Chapter 3 – Cloud Infrastructure

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 1
Contents
 IaaS services from Amazon.
 Regions and availability zones for Amazon Web Services.
 Instances – attributes and cost.
 A repertoire of Amazon Web Services.
 SaaS and PaaS services from Google.
 SaaS and PaaS services from Microsoft.
 Open-source platforms for private clouds.
 Cloud storage diversity and vendor lock-in.
 Cloud interoperability; the Intercloud.
 Energy use and ecological impact large datacenters .
 Service and compliance level agreements.
 Responsibility sharing between user and the cloud service provider.
 User security concerns.
 User motivation.
Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 2
Existing cloud infrastructure
 The cloud computing infrastructure at Amazon, Google, and Microsoft
(as of mid 2012).
 Amazon is a pioneer in Infrastructure-as-a-Service (IaaS).
 Google's efforts are focused on Software-as-a-Service (SaaS) and
Platform-as-a-Service (PaaS).
 Microsoft is involved in PaaS.
 Private clouds are an alternative to public clouds. Open-source cloud
computing platforms such as:
 Eucalyptus,
 OpenNebula,
 Nimbus,
 OpenStack
can be used as a control infrastructure for a private cloud.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 3
Amazon Web Services (AWS)

 AWS  IaaS cloud computing services launched in 2006.


 Businesses in 200 countries used AWS in 2012.

 The infrastructure consists of compute and storage servers

interconnected by high-speed networks and supports a set of
services.

 An application developer:
 Installs applications on a platform of his/her choice.
 Manages resources allocated by Amazon.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 4
AWS regions and availability zones
 Amazon offers cloud services through a network of data centers on
several continents.
 In each region there are several availability zones interconnected by
high-speed networks.
 An availability zone is a data center consisting of a large number of
servers.

 Regions do not share resources and communicate through the

Internet.
Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 5
 Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 6
AWS instances
 An instance is a virtual server with a well specified set of
resources including: CPU cycles, main memory, secondary
storage, communication and I/O bandwidth.
 The user chooses:
 The region and the availability zone where this virtual server
should be placed.
 An instance type from a limited menu of instance types.
 When launched, an instance is provided with a DNS name; this
name maps to a
 private IP address  for internal communication within the

internal EC2 communication network.

 public IP address  for communication outside the internal
Amazon network, e.g., for communication with the user that
launched the instance.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 7
AWS instances (cont’d)

 Network Address Translation (NAT) maps external IP addresses to

internal ones.

 The public IP address is assigned for the lifetime of an instance.

 An instance can request an elastic IP address, rather than a public IP

address. The elastic IP address is a static public IP address allocated
to an instance from the available pool of the availability zone.

 An elastic IP address is not released when the instance is stopped or

terminated and must be released when no longer needed.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 8
 EC2
instance
Compute server
EC2 instance
Instance
EC2
Compute server instance
SQS

Cloud watch Compute server

Cloud front
NAT
Cloud interconnect
Elastic cache
Internet
Cloud formation

Elastic beanstalk

Elastic load balancer

AWS management
console S3 EBS SDB
S3 EBS SDB
Servers running AWS
services S3 SDB
S3
Simple DB

AWS storage servers

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 9
Steps to run an application

 Retrieve the user input from the front-end.

 Retrieve the disk image of a VM (Virtual Machine) from a

repository.

 Locate a system and requests the VMM (Virtual Machine Monitor)

running on that system to setup a VM.

 Invoke the Dynamic Host Configuration Protocol (DHCP) and the

IP bridging software to set up MAC and IP addresses for the VM.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 10
User interactions with AWS

 The AWS Management Console. The easiest way to access all

services, but not all options may be available.

 AWS SDK libraries and toolkits are provided for several

programming languages including Java, PHP, C#, and Objective-C.

 Raw REST requests.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 11
Examples of Amazon Web Services
 AWS Management Console - allows users to access the services
offered by AWS .
 Elastic Cloud Computing (EC2) - allows a user to launch a variety
of operating systems.
 Simple Queuing Service (SQS) - allows multiple EC2 instances to
communicate with one another.
 Simple Storage Service (S3), Simple DB, and Elastic Bloc Storage
(EBS) - storage services.
 Cloud Watch - supports performance monitoring.
 Auto Scaling - supports elastic resource management.
 Virtual Private Cloud - allows direct migration of parallel
applications.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 12
 CloudWatch

EC2

Linux, Debian,
Fedora,OpenSolaris,
Open Suse, Red Hat, S3
Ubuntu, Windows, Suse
Linux

EBS
SQS -Simple Queue Service

EC2

Linux, Debian, Simple DB

Fedora,OpenSolaris,
Open Suse, Red Hat,
Ubuntu, Windows, Suse
Linux

Virtual Private Cloud

Autoscaling

AWS Management Console

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 13
EC2 – Elastic Cloud Computing
 EC2 - web service for launching instances of an application under
several operating systems, such as:
 Several Linux distributions.
 Microsoft Windows Server 2003 and 2008.
 OpenSolaris.
 FreeBSD.
 NetBSD.
 A user can
 Load an EC2 instance with a custom application environment.
 Manage network’s access permissions.
 Run the image using as many or as few systems as desired.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 14
 EC2 (cont’d)
 Import virtual machine (VM) images from the user environment to an
instance through VM import.
 EC2 instances boot from an AMI (Amazon Machine Image) digitally
signed and stored in S3.
 Users can access:
 Images provided by Amazon.
 Customize an image and store it in S3.
 An EC2 instance is characterized by the resources it provides:
 VC (Virtual Computers) – virtual systems running the instance.
 CU (Compute Units) – measure computing power of each system.
 Memory.
 I/O capabilities.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 15
Instance types
 Standard instances: micro (StdM), small (StdS), large (StdL), extra
large (StdXL); small is the default.
 High memory instances: high-memory extra large (HmXL), high-
memory double extra large (Hm2XL), and high-memory quadruple
extra large (Hm4XL).
 High CPU instances: high-CPU extra large (HcpuXL).
 Cluster computing: cluster computing quadruple extra large (Cl4XL).

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 16
Instance cost
 A main attraction of the Amazon cloud computing is the low cost.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 17
S3 – Simple Storage System
 Service designed to store large objects; an application can handle
an unlimited number of objects ranging in size from 1 byte to 5 TB.
 An object is stored in a bucket and retrieved via a unique,
developer-assigned key; a bucket can be stored in a Region
selected by the user.
 Supports a minimal set of functions: write, read, and delete; it does
not support primitives to copy, to rename, or to move an object
from one bucket to another.
 The object names are global.
 S3 maintains for each object: the name, modification time, an
access control list, and up to 4 KB of user-defined metadata.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 18
S3 (cont’d)
 Authentication mechanisms ensure that data is kept secure.
 Objects can be made public, and rights can be granted to other
users.
 S3 computes the MD5 of every object written and returns it in a
field called ETag.
 A user is expected to compute the MD5 of an object stored or
written and compare this with the ETag; if the two values do
not match, then the object was corrupted during transmission
or storage.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 19
Elastic Block Store (EBS)
 Provides persistent block level storage volumes for use with EC2
instances; suitable for database applications, file systems, and
applications using raw data devices.
 A volume appears to an application as a raw, unformatted and reliable
physical disk; the range 1 GB -1 TB.
 An EC2 instance may mount multiple volumes, but a volume cannot
be shared among multiple instances.
 EBS supports the creation of snapshots of the volumes attached to an
instance and then uses them to restart the instance.
 The volumes are grouped together in Availability Zones and are
automatically replicated in each zone.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 20
SimpleDB
 Non-relational data store. Supports store and query functions
traditionally provided only by relational databases.

 Supports high performance Web applications; users can store and

query data items via Web services requests.

 Creates multiple geographically distributed copies of each data item.

 It manages automatically:
 The infrastructure provisioning.
 Hardware and software maintenance.
 Replication and indexing of data items.
 Performance tuning.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 21
SQS - Simple Queue Service
 Hosted message queues are accessed through standard SOAP
and Query interfaces.

 Supports automated workflows - EC2 instances can coordinate by

sending and receiving SQS messages.

 Applications using SQS can run independently and

asynchronously, and do not need to be developed with the same
technologies.

 A received message is “locked'' during processing; if processing

fails, the lock expires and the message is available again.

 Queue sharing can be restricted by IP address and time-of-day.

Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 22
CloudWatch
 Monitoring infrastructure used by application developers,
users, and system administrators to collect and track metrics
important for optimizing the performance of applications and for
increasing the efficiency of resource utilization.
 Without installing any software a user can monitor either seven
or eight pre-selected metrics and then view graphs and
statistics for these metrics.
 When launching an Amazon Machine Image (AMI) the user
can start the CloudWatch and specify the type of monitoring:
 Basic Monitoring - free of charge; collects data at five-minute
intervals for up to seven metrics.
 Detailed Monitoring - subject to charge; collects data at one
minute interval.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 23
AWS services introduced in 2012
 Route 53 - low-latency DNS service used to manage user's DNS
public records.
 Elastic MapReduce (EMR) - supports processing of large amounts of
data using a hosted Hadoop running on EC2.
 Simple Workflow Service (SWF) - supports workflow management;
allows scheduling, management of dependencies, and coordination
of multiple EC2 instances.
 ElastiCache - enables web applications to retrieve data from a
managed in-memory caching system rather than a much slower disk-
based database.
 DynamoDB - scalable and low-latency fully managed NoSQL
database service.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 24
AWS services introduced in 2012 (cont’d)
 CloudFront - web service for content delivery.
 Elastic Load Balancer - automatically distributes the incoming
requests across multiple instances of the application.
 Elastic Beanstalk - handles automatically deployment, capacity
provisioning, load balancing, auto-scaling, and application
monitoring functions.
 CloudFormation - allows the creation of a stack describing the
infrastructure for an application.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 25
Elastic Beanstalk
 Handles automatically the deployment, capacity provisioning, load
balancing, auto-scaling, and monitoring functions.
 Interacts with other services including EC2, S3, SNS, Elastic Load
Balance and AutoScaling.
 The management functions provided by the service are:
 Deploy a new application version (or rollback to a previous version).
 Access to the results reported by CloudWatch monitoring service.
 Email notifications when application status changes or application
servers are added or removed.
 Access to server log files without needing to login to the application
servers.
 The service is available using: a Java platform, the PHP server-side
description language, or the .NET framework.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 26
SaaS services offered by Google
(Cloud computing: the Google perspective)
 Gmail - hosts Emails on Google servers and provides a web
interface to access the Email.
 Google docs - a web-based software for building text documents,
spreadsheets and presentations.
 Google Calendar - a browser-based scheduler; supports multiple
user calendars, calendar sharing, event search, display of
daily/weekly/monthly views, and so on.
 Google Groups - allows users to host discussion forums to create
messages online or via Email.
 Picasa - a tool to upload, share, and edit images.
 Google Maps - web mapping service; offers street maps, a route
planner, and an urban business locator for numerous countries
around the world

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 27
PaaS services offered by Google
 AppEngine - a developer platform hosted on the cloud.
 Initially supported Python, Java was added later.
 The database for code development can be accessed with GQL
(Google Query Language) with a SQL-like syntax.
 Google Co-op - allows users to create customized search engines
based on a set of facets/categories.
 Google Drive - an online service for data storage.
 Google Base - allows users to load structured data from different
sources to a central repository, a very large, self-describing, semi-
structured, heterogeneous database.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 28
PaaS and SaaS services from Microsoft
 Windows Azure - an operating system; has 3 components:
 Compute - provides a computation environment.
 Storage - for scalable storage.
 Fabric Controller - deploys, manages, and monitors applications.

 SQL Azure - a cloud-based version of the SQL Server.

 Azure AppFabric, formerly .NET Services - a collection of services

for cloud applications.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 29
 Azure

Connect Applications and Data CDN

Compute Storage

Blobs Tables Queues

Fabric Controller

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 30
Azure
 Azure and Online Services are, respectively, PaaS and SaaS cloud platforms from
Microsoft. Windows
 Azure is an operating system,SQLAzure is a cloud-based version of theSQLServer,
and AzureAppFabric(formerly .NET Services) is a collection of services for cloud
applications.
 Windows Azure has three core components (see Figure 3.3): Compute, which
provides a computationenvironment; Storage for scalable storage; and Fabric
Controller, which deploys, manages, and monitorsapplications; it interconnects
nodes consisting of servers, high-speed connections, and switches.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 31
Azure
 The Content Delivery Network (CDN) maintains cache copies of data to speed up
computations. The Connect subsystem supports IP connections between the users
and their applications running on Windows Azure. The API interface to Windows
Azure is built on REST, HTTP, and XML. The platform includes five services:
Live Services, SQL Azure, AppFabric, SharePoint, and Dynamics CRM. A client
 library and tools are also provided for developing cloud applications in Visual
Studio.
 The computations carried out by an application are implemented as one or more
roles; an application typically runs multiple instances of a role. We can
distinguish (i)Web role instances used to create Web applications; (ii) Worker role
instances used to run Windows-based code; and (iii) VM role instances that run a
user-provided Windows Server 2008 R2 image.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 32
Azure
 Scaling, load balancing, memory management, and reliability are ensured by a
fabric controller, a distributed application replicated across a group of machines
that owns all of the resources in its environment – computers, switches, load
balancers – and it is aware of every Windows Azure application.
 The fabric controller decides where new applications should run; it chooses the
physical servers to optimize utilization using configuration information uploaded
with each Windows Azure application.
 The configuration information is an XML-based description of how many Web
role instances, how many Worker role instances, and what other resources the
application needs. The fabric controller uses this configuration file to determine
how many VMs to create.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 33
Azure
 Blobs, tables, queues, and drives are used as scalable storage. A blob contains
binary data; a container consists of one or more blobs. Blobs can be up to a
terabyte and they may have associated metadata (e.g., the information about
where a JPEG photograph was taken). Blobs allow a Windows Azure role instance
to interact with persistent storage as though it were a local NTFS6 file system.
Queues enable Web role instances to communicate asynchronously with Worker
role instances.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 34
Open-source platforms for private clouds
 Private clouds provide a cost-effective alternative for very large organizations. A
private cloud has essentially the same structural components as a commercial one:
the servers, the network, virtual machines monitors (VMMs) running on
individual systems, an archive containing disk images of virtual machines (VMs),
a front end for communication with the user, and a cloud control infrastructure.
Opensource cloud computing platforms such as Eucalyptus, OpenNebula, and
Nimbus can be used as a control infrastructure for a private cloud.
 Schematically, a cloud infrastructure carries out the following steps to run an
application:
 • Retrieves the user input from the front end.
 • Retrieves the disk image of a VM from a repository.
 • Locates a system and requests the VMM running on that system to set up a VM.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 35
Open-source platforms for private clouds
 Eucalyptus - can be regarded as an open-source counterpart of
Amazon's EC2.

 Open-Nebula - a private cloud with users actually logging into the

head node to access cloud functions. The system is centralized
and its default configuration uses the NFS file system.

 Nimbus - a cloud solution for scientific applications based on

Globus software; inherits from Globus:
 The image storage.
 The credentials for user authentication.
 The requirement that a running Nimbus process can ssh into all
compute nodes.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 36
Eucalyptus
 Virtual Machines - run under several VMMs including Xen, KVM,
and VMware.
 Node Controller - runs on server nodes hosting a VM and controls
the activities of the node.
 Cluster Controller - controls a number of servers.
 Cloud Controller - provides the cloud access to end-users,
developers, and administrators.
 Storage Controller - provides persistent virtual hard drives to
applications. It is the correspondent of EBS.
 Storage Service (Walrus) - provides persistent storage; similar to
S3, it allows users to store objects in buckets.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 37
Eucalyptus
 The procedure to construct a virtual machine is based on the generic one
 Theeuca2ools front end is used to request a VM.
 • The VM disk image is transferred to a compute node.
 • This disk image is modified for use by the VMM on the compute node.
 • The compute node sets up network bridging to provide a virtual network
interface controller (NIC)8
 with a virtual Media Access Control (MAC) address.9
 • In the head node the DHCP is set up with the MAC/IP pair.
 • VMM activates the VM.
 • The user can now ssh10 directly into the VM.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 38
 Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 39
 Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 40
Open-Nebula (www.opennebula.org)
 a private cloud with users actually logging into the head node
 to access cloud functions. The system is centralized and its default configuration
uses NFS (Network File System). The procedure to construct a virtual machine
consists of several steps:
 (i) the user signs into the head node using ssh;
 (ii) the system uses the onevm command to request a VM;
 (iii) the VM template disk image is transformed to fit the correct size and
configuration within the NFS directory on the head node;
 (iv) the oned daemon on the head node uses ssh to log into a compute node; (v)
the compute node sets up network bridging to provide a virtual NIC with a virtual
MAC;
 (vi) the files needed by the VMM are transferred to the compute node via the
NFS; (vii) the VMM on the compute node starts the VM; and
 (viii) the user is able to ssh directly to the VM on the compute node.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 41
Nimbus (www.nimbusproject.org)
 A cloud solution for scientific applications based on the Globus
 software. The system inherits from Globus the image storage, the credentials for
user authentication, and the requirement that a running Nimbus process can ssh
into all compute nodes. Customization in this system can only be done by the
system administrators.
 The conclusions of the comparative analysis are as follows:
 Eucalyptus is best suited for a large corporation with its own private cloud
because it ensures a degree of protection from user malice and mistakes
 . OpenNebula is best suited for a testing environment with a few servers. Nimbus
is more adequate for a scientific community less interested in the technical
internals of the system than with broad customization requirements.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 42
OpenStack (www.rackspace.com)
 OpenStack is an open-source project started in 2009 at the National Aeronautics
and Space Administration (NASA) in collaboration with Rackspace
(www.rackspace.com) to develop a scalable cloud operating system for farms of
servers using standard hardware.
 Though recently NASA has moved its cloud infrastructure to AWS in addition to
Rackspace, several other companies, including HP, Cisco, IBM, and Red Hat,
have an interest in OpenStack.
 The current version of the system supports a wide range of features such as
application programming interfaces (APIs) with rate limiting and authentication;
live VM management to run, reboot, suspend, and terminate instances; role-based
access control; and the ability to allocate, track, and limit resource utilization.
The administrators and the users control their resources using an extensible Web
application called the Dashboard.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 43
Cloud storage diversity and vendor lock-in
 Risks when a large organization relies on a single cloud service
provider:
 Cloud services may be unavailable for a short or an extended
period of time.
 Permanent data loss in case of a catastrophic system failure.
 The provider may increase the prices for service.

 Switching to another provider could be very costly due to the large

volume of data to be transferred from the old to the new provider.

 A solution is to replicate the data to multiple cloud service

providers, similar to data replication in RAID.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 44
 RAID 5 controller

a1 a2 a3 aP

b1 b2 bP b3

c1 cP c2 c3

dP d1 d2 d3

Disk 1 Disk 2 Disk 3 Disk 4

(a)
Cloud 1 Cloud 2

a1
b1 a2
c1 b2
d1
dP c1
cP
d1

Client Proxy
a3
bP
c2
d2
aP
d3
b3
c3 Cloud 3
d3

Cloud 4
(b)
Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 45
Cloud interoperability; the Intercloud
 An Intercloud  a federation of clouds that cooperate to provide a
better user experience.
 Is an Intercloud feasible?
 Not likely at this time:
 There are no standards for either storage or processing.
 The clouds are based on different delivery models.
 The set of services supported by these delivery models is large
and open; new services are offered every few months.
 CSPs (Cloud Service Providers) believe that they have a
competitive advantage due to the uniqueness of the added value of
their services.
 Security is a major concern for cloud users and an Intercloud could
only create new threats.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 46
Energy use and ecological impact

 The energy consumption of large-scale data centers and their costs

for energy and for cooling are significant.
 In 2006, the 6,000 data centers in the U.S consumed 61x10 9 KWh of
energy, 1.5% of all electricity consumption, at a cost of $4.5 billion.
 The energy consumed by the data centers was expected to double
from 2006 to 2011 and peak instantaneous demand to increase from
7 GW to 12 GW.
 The greenhouse gas emission due to the data centers is estimated to
increase from 116 x109 tones of CO2 in 2007 to 257 tones in 2020
due to increased consumer demand.
 The effort to reduce energy use is focused on computing, networking,
and storage activities of a data center.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 47
Energy use and ecological impact (cont’d)

 Operating efficiency of a system is captured by the performance

per Watt of power.
 The performance of supercomputers has increased 3.5 times faster
than their operating efficiency – 7,000% versus 2,000% during the
period 1998 – 2007.
 A typical Google cluster spends most of its time within the 10-50%
CPU utilization range; there is a mismatch between server
workload profile and server energy efficiency.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 48
Energy-proportional systems
 An energy-proportional system consumes no power when idle, very
little power under a light load and, gradually, more power as the
load increases.
 By definition, an ideal energy-proportional system is always
operating at 100% efficiency.
 Humans are a good approximation of an ideal energy proportional
system; about 70 W at rest, 120 W on average on a daily basis,
and can go as high as 1,000 – 2,000 W during a strenuous, short
time effort.
 Even when power requirements scale linearly with the load, the
energy efficiency of a computing system is not a linear function of
the load; even when idle, a system may use 50% of the power
corresponding to the full load.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 49
 Percentage of
power usage

100
Typical operating
90 region
Power
80

70
Energy
60
efficiency

50

40

30

20

10
Percentage
0 of system
0 10 20 30 40 50 60 70 80 90 100 utilization

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 50
Service Level Agreement (SLA)

 SLA - a negotiated contract between the customer and CSP; can be

legally binding or informal. Objectives:
 Identify and define the customer’s needs and constraints including
the level of resources, security, timing, and QoS.
 Provide a framework for understanding; a critical aspect of this
framework is a clear definition of classes of service and the costs.
 Simplify complex issues; clarify the boundaries between the
responsibilities of clients and CSP in case of failures.
 Reduce areas of conflict.
 Encourage dialog in the event of disputes.
 Eliminate unrealistic expectations.
 Specifies the services that the customer receives, rather than how
the cloud service provider delivers the services.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 51
Responsibility sharing between user and CSP User responsibility

SaaS PaaS IaaS

C
Interface Interface Interface L
O
U
D
Application Application Application

U
S
Operating system Operating system Operating system E
R
S
Hypervisor Hypervisor Hypervisor
E
R
V
Computing service Computing service Computing service I
C
E
Storage service Storage service Storage service
P
R
O
Network Network Network
V
I
D
Local infrastructure Local infrastructure Local infrastructure E
R
Cloud Computing: Theory and Practice.
Dr. E. Saravana Kumar Chapter 3 52
User security concerns
 Potential loss of control/ownership of data.
 Data integration, privacy enforcement, data encryption.
 Data remanence after de-provisioning.
 Multi tenant data isolation.
 Data location requirements within national borders.
 Hypervisor security.
 Audit data integrity protection.
 Verification of subscriber policies through provider controls.
 Certification/Accreditation requirements for a given cloud service.

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 53
Reasons driving decision to use public clouds

Cloud Computing: Theory and Practice.

Dr. E. Saravana Kumar Chapter 3 54