Quantifying and improving I/O predictability in virtualized systems

Author

Cheng Li ; Goiri, Inigo ; Bhattacharjee, Arup ; Bianchini, R. ; Nguyen, T.D.

Author_Institution

Dept. of Comput. Sci., Rutgers Univ., Piscataway, NJ, USA

fYear

2013

fDate

3-4 June 2013

Firstpage

1

Lastpage

6

Abstract

Virtualization enables the consolidation of virtual machines (VMs) to increase the utilization of physical servers in Infrastructure-as-a-Service (IaaS) cloud providers. However, our experience shows that storage I/O performance varies wildly in the face of consolidation. Since many users may desire consistent performance, we argue that IaaS providers should offer a class of predictable-performance service in addition to existing (predictability-oblivious) services. Thus, we propose VirtualFence, a storage system that provides predictable VM performance. VirtualFence uses three main techniques: (1) non-work-conserving time-division I/O scheduling, (2) a small solid-state (SSD) cache in front of a much larger hard disk drive (HDD), and (3) space-partitioning of both the SSD cache and the HDD. Our evaluation shows that VirtualFence improves predictability significantly, while allowing cloud providers to reach any desired compromise between predictability and performance.

Keywords

cache storage; cloud computing; disc drives; hard discs; scheduling; virtual machines; virtual storage; virtualisation; HDD; I-O predictability improvement; I-O predictability quantification; IaaS cloud providers; SSD cache; VM; VirtualFence; hard disk drive; infrastructure-as-a-service cloud providers; nonwork-conserving time-division I-O scheduling; predictability-oblivious services; small solid-state cache; space-partitioning; storage I-O performance; storage I/O performance; virtual machines; virtualization; virtualized systems; Face; Interference; Performance evaluation; Resource management; Servers; Throughput; Time factors;

fLanguage

English

Publisher

ieee

Conference_Titel

Quality of Service (IWQoS), 2013 IEEE/ACM 21st International Symposium on

Conference_Location

Montreal, QC

ISSN

1548-615X

Print_ISBN

978-1-4799-0589-8

Type

conf

DOI

10.1109/IWQoS.2013.6550269

Filename

6550269