"LISA is the conference that I send my system administrators to so they can bring the latest tools and techniques back to the rest of the team. Much of our current environment can be traced directly back to LISA."
Cory Lueninghoener, Deputy Group Leader of Production High Performance Computing at Los Alamos National Laboratory
"LISA is the conference that I send my system administrators to so they can bring the latest tools and techniques back to the rest of the team. Much of our current environment can be traced directly back to LISA."
Cory Lueninghoener, Deputy Group Leader of Production High Performance Computing at Los Alamos National Laboratory
"LISA is where professionals share what's hot in designing, building, and maintaining critical systems."
Tom Limoncelli, author, speaker, and system administrator
"I keep coming back for the technical content and the personal networking opportunities. I attend for career development."
LISA '13 Attendee
"I use LISA to benchmark the SA activities in my company."
LISA '13 Attendee
"LISA is where I find direction for evolving the my core professional skills."
LISA '13 Attendee
"Information from LISA helps us push the envelope on automation and scaling, allowing a team of four to manage over 3000 Firefox build and test systems running 15 different operating systems."
Amy Rich, Manager of Release Engineering Operations at Mozilla
"LISA is where professionals share what's hot in designing, building, and maintaining critical systems."
Tom Limoncelli, author, speaker, and system administrator
"LISA is the place where industry best practices and cutting-edge research come together to advance system administration."
HotRestore: A Fast Restore System for Virtual Machine Cluster
Refereed Paper
Wednesday, November 12, 2014 - 2:00pm-2:15pm
Authors:
Lei Cui, Jianxin Li, Tianyu Wo, Bo Li, Renyu Yang, Yingjie Cao, and Jinpeng Huai, Beihang University
Abstract:
A common way for virtual machine cluster (VMC) to tolerate failures is to create distributed snapshot and then restore from the snapshot upon failure. However, restoring the whole VMC suffers from long restore latency due to large snapshot files. Besides, different latencies would lead to discrepancies in start time among the virtual machines. The prior started virtual machine (VM) thus cannot communicate with the VM that is still restoring, consequently leading to the TCP backoff problem.
In this paper, we present a novel restore approach called HotRestore, which restores the VMC rapidly without compromising performance. Firstly, HotRestore restores a single VM through an elastic working set which prefetches the working set in a scalable window size, thereby reducing the restore latency. Second, HotRestore constructs the communication-induced restore dependency graph, and then schedules the restore line to mitigate the TCP backoff problem. Lastly, a restore protocol is proposed to minimize the backoff duration. In addition, a prototype has been implemented on QEMU/ KVM. The experimental results demonstrate that HotRestore can restore the VMC within a few seconds whilst reducing the TCP backoff duration to merely dozens of milliseconds.
USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.
@inproceedings {186491,
author = {Lei Cui and Jianxin Li and Tianyu Wo and Bo Li and Renyu Yang and Yinglie Cao and Jinpeng Huai},
title = {HotRestore: A Fast Restore System for Virtual Machine Cluster},
booktitle = {28th Large Installation System Administration Conference (LISA14)},
year = {2014},
isbn = {978-1-931971-17-1},
address = {Seattle, WA},
pages = {10--25},
url = {https://www.usenix.org/conference/lisa14/conference-program/presentation/cui},
publisher = {{USENIX} Association},
}
Twitter
Facebook
LinkedIn
Google+
YouTube
connect with us