4 S.S. Manvi, G.

Krishna Shyam / Journal of Network and Computer Applications ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Table 2 the computer and the database. Cloud storage systems generally
Physical and logical resources. rely on hundreds of data servers. Because computers occasionally
require maintenance or repair, it is important to store the same
Physical resources Logical resources
information on multiple machines. This is called redundancy.
CPU Operating system Without redundancy, a cloud storage system could not ensure
Memory Energy clients that they could access their information at any given time.
Storage Network throughput/ bandwidth Most systems store the same data on servers that use different
Workstations Information security, Protocols
Network elements APIs
power supplies. That way, clients can access their data even if one
Sensors/actuators Network loads, delays power supply fails. The two biggest concerns about cloud storage
are reliability and security. Clients are not likely to entrust their
data to another company without a guarantee that they will be
estimation. In this section, we present significance of resource able to access their information whenever they want and no one
types and issues in resource management. else will be able to get it. Through this, clouds provide storage as a
service.
(iv) Workstations: IT managers are seeing a trend where more
3.1. Type of resources
powerful PCs could be classified as workstations. Bigger CPUs,
faster graphics, and upwards of 20 Gb of RAM–these are machines
A resource is any physical or virtual component of limited
that are designed to do a lot of local processing. The challenge is to
availability within a computer system. Every device connected to a
take advantage of cloud technology and economics and use
computer system is a resource. Every internal system component
workstations to bring high-performance computing (HPC) cap-
is a resource. We have listed out various physical and logical
abilities to the corporate user or any user connected to the Internet
resources in Table 2. Now we shall discuss the impact of the above
with good connectivity. To accomplish this, cloud vendors are
mentioned resources on the performance of the clouds.
working with applications that live and do some work on the local
workstation, but then shunt much of the workload out into the
3.1.1. Physical resources cloud and back again. In order to see that the users get the best
Physical resources typically include processors, memory, and services at all times, proper functioning of workstations are very
peripheral devices. Physical resources vary fairly dramatically from much desired. The systems used as workstations should be of
computer to computer. For example, a typical PC system might higher configuration so that they withstand the overloaded net-
have 640 K of memory, one 20 Mbyte Winchester disk, one floppy work. The issue here is to see that workstations operate on its
disk drive, a single keyboard, and a single video display. A typical own, without human interference. Through this, clouds provide
mainframe system has several parallel processors, hundreds of monitoring as a service.
disks, tens of millions of bytes of memory, hundreds of terminals, (v) Network elements: Managing millions of network compo-
tapes, and other special purpose peripherals, and is connected to a nents (hubs, bridges, switches etc.) lead to unsustainable admin-
global network with thousands of other similar computers. Cloud istrator costs, requiring automated methods for typical system
providers can offer resource provisioning plans to consumers, management tasks. The automated methods need to deal with
namely short-term on-demand and long-term reservation plans. increased monitoring network sizes of several orders of magni-
Outsourcing techniques can be used to take advantage of cloud tudes higher than current systems. Through this, clouds provide
computing infrastructure for providing scalability and high avail- communication as a service.
ability capabilities to the web applications deployed on it; this (vi) Sensors/actuators: Proliferation of applications involving
would definitely increase the number of cloud consumers and Internet-connected objects, has recently given rise to the notion of
hence increase the resource utilization of clouds. Now let us clouds of Internet-connected objects (i.e. sensors, actuators, devices),
discuss some of the important physical resources and its impact which are promoted as large-scale networks of spatially distributed
on cloud computing. entities with scalable processing and storage capabilities.
(i) CPU (central processing unit): It performs most of the
processing inside a computer. The issue in cloud computing is
the CPU utilization. CPU utilization refers to a computer's usage of 3.1.2. Logical resources
processing resources, or the amount of work handled by a CPU. Logical resources are system abstractions which have tempor-
Actual CPU utilization in cloud varies depending on the amount ary control over physical resources. They can support in develop-
and type of managed computing tasks. Certain tasks require heavy ment of applications and efficient communication protocols.
CPU time, while others require less because of non-CPU resource The significance of logical resources in cloud computing is as
requirements. Proper CPU usage makes it easy to consume massive follows.
amounts of compute power for batch processing, data analysis, (i) Operating system: It provides users with “logical” well-
and high performance computing needs. behaved environment to manage physical (hardware) resources
(ii) Memory: The cloud computer architecture asks for a as well as offers mechanisms and policies for the control of object/
clustered structure of the memory resources in the form of virtual resources. The operating system performs file management, device
entities. Gone are the days when memory management was done management, performance management, security and fault toler-
using the static methods. As cloud environment is dynamic and ance management, thereby facilitating efficient utilization of the
volatile, there is a strong need to inculcate the dynamic memory available resources.
allocation trends in the cloud based systems. The increased (ii) Energy: The main technique applied to minimize energy
number of cores in cloud servers combined with the rapid consumption is concentrating the workload to the minimum of
adoption of virtualization technologies also creates huge demand physical nodes and switching idle nodes off. This approach
for memory. requires dealing with the power/performance trade-off, as perfor-
(iii) Storage: It refers to saving data to an off-site storage system mance of applications can be degraded due to the workload
maintained by a third party. Instead of storing information to consolidation.
computer's hard drive or other local storage device, we save it to a (iii) Network throughput/bandwidth: In cloud computing, we are
remote database. The Internet provides the connection between concerned about measuring the maximum data throughput in bits

Please cite this article as: Manvi SS, Krishna Shyam G. Resource management for Infrastructure as a Service (IaaS) in cloud computing:
A survey. Journal of Network and Computer Applications (2013), http://dx.doi.org/10.1016/j.jnca.2013.10.004i