Un-IOV: Achieving Bare-Metal Level I/O Virtualization Performance for Cloud Usage With Migratability, Scalability and Transparency

被引:0
作者
Zhang, Zongpu [1 ]
Xia, Chenbo [1 ]
Liang, Cunming [2 ]
Li, Jian [1 ]
Yu, Chen [3 ]
Bie, Tiwei [2 ]
Martin, Roberts [2 ]
Dan, Daly [2 ]
Wang, Xiao [2 ]
Liu, Yong [2 ]
Guan, Haibing [1 ]
机构
[1] Shanghai Jiao Tong Univ, Dept Comp Sci & Engn, Shanghai 200240, Peoples R China
[2] Intel Corp, Shanghai 200241, Peoples R China
[3] Huazhong Univ Sci & Technol, Sch Comp Sci & Technol, Wuhan 430074, Peoples R China
基金
中国国家自然科学基金;
关键词
Virtual machine monitors; Virtualization; Performance evaluation; Hardware; Cloud computing; Switches; Kernel; I/O virtualization; live migration; SR-IOV;
D O I
10.1109/TC.2024.3375589
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
I/O virtualization is utilized by cloud platforms to provide tenants with efficient, scalable, and manageable network and storage services. The de-facto industrial standard, paravirtualization, offers rich cloud functionality by introducing split front-end and back-end drivers in the guest and host operating systems, respectively. Given this fact, paravirtualization incurs host inefficiency and performance overhead. Thus, emerging hardware virtio accelerators (i.e., SRIOV-capable devices that conform to virtio specification) with device passthrough technologies mitigate the performance issue. However, adopting these devices presents the challenge of insufficient support for live migration. This paper proposes Un-IOV, a novel I/O virtualization system that simultaneously achieves bare-metal level I/O performance and migratability. The key idea is to develop a new hybrid virtualization stack with: (1) a host-bypassed direct data path for virtio accelerators, and (2) a relayed control path guaranteeing seamless live migration support. Un-IOV achieves high scalability by consuming minimum host resources. Extensive experiment results demonstrate that Un-IOV achieves superior network and storage virtualization performance than software implementations with comparable performance of direct passthrough I/O virtualization, while imposing zero guest modification (i.e., guest transparency).
引用
收藏
页码:1655 / 1668
页数:14
相关论文
共 44 条
[1]  
Ahmad I., 2011, 2011 USENIX ANN TECH, P45
[2]  
[Anonymous], Inventec FPGA IPU C5020X adapter
[3]  
[Anonymous], P-Tile Avalon Streaming Intel FPGA IP for PCI Express User Guide
[4]  
[Anonymous], Mellanox BlueField SmartNIC for Ethernet
[5]  
Belay Adam, 2014, Proceedings of the 11th USENIX Symposium on Operating Systems Design and Implementation (OSDI '14). OSDI '14, P49
[6]  
Cheng Kevin, 2020, VEE '20: Proceedings of the 16th SIGPLAN/SIGOPS International Conference on Virtual Execution Environments, P45, DOI 10.1145/3381052.3381317
[7]  
Cooper B.F., 2010, ACM SOCC, P143, DOI DOI 10.1145/1807128.1807152
[8]  
Dalton M, 2018, PROCEEDINGS OF THE 15TH USENIX SYMPOSIUM ON NETWORKED SYSTEMS DESIGN AND IMPLEMENTATION (NSDI'18), P373
[9]  
Dey K., 2014, P ACM S CLOUD COMP
[10]   PacketMill: Toward Per-Core 100-Gbps Networking [J].
Farshin, Alireza ;
Barbette, Tom ;
Roozbeh, Amir ;
Maguire, Gerald Q., Jr. ;
Kostic, Dejan .
ASPLOS XXVI: TWENTY-SIXTH INTERNATIONAL CONFERENCE ON ARCHITECTURAL SUPPORT FOR PROGRAMMING LANGUAGES AND OPERATING SYSTEMS, 2021, :1-17