ES2: Building an Efficient and Responsive Event Path for I/O Virtualization

被引:1
|
作者
Hu, Xiaokang [1 ]
Li, Jian [2 ]
Ma, Ruhui [1 ]
Guan, Haibing [1 ]
机构
[1] Shanghai Jiao Tong Univ, Dept Comp Sci & Engn, Shanghai 200240, Peoples R China
[2] Shanghai Jiao Tong Univ, Sch Software, Shanghai 200240, Peoples R China
基金
中国国家自然科学基金;
关键词
I/O virtualization; hypervisor; VM exits; virtual interrupts; I/O requests; I/O performance;
D O I
10.1109/TCC.2020.2969660
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Hypervisor intervention in the virtual I/O event path is a main performance bottleneck for I/O virtualization because of the incurred costly VM exits. The shortcomings of prior software solutions against virtual interrupt delivery, a major source of VM exits, promoted the emergence of the hardware-based Posted-Interrupt (PI) technology. PI can provide non-exit interrupt delivery without compromising any virtualization benefit. However, it only acts on the half of the event path, i.e., the interrupt path, while guests I/O requests may also trigger a large amount of VM exits. Additionally, PI may still suffer a severe latency from the vCPU scheduling while delivering interrupts. Aiming at an optimal event path, we propose ES2 to simultaneously improve bidirectional I/O event delivery between guests and their devices. On the basis of PI, ES2 introduces hybrid I/O handling scheme for efficient I/O request delivery and intelligent interrupt redirection for enhanced I/O responsiveness. It does not require any modification to guest OS. We demonstrate that ES2 greatly reduces I/O-related VM exits with the exit handling time (EHT) below 2.5 percent for TCP streams and 0.1 percent for UDP streams, increases guest throughput by 1.9x for Memcached and 1.6x for Nginx, and keeps guest latency at a low level.
引用
收藏
页码:1358 / 1372
页数:15
相关论文
共 50 条
  • [1] ES2: Aiming at an Optimal Virtual I/O Event Path
    Hu, Xiaokang
    Zhang, Wang
    Li, Jian
    Ma, Ruhui
    Wu, Feng
    Guan, Haibing
    2017 46TH INTERNATIONAL CONFERENCE ON PARALLEL PROCESSING (ICPP), 2017, : 141 - 150
  • [2] sEBP: Event Based Polling for Efficient I/O Virtualization
    Tian, Kun
    Dong, Yaozu
    Mi, Xiang
    Guan, Haibing
    2012 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER), 2012, : 135 - 143
  • [3] Construction and analysis of Es2 efficient supersaturated designs
    Liu, Yufeng
    Ruan, Shiling
    Dean, Angela M.
    JOURNAL OF STATISTICAL PLANNING AND INFERENCE, 2007, 137 (05) : 1516 - 1529
  • [4] I/O Passthru: Upstreaming a flexible and efficient I/O Path in Linux
    Joshi, Kanchan
    Gupta, Anuj
    Gonzalez, Javier
    Kumar, Ankit
    Reddy, Krishna Kanth
    George, Arun
    Lund, Simon
    Axboe, Jens
    PROCEEDINGS OF THE 21ST USENIX SYMPOSIUM ON NETWORKED SYSTEMS DESIGN AND IMPLEMENTATION, NSDI 24, 2024, : 107 - 121
  • [5] I/O Passthru: Upstreaming a flexible and efficient I/O Path in Linux
    Joshi, Kanchan
    Gupta, Anuj
    Gonzalez, Javier
    Kumar, Ankit
    Reddy, Krishna Kanth
    George, Arun
    Lund, Simon
    Axboe, Jens
    PROCEEDINGS OF THE 22ND USENIX CONFERENCE ON FILE AND STORAGE TECHNOLOGIES, FAST 24, 2024, : 107 - 121
  • [6] Transparent I/O-Aware GPU Virtualization for Efficient Resource Consolidation
    Gonzalez, Nelson Mimura
    Elengikal, Tonia
    2021 IEEE 35TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM (IPDPS), 2021, : 131 - 140
  • [7] ReNIC: Architectural Extension to SR-IOV I/O Virtualization for Efficient Replication
    Dong, Yaozu
    Chen, Yu
    Pan, Zhenhao
    Dai, Jinquan
    Jiang, Yunhong
    ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION, 2012, 8 (04)
  • [8] The fission yeast ES2 homologue, Bis1, interacts with the Ish1 stress-responsive nuclear envelope protein
    Taricani, L
    Tejada, ML
    Young, PG
    JOURNAL OF BIOLOGICAL CHEMISTRY, 2002, 277 (12) : 10562 - 10572
  • [9] Optimizing Network I/O Virtualization with Efficient Interrupt Coalescing and Virtual Receive Side Scaling
    Dong, Yaozu
    Xu, Dongxiao
    Zhang, Yang
    Liao, Guangdeng
    2011 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER), 2011, : 26 - 34
  • [10] I/O Virtualization Utilizing an Efficient Hardware System-level Memory Management Unit
    Kornaros, George
    Harteros, Konstantinos
    Christoforakis, Ioannis
    Astrinaki, Maria
    2014 INTERNATIONAL SYMPOSIUM ON SYSTEM-ON-CHIP (SOC), 2014,