualization for High Performance Computing
A network interface card architecture for I/O virtualization in embedded systems
WIOV'10 Proceedings of the 2nd conference on I/O virtualization
ACM SIGOPS Operating Systems Review
Agent-based thermal management using real-time I/O communication relocation for 3D many-cores
PATMOS'11 Proceedings of the 21st international conference on Integrated circuit and system design: power and timing modeling, optimization, and simulation
Hi-index | 0.00 |
The recent emergence of network interface cards (NICs) with diverse hardware features for I/O virtualization poses an important challenge for virtual machine environments, particularly in the area of system management. In this paper, we make the case for developing a high-level network I/O virtualization management system that can translate user-relevant policy specifications into the hardware and software-specific configurations that are needed on each particular hardware platform. As a first step toward this goal, we describe and classify configuration options that are presented by a wide variety of mechanisms for NIC hardware support for virtualization, and discuss workload and policy considerations that should be factored into configuration decisions. In addition, we propose new mechanisms for intra-node guest-to-guest networking that leverage NIC hardware switching support, and we present a unified system architecture for network I/O virtualization that exposes the configuration options that we identified to high-level management layers.