International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1472
SIR (System Information Reporter) Framework for Performance
Comparison of Different Operating Systems with KVM Hypervisor
Abhishek K.L
1
, Anupama K.C
2
1
M.Tech, IV Sem, Dept. of ISE, Bangalore Institute of Technology, Bangalore, Karnataka, India
2
Assistant Professor, Dept. of ISE, Bangalore Institute of Technology, Bangalore, Karnataka, India
---------------------------------------------------------------------***--------------------------------------------------------------------
Abstract-Hypervisors enable multiple operating systems
to run above it with the help of virtualization technique
by sharing underlying physical resources. KVM (Kernel-
based Virtual Machine) hypervisor is chosen here which
uses hybrid virtualization technique i.e., it uses full
virtualization technique along with hardware assisted
virtualization. It is motivating to analyze different
Operating Systems (OSs) performance with KVM
hypervisor. We have chosen three guest operating
systems, namely Windows 7, CentOS7 and Ubuntu 14.02
for the experimentation. The three Operating Systems
(OSs) are prudently chosen to represent three categories
namely Hardware virtualized gust, para-virtualized
commercial guest and para-virtualized free guest. The
Operating Systems are compared by performance tests of
CPU utilization, memory management, disk activity, and
network communication. Important System information
is gathered using SIR (System Information Reporter)
framework on the respective guest Operating System on
KVM hypervisor.
Keywords: virtualization, hypervisor, SIR, KVM.
I. INTRODUCTION
Virtualization technology is explanation of
something a well known as became lost in basic
material, computerized information devices, occupied
systems or a server which will be created in virtual
(rather than actual) way of doing thing for the usage.
These technologies are accepted on a large scale in the
society right to its expansion in show and tell the
resources across the determined regions. Virtualization
comes facing the romp where there is a World Wide
Web connecting places to places. Virtualization allows
active on many occupied systems at a presage which
will give stage to associate those hired systems
performance [1].
Hypervisors empower distributed computing
model to give adaptable foundations and on interest
access to processing assets as they backing numerous
working frameworks to keep running on one physical
server simultaneously. Hypervisor sellers do assert that
they have invalidated virtualization overhead totally
contrast with local framework, yet there exists certain
moment virtualization overhead on the grounds that
virtual machines need to speak with center layer
hypervisor to get to the hidden physical equipment
furthermore there is an effect of other virtual machines
running on the same hypervisor Different virtualization
systems like full virtualization, para-virtualization and
cross breed model virtualization are utilized as a part of
hypervisors advancement [2].
Performance evaluation involves four types of
parameters, to be compared in three different
stages.CPU performance assessment, network
performance assessment, I/O read-write performance
assessment, memory performance assessment are the
four parameters which will be providing information
about the operating systems performances. Three
stages of comparison of performances are low
workload, medium workload and high workload which
will provide the performance results in three different
levels leads to easy evaluation of performances. SIR
framework will give the information about the CPU
processing time, memory used and other valuable
information as well, which can be plotted to find the
results changes in the performance assessments of all
the parameters which are involved [3].
II. RELATED WORK
P. VijayaVardhan Reddy, et-al [4]the paper,
DzEvaluation of different hypervisors performance in the
private clouddz evaluates the performance of three
hypervisors ESXi, XenServerǯs and KVM using S)GAR
framework for system information and Passmark for system
workloads in the private cloud environment. Private cloud
has been designed using open source cloud computing
software CloudStack. Hypervisors are deployed as hosts in
the CloudStack. This paper recommends best suited
hypervisors for respective workloads in the private cloud
based on the performance of system information and system
workloads.