BCS601 Cloud Computing Module-3

Module-3
Cloud Platform Architecture over Virtualized Datacenters: Cloud Computing and Service Models,
Data Center Design and Interconnection Networks, Architectural Design of Compute and Storage
Clouds, Public Cloud Platforms: GAE, AWS and Azure, Inter-Cloud Resource Management.
Textbook 1: Chapter 4: 4.1 to 4.5

4.1 Cloud Computing and Service Models

4.1.1 Public, Private, and Hybrid Clouds
• Public Cloud:
o Hosted and maintained by third-party cloud service providers.
o Resources such as computing power, storage, and networking are shared among
multiple customers.
o Offers cost-effectiveness, scalability, and flexibility.
o Examples: Amazon Web Services (AWS), Microsoft Azure, Google Cloud
Platform (GCP). o Disadvantages: Limited control over infrastructure, potential
security concerns due to shared environments.
• Private Cloud:
o Dedicated cloud infrastructure for a single organization, either hosted
onpremises or by a third-party provider.
o Provides enhanced security, compliance, and performance optimization.
o Ideal for industries with strict regulatory requirements such as healthcare and
finance. o Higher operational costs due to maintenance and hardware
requirements.
o Examples: VMware vSphere, OpenStack-based private clouds.
• Hybrid Cloud:
o A combination of public and private cloud infrastructures to optimize
performance and cost-efficiency.
o Workloads can be dynamically shifted between private and public clouds based
on business needs.
o Benefits include greater flexibility, scalability, and disaster recovery options.
o Example: AWS Outposts, Microsoft Azure Stack.

pg. 1
BCS601 Cloud Computing Module-3

o
4.1.1.5 Data-Center Networking Structure o Cloud computing enables dynamic resource
allocation through virtual clusters, with gateway nodes serving as access points and security
controls. Unlike traditional grids, which rely on static resource allocation, cloud platforms
handle fluctuating workloads dynamically. Private clouds, when well-designed, can
efficiently meet these demands.
o Data centers and supercomputers share some similarities but differ significantly in scale,
architecture, and networking. Data centers rely on large clusters of servers with
integrated storage and cache, whereas supercomputers use separate data farms.
Networking also differs: supercomputers utilize high-bandwidth, custom networks,
while data centers rely on IP-based networks, such as 10 Gbps Ethernet.
o Private cloud examples include NASA’s cloud for climate modeling and CERN’s cloud
for distributing research resources globally. These clouds require varying levels of
performance, security, and data protection, governed by Service Level Agreements
(SLAs). Cloud computing builds upon grid computing but focuses more on scalable,
abstracted services rather than just storage and computing resources.

Cloud Development Trends

pg. 2
BCS601 Cloud Computing Module-3

Private clouds are expected to grow faster than public clouds due to their enhanced security and
trust within organizations. As they mature, they may transition into public or hybrid clouds,
blurring the distinction between cloud types. Hybrid clouds are likely to dominate the future.
Cloud services are categorized into different types of nodes: service-access nodes handle user
interactions, runtime supporting service nodes assist cloud operations, and independent service
nodes provide specialized services. Cloud computing optimizes performance by minimizing
data movement, reducing Internet traffic, and addressing petascale I/O challenges. However,
cloud performance, security, and QoS still require further validation through real-world
applications.

4.1.2 Cloud Ecosystem and Enabling Technologies

Cloud computing aims to shift processing, storage, and software delivery from desktops to data
centers, ensuring scalability, efficiency, and economic viability through a pay-as-you-go model.
Key design objectives include:
1. Shifting Computing to Data Centers – Moving resources from local machines to
centralized cloud infrastructure.
2. Service Provisioning & Economics – Efficient resource utilization with SLAs and cost
effective pricing models.
3. Scalability – Cloud platforms must support increasing user demands seamlessly.
4. Data Privacy Protection – Ensuring trust in cloud providers to handle sensitive data
securely.
5. High-Quality Services – Standardizing QoS for interoperability across cloud providers.
6. New Standards & Interfaces – Addressing data lock-in issues with universal APIs and
flexible access protocols.
Cloud computing reduces costs by eliminating capital expenses for hardware acquisition and
shifting to a pay-per-use model. Traditional IT systems incur both fixed and increasing
operational costs as users grow, whereas cloud computing minimizes upfront investment,
making it attractive for startups and enterprises.

• Cloud Ecosystem Components:

pg. 3
BCS601 Cloud Computing Module-3

•
o Cloud Providers: AWS, Google Cloud, Microsoft Azure, IBM Cloud, Oracle
Cloud.
o Cloud Consumers: Businesses, developers, end-users leveraging cloud
services.
o Cloud Brokers: Intermediaries that manage service usage and performance.
o Security and Compliance Solutions: Identity management, encryption, and
regulatory compliance tools.
• Key Enabling Technologies:
o Virtualization: Allows multiple virtual machines (VMs) to run on a single
physical server, improving efficiency and flexibility. o Containerization:
Docker and Kubernetes enable efficient application deployment across cloud
environments.
o Microservices Architecture: Supports modular application development,
enhancing scalability and manageability. o Artificial Intelligence (AI) &
Machine Learning (ML): Enables cloud-based predictive analytics,
automation, and intelligent decision-making.
o Blockchain Technology: Used for secure transactions, data integrity, and
decentralized cloud services.
4.1.3 Infrastructure-as-a-Service (IaaS)

pg. 4
BCS601 Cloud Computing Module-3

• Definition:
o Provides essential computing resources (compute, storage, networking) as
virtualized services over the internet.
• Key Characteristics:
o Pay-as-you-go pricing model.
o Highly scalable infrastructure.
o Automated resource provisioning.
• Examples:

•
• o AWS EC2: Provides virtual machines with flexible configurations.

pg. 5
BCS601 Cloud Computing Module-3

o Google Compute Engine (GCE): Scalable VM instances for cloud workloads.

o Microsoft Azure Virtual Machines: Supports both Windows and Linux
environments.
• Advantages:
o Eliminates the need for physical hardware investment.
o Provides high availability and disaster recovery solutions.
o Offers global reach and reduced latency through distributed data centers.
• Challenges:
o Requires advanced cloud expertise for configuration and management.
o Potential security risks if not properly configured.
4.1.4 Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS)
• Platform-as-a-Service (PaaS):
o Provides a cloud-based environment for application development and
deployment. o Includes integrated tools such as databases, runtime
environments, and development frameworks.
o Examples:

pg. 6
BCS601 Cloud Computing Module-3

o
 Google App Engine: Scalable application hosting.
 AWS Elastic Beanstalk: Automated deployment and scaling of web
applications.
 Microsoft Azure App Services: PaaS offering for .NET, Java, Python,
and more.


o Benefits:
 Reduces the complexity of software development.
 Allows developers to focus on writing code instead of managing
infrastructure.
 Scalable and flexible resource allocation.
o Challenges:
 Limited control over the underlying infrastructure. 
Vendor lock-in concerns.
• Software-as-a-Service (SaaS):

pg. 7
BCS601 Cloud Computing Module-3

o Delivers software applications over the internet without requiring installation on

local machines.
Users access software via web browsers on a subscription basis. o
Examples:
 Google Workspace (Docs, Sheets, Gmail): Cloud-based productivity
tools.
 Microsoft 365: Office applications and collaboration tools.
 Salesforce: Customer relationship management (CRM) software.
o Benefits:
 Easy access from anywhere with an internet connection.
 Automatic updates and maintenance handled by the provider.
 Lower upfront costs compared to traditional software licensing.
o Challenges:
 Data security and privacy concerns.
 Dependence on internet connectivity.
 Limited customization compared to on-premises solutions.
Mashup of Cloud Services
At the time of this writing, public clouds are in use by a growing number of users. Due to
the lack of trust in leaking sensitive data in the business world, more and more enterprises,
organizations, and communities are developing private clouds that demand deep
customization. An enterprise cloud is used by multiple users within an organization. Each
user may build some strategic applications on the cloud, and demands customized
partitioning of the data, logic, and database in the metadata representation. More private
clouds may appear in the future.
Based on a 2010 Google search survey, interest in grid computing is declining rapidly.
Cloud mashups have resulted from the need to use multiple clouds simultaneously or in
sequence. For example, an industrial supply chain may involve the use of different cloud
resources or services at different stages of the chain. Some public repository provides
thousands of service APIs and mashups for web commerce services. Popular APIs are
provided by Google Maps, Twitter, YouTube, Amazon eCommerce, Salesforce.com, etc

4.2 Data-Center Design and Interconnection Networks

4.2.1 Warehouse-Scale Data-Center Design
• Definition: Large-scale data centers that host thousands of servers to support cloud
operations.
• Key Design Factors:
o Redundancy: Ensures fault tolerance and high availability. o Scalability:
Allows the data center to expand resources efficiently.
o Energy efficiency: Uses cooling techniques such as liquid cooling and free air
cooling.

pg. 8
BCS601 Cloud Computing Module-3

o
o Security measures: Includes biometric access control, surveillance, and fire
suppression systems.
• Example: Google’s warehouse-scale computing infrastructure, which features
custombuilt hardware and software optimization.
• Data centers use raised floors to manage power cables and distribute cool air efficiently.
The cooling system relies on Computer Room Air Conditioning (CRAC) units,
which pressurize the raised floor plenum with cold air. Perforated tiles release this air in
front of server racks, which are arranged in alternating cold and hot aisles to prevent
heat mixing. Hot air from servers is recirculated back to the CRAC units, cooled, and
reintroduced into the system.
• Modern data centers enhance cooling with water-based free cooling and cooling
towers, which pre-cool condenser water before it reaches the chiller, improving
efficiency and reducing energy consumption.

•
4.2.2 Data-Center Interconnection Networks
• Purpose: Facilitates communication between servers, storage, and network devices.
• Common Network Topologies:

pg. 9
BCS601 Cloud Computing Module-3

o Fat-tree architecture: Provides high bandwidth and low latency, used in cloud
data centers.
Clos network: Ensures efficient routing and redundancy.
o Software-Defined Networking (SDN): Enables dynamic
network configuration and automation.
• Load Balancing: Distributes network traffic evenly to prevent bottlenecks.
• Fault Tolerance: Ensures network resilience through redundant connections.

•
4.2.3 Modular Data Centers in Shipping Containers
• Definition: Prefabricated data centers housed in shipping containers.
• Advantages: o Portability: Can be transported to different locations as needed. o
Scalability: Easy to add more modules for additional capacity.
o Energy Efficiency: Designed with advanced cooling systems to reduce power
consumption.
• Use Cases: Disaster recovery, military applications, rapid deployment in remote areas.
• Example: Microsoft’s Azure Modular Data Centers, which support cloud workloads
in various locations.
4.2.4 Interconnection of Modular Data Centers
• Techniques Used:
o Fiber-optic networking: Ensures high-speed data transfer between modules. o
SDN integration: Manages dynamic network reconfigurations.
o Data replication strategies: Ensures data consistency across distributed
centers.
• Challenges:
o Latency issues: Distance between modular centers can impact
performance. o Security risks: Data transmission over long distances
requires encryption

pg. 10
BCS601 Cloud Computing Module-3

o o
.
4.2.5 Data-Center Management Issues
Here are basic requirements for managing the resources of a data center. These suggestions have
resulted from the design and operational experiences of many data centers in the IT and service
industries.
• Making common users happy The data center should be designed to provide quality
service to the majority of users for at least 30 years.
• Controlled information flow Information flow should be streamlined. Sustained services
and high availability (HA) are the primary goals.
• Multiuser manageability The system must be managed to support all functions of a data
center, including traffic flow, database updating, and server maintenance.
• Scalability to prepare for database growth The system should allow growth as workload
increases. The storage, processing, I/O, power, and cooling subsystems should be
scalable.
• Reliability in virtualized infrastructure Failover, fault tolerance, and VM live migration
should be integrated to enable recovery of critical applications from failures or disasters.
• Low cost to both users and providers The cost to users and providers of the cloud system
built over the data centers should be reduced, including all operational costs.
• Security enforcement and data protection Data privacy and security defense mechanisms
must be deployed to protect the data center against network attacks and system interrupts
and to maintain data integrity from user abuses or network attacks.
• Green information technology Saving power consumption and upgrading energy
efficiency are in high demand when designing and operating current and future data
centers.
• Key Challenges:

pg. 11
BCS601 Cloud Computing Module-3

o Resource allocation: Ensuring optimal usage of compute, storage, and network

resources.
Security and compliance: Adhering to industry standards such as GDPR and
HIPAA.
o Energy efficiency: Implementing cooling solutions and renewable energy
sources. o Infrastructure monitoring: Using AI-powered analytics to predict
failures and optimize performance. o Automation: Implementing orchestration
tools like Kubernetes for workload management.
4.3 Architectural Design of Compute and Storage Clouds
Cloud Platform Design Goals: Cloud computing platforms are designed with four key goals:
scalability, virtualization, efficiency, and reliability. They support Web 2.0 applications by
managing user requests, allocating resources, and provisioning services across both physical
and virtual machines. Security in shared environments remains a critical challenge.
To establish a large-scale HPC infrastructure, cloud platforms integrate hardware and software
for seamless operation. Scalability is achieved through cluster architecture, allowing for easy
expansion of processing power, storage, and bandwidth. Reliability is enhanced by storing data
in multiple geographically distributed locations, ensuring accessibility even in case of failures.
Cloud architectures can scale efficiently by adding more servers and expanding network
connectivity as needed.
Enabling Technologies for Clouds
Cloud computing is driven by advancements in broadband and wireless networking, declining
storage costs, and improvements in Internet computing software. These technologies enable on-
demand capacity scaling, cost reduction, and service experimentation for users, while providers
benefit from higher system utilization through multiplexing, virtualization, and dynamic
resource provisioning.
Clouds rely on continuous progress in hardware, software, and networking, which collectively
enhance performance, efficiency, and scalability. These advancements ensure that cloud
platforms can dynamically allocate resources while maintaining cost-effectiveness and
flexibility. Cloud computing has become a reality due to advancements in hardware,
virtualization, and service-oriented architecture (SOA). Multicore CPUs, memory chips,
and disk arrays enable faster data centers with vast storage capacity. Virtualization supports
rapid cloud deployment and disaster recovery, while SOA enhances cloud service integration.
Progress in SaaS, Web 2.0 standards, and Internet performance has driven cloud adoption,
allowing for large-scale, multi-tenant services over massive datasets. Distributed storage
systems form the backbone of modern data centers, while improvements in license
management and automatic billing further streamline cloud operations.

pg. 12
BCS601 Cloud Computing Module-3

4.3.1 A Generic Cloud Architecture Design

• Three-layer architecture:
1. Compute layer: Manages virtual machines (VMs), containers, and serverless
functions.
2. Storage layer: Includes block storage, object storage, and file storage solutions.
3. Networking layer: Handles firewalls, load balancers, and Software-Defined
Networking (SDN).

4.
• Key Characteristics:
o On-demand resource allocation.
o Elastic scalability to meet workload demands. o Centralized management and
orchestration.
o Automated provisioning using APIs and infrastructure as code (IaC).

pg. 13
BCS601 Cloud Computing Module-3

4.3.2 Layered Cloud Architectural Development

• Service-oriented architecture (SOA): Enhances modular development and

interoperability.
• Multi-tenancy: Supports multiple users in a shared environment while ensuring
security.
• Quality of Service (QoS) considerations:
o Latency: Minimizing delays in service response. o Bandwidth: Ensuring
efficient data transmission.
o Performance optimization: Using caching, load balancing, and autoscaling.
Market-Oriented Cloud Architecture
Cloud providers must ensure QoS (Quality of Service) to meet consumer demands, as defined
in SLAs (Service Level Agreements). Traditional system-centric resource management is
insufficient; instead, market-oriented resource management is used to balance supply and
demand of cloud resources.
The architecture includes:
• Users/Brokers: Submit service requests from anywhere.
• SLA Resource Allocator: Interfaces between cloud providers and users.
• Service Request Examiner: Interprets requests and prioritizes allocations.
• Accounting Mechanism: Tracks resource usage for billing and optimization.
• VM Monitor: Monitors VM availability and entitlements.
• Dispatcher: Assigns accepted service requests to VMs.
• Service Request Monitor: Tracks request execution progress.
Clouds dynamically allocate multiple VMs on a single physical machine, allowing resource
partitioning and supporting different OS environments. The Pricing Mechanism determines
costs based on factors like time of submission (peak/off-peak), fixed vs. dynamic rates, and
resource availability.

pg. 14
BCS601 Cloud Computing Module-3

Quality of Service Factors

A cloud data center consists of multiple computing servers that allocate resources based on
service demands. QoS (Quality of Service) parameters such as time, cost, reliability, and
security are essential for commercial cloud offerings. Since business operations are dynamic,
QoS requirements must adapt over time.
Key Components of Market-Oriented Cloud Architecture:
1. SLA Resource Allocator: Manages agreements between users and cloud providers.
2. Virtual & Physical Machines: Resources are allocated dynamically.
3. Pricing & Accounting: Tracks resource usage and charges users accordingly.
4. Dispatcher & Monitors: Assigns tasks to virtual machines (VMs) and tracks execution.
5. Service Request Examiner & Admission Control: Ensures efficient resource
allocation without overloading.
Key Features:
• Customer-Driven Service Management: Adapts resources based on user requirements.
• Computational Risk Management: Identifies and mitigates risks in cloud operations.
• Autonomic Resource Management: Self-managing models adjust resource distribution
dynamically.
• Dynamic SLA Negotiation: Supports real-time adjustments based on changing demands.
By leveraging VM technology, this architecture optimizes resource allocation, ensuring
efficient, cost-effective, and reliable cloud services while adapting to real-time fluctuations in
demand.

4.3.3 Virtualization Support and Disaster Recovery

System virtualization software plays a crucial role in cloud computing by simulating hardware
execution and enabling the operation of multiple virtual environments on shared physical
infrastructure.

pg. 15
BCS601 Cloud Computing Module-3

Key Functions of Virtualization Software in Cloud Computing:

1. Hardware Virtualization: Allows unmodified operating systems to run on virtualized
hardware, enabling legacy software support.
2. Flexible Development & Deployment:
o Developers can use any OS or programming environment without infrastructure
limitations. o The same environment is used for both development and deployment,
reducing runtime errors.
3. Multi-Tenant Isolation:
o Virtual Machines (VMs) securely separate users while maximizing resource
utilization.
o Unlike traditional cluster resource sharing, virtualization ensures better security
and customization.
4. Hosting Third-Party Applications: Cloud platforms use VMs to host external
applications, offering users complete control over their software stack.
5. Scalability & Customization: Users can configure VMs to suit their needs without
interfering with other cloud tenants.
By leveraging virtualization, cloud computing platforms provide enhanced flexibility, security,
and resource efficiency, making it easier to deploy, manage, and scale applications across
diverse environments.
Virtualization Support in Public Clouds
Virtualization plays a significant role in cloud computing by abstracting physical resources,
enhancing flexibility, and improving disaster recovery. Comparison of Virtualization Support
in Public Clouds
Cloud Provider Virtualization Support Key Feature

AWS (Amazon Web Extreme flexibility using

Users can run custom applications
Services) VMs

Programming-level
Microsoft Azure virtualization (.NET) Supports application development

Google App Engine Application-level Users can only build apps within
(GAE) virtualization Google’s service framework
Each cloud provider supports different virtualization mechanisms based on their architecture
and service models.

Storage Virtualization for Green Data Centers

• Energy Consumption Concern: IT power consumption in the U.S. has doubled,
accounting for 3% of total energy use.
• Corporate Response: Fortune 500 companies are adopting energy-efficient strategies.
• Impact of Virtualization:

pg. 16
BCS601 Cloud Computing Module-3

o Reduces power usage by consolidating multiple workloads on fewer physical

servers. o Lowers operational costs and improves energy efficiency. o Supports
Green Computing initiatives.
Virtualization for IaaS (Infrastructure as a Service)
Cloud computing leverages VM technology for customizable environments, offering:
1. Workload Consolidation: Optimizes underutilized servers, reducing operational costs.
2. Legacy Application Support: Runs older software without compatibility issues.
3. Security Enhancements: Sandboxing prevents malware and application failures from
affecting the entire system.
4. Performance Isolation: Enables cloud providers to ensure better QoS (Quality of
Service) for users.

VM Cloning for Disaster Recovery

• Traditional Recovery: Slow and expensive, requiring reinstallation of hardware, OS, and
software.
• VM-Based Recovery: Faster (40% of traditional recovery time), using snapshots for live
migration.
• Disaster Recovery (DR) Strategies:
1. Data replication: Synchronous and asynchronous replication techniques.
2. Snapshot backups: Creating periodic copies of virtual machines and storage
volumes.
3. Failover mechanisms: Redundant systems automatically take over in case of
failure. 4. Geographic redundancy: Storing backup data across multiple cloud
regions.

4.3.4 Architectural Design Challenges

• Security risks: Ensuring proper authentication, encryption, and compliance measures.
• Scalability: Handling increased demand while maintaining performance.
• Interoperability: Integrating different cloud service providers and platforms.
• Data sovereignty and regulatory compliance: Ensuring adherence to GDPR, HIPAA,
and other regulations.
Challenges in Cloud Computing and Virtualization
Cloud computing has revolutionized IT infrastructure, but it presents multiple challenges
related to availability, security, performance, and standardization.
Challenge 1: Service Availability and Data Lock-in
• Single Point of Failure: Cloud services managed by a single provider can be vulnerable
to failures.
• Multi-Cloud Solutions: Using multiple cloud providers enhances high availability
(HA).
• DDoS Attacks: Can disrupt SaaS providers, requiring rapid scale-ups for defense.

pg. 17
BCS601 Cloud Computing Module-3

• Data Lock-in:
o APIs remain proprietary, restricting data and application portability.
o Standardizing APIs would enable multi-cloud deployment and “surge
computing” (using public clouds for extra workloads).

Challenge 2: Data Privacy and Security Concerns

• Public Cloud Exposure: Cloud platforms are open to security threats.
• Encryption & Firewalls: Data encryption, VLANs, and network middleboxes can
mitigate risks.
• Security Threats:
o Traditional: DoS attacks, malware, rootkits, spyware.
o Cloud-Specific: Hypervisor malware, guest hopping, VM hijacking, man-
inthe-middle attacks on VM migration.
• Data Sovereignty Laws: Many nations require customer data storage within
national boundaries.

Challenge 3: Unpredictable Performance and Bottlenecks

• I/O Bottlenecks:
o VMs efficiently share CPU and memory, but I/O performance suffers.
o Example: Amazon EC2
 STREAM benchmark: 1,355 MB/sec bandwidth needed. 
Disk writes: 55 MB/sec, showing I/O limitations.
• Data Placement Complexity:
o Data-intensive applications require optimized data transfer strategies.
o Amazon CloudFront minimizes data transport costs.
• Solutions:
o Improve I/O virtualization architectures. o
Remove bottleneck links and weak
servers.

Challenge 4: Distributed Storage and Software Bugs

• Scalable Storage Needs:
o Cloud databases must scale on-demand while ensuring high availability (HA).
o Distributed SANs improve storage resilience.
• Debugging Challenges:
o Large-scale cloud bugs cannot be easily reproduced. o VM-
based debugging can help capture cloud runtime issues. o
Simulated debugging (if well-designed) offers another solution.

pg. 18
BCS601 Cloud Computing Module-3

Challenge 5: Cloud Scalability, Interoperability, and Standardization

• Diverse Pricing Models:
o GAE (Google App Engine) charges per cycle used.
o AWS charges hourly for VM instances, even if idle.
• Standardization Efforts:
o Open Virtualization Format (OVF):
 Portable VM packaging across platforms.
 Allows cross-hypervisor VM migration (e.g., Intel ↔ AMD).
o Challenges:
 Need hypervisor-agnostic VMs.
 Support for legacy hardware in cloud load
balancing. Challenge 6: Software Licensing and Reputation
Sharing
• Licensing Challenges:
o Open-source software dominates due to flexible licensing.
o Commercial software vendors must adapt pay-as-you-go or bulk-use models.
• Reputation Management:
o Cloud-wide blacklisting (e.g., spam-prevention services blacklisting EC2 IPs).
o Potential reputation-guarding services to prevent cloud-wide bans.
• Legal Liability:
o Cloud providers and customers debate who holds legal liability for failures or
security breaches. o SLAs (Service Level Agreements) must clearly define liability.

4.4 Public Cloud Platforms: GAE, AWS, and Azure

4.4.1 Public Clouds and Service Offerings

• Major Public Cloud Providers:

pg. 19
BCS601 Cloud Computing Module-3

o Amazon Web Services (AWS): Leading IaaS and PaaS provider with services
like EC2, S3, Lambda. o Google Cloud Platform (GCP): AI-driven cloud
infrastructure with services like GKE, Cloud Run, and BigQuery. o Microsoft
Azure: Strong hybrid cloud capabilities with services like Azure Virtual Machines,
Azure Kubernetes Service, and Azure AI.
• Common Service Offerings:
o Compute power (VMs, serverless functions, containers).
o Storage solutions (object, block, file storage).
o AI/ML services, big data analytics, and security tools. o Managed databases
(SQL, NoSQL, cloud-native DBs).

o
4.4.2 Google App Engine (GAE)
• PaaS solution that allows developers to build and deploy web applications without
managing the infrastructure.
• Key Features:
o Automatic scaling of applications based on demand.
o Supports multiple programming languages (Python, Java, Go, Node.js, PHP,
etc.). o Integrated development tools and monitoring.
o Pay-as-you-go pricing model.
Google App Engine (GAE) Architecture
Google App Engine (GAE) is a Platform-as-a-Service (PaaS) that allows developers to build
and deploy applications on Google’s cloud infrastructure. It eliminates the need for server
maintenance and provides scalable cloud services.
Key Building Blocks of Google Cloud Infrastructure

pg. 20
BCS601 Cloud Computing Module-3

1. Google File System (GFS): Manages large-scale data storage.

2. MapReduce: Supports distributed computing for application development.
3. BigTable: Provides structured and semi-structured data storage.
4. Chubby: A distributed lock service for synchronization.
These components form the foundation of Google’s cloud services, enabling seamless data
storage, processing, and application management.
GAE Features and Architecture
• GAE runs user programs on Google’s infrastructure, offering scalability and
maintenance-free operation for developers.
• The frontend serves as an application framework, similar to ASP, J2EE, and JSP,
supporting Python and Java for application development.
• The GAE infrastructure supports multiple clusters of servers running GFS,
MapReduce, BigTable, and Chubby to provide efficient resource management.
• Web-based interactions allow both users and third-party developers to access Google
applications seamlessly.
This architecture ensures that applications hosted on GAE can handle large-scale data and traffic
efficiently while leveraging Google’s high-performance cloud infrastructure.

o
Summary of GAE Components and Services
Google App Engine (GAE) is an application development platform rather than an
infrastructure platform, providing tools and services for building and deploying cloud
applications.
Major Components of GAE
1. Datastore:
o Provides object-oriented, distributed, structured data storage based on
BigTable.
o Ensures secure data management operations.
2. Application Runtime Environment:

pg. 21
BCS601 Cloud Computing Module-3

o Supports scalable web programming and execution. o Compatible with

Python and Java for development.
3. Software Development Kit (SDK):
o Facilitates local application development and testing. o Allows developers
to upload application code seamlessly.
4. Administration Console:
o Simplifies the management of application development cycles.
o Focuses on software management rather than physical infrastructure.
5. GAE Web Service Infrastructure:
o Provides special interfaces for efficient storage and network resource
management.
GAE Services and Usage
• Free Access for Gmail Users:
o Users can register using a Gmail account and utilize free services within a
quota.
o Exceeding the quota requires a paid plan.
• Programming Language Support:
o GAE supports Python, Ruby, and Java, but does not provide
Infrastructureas-a-Service (IaaS).
o Unlike AWS (which offers IaaS and PaaS), GAE is strictly a PaaS model for
deploying user-built applications.
• Comparison with Other Cloud Platforms:
o GAE: Focuses on application hosting and development. o AWS:
Provides both IaaS (infrastructure) and PaaS (platform) services.
o Azure: Supports a similar model to GAE but focuses on .NET applications.
GAE simplifies cloud application deployment by abstracting infrastructure
complexities, making it an ideal choice for developers focused on web and
cloud-based applications.
4.4.3 Amazon Web Services (AWS)
AWS Components and Services
Amazon Web Services (AWS) is a leading provider of public cloud services using the
Infrastructure-as-a-Service (IaaS) model. AWS enables flexible and secure computing resource
sharing through Virtual Machines (VMs).
Key AWS Components:
1. EC2 (Elastic Compute Cloud):
o Provides virtualized computing platforms for running cloud applications.
2. S3 (Simple Storage Service):
o Offers object-oriented storage for data management.

pg. 22
BCS601 Cloud Computing Module-3

3. EBS (Elastic Block Service):

o Supports block storage for traditional applications.
4. SQS (Simple Queue Service):
o Ensures reliable message processing between processes. o Messages persist
even if the receiver process is inactive.
5. ELB (Elastic Load Balancing):
o Distributes incoming traffic across multiple EC2 instances.
oPrevents overloading and manages fault tolerance.
6. CloudWatch:
o Monitors AWS resources (e.g., CPU, disk I/O, network traffic).
oSupports autoscaling and load balancing via ELB.
7. RDS (Relational Database Service):
o Provides a fully managed relational database service.
8. Elastic MapReduce (EMR):
o Equivalent to Hadoop on EC2 for big data processing.
9. AWS Import/Export:
o Allows physical data transfer via shipping disks for high-bandwidth data
migration.
10. CloudFront:
o Implements content distribution for faster web application delivery.
11. FPS (Flexible Payment Service):
o Enables developers to charge AWS users for paid services.
12. FWS (Fulfillment Web Service):
o Provides order fulfillment services through Amazon’s logistics.
13. MPI Clusters & Cluster Compute Instances (July 2010):
o AWS introduced hardware-assisted virtualization for high-
performance computing (HPC).
o Uses EBS-based booting instead of para-virtualization.
AWS provides a flexible cloud computing environment, making it ideal for small and medium
businesses looking to scale their operations while serving large numbers of users efficiently.

pg. 23
BCS601 Cloud Computing Module-3

• Market leader in cloud services, providing comprehensive

solutions for enterprises.
• Advantages: o Global infrastructure with multiple availability
zones. o Broad range of services catering to various industry needs.
o Strong security and compliance measures.
4.4.4 Microsoft Azure

Microsoft Azure Cloud Platform

Launch Year: 2008 Objective: To address cloud computing challenges using Microsoft data
centers.
Architecture: Built on Windows OS and Microsoft virtualization technology.
Key Components of Azure:
1. Windows Azure:
o Provides a cloud platform based on Windows OS.
o Manages VMs, storage, and network resources in data centers.
2. Core Azure Services:

pg. 24
BCS601 Cloud Computing Module-3

o Live Service: Supports Microsoft Live applications and multi-machine data

access.
o .NET Service: Enables local application development and cloud execution. o
SQL Azure: Provides cloud-based relational database services with SQL Server.
o SharePoint Service: Helps in scalable business application development.
o Dynamic CRM Service: Supports business management applications (finance,
marketing, sales).
3. Integration with Microsoft Products:
o Azure services work seamlessly with Windows Live, Office Live, Exchange
Online, SharePoint Online, and Dynamic CRM Online.
4. Communication Protocols:
o Uses SOAP and REST for web-based communication and third-party cloud
integration.
5. Azure Development Kit (SDK):
o Allows developers to run a local version of Azure, develop applications, and
debug them on Windows hosts.
Key Differentiators of Azure:
• Seamless integration with Microsoft products.
• Strong support for enterprise applications (SharePoint, CRM, SQL).
• Developer-friendly SDK for local development and testing.
• Hybrid cloud capabilities with third-party cloud integration.
Microsoft Azure offers a powerful and scalable cloud solution, making it a preferred choice for
enterprises that rely on Microsoft technologies.
INTER-CLOUD RESOURCE MANAGEMENT
Extended Cloud Computing Services : Cloud computing is structured into different service
layers, each catering to distinct functionalities and user needs. These services range from
hardware and networking to software applications and runtime support.

Six Layers of Cloud Services

pg. 25
BCS601 Cloud Computing Module-3

1. Hardware as a Service (HaaS) o Provides physical

infrastructure (servers, storage, networking).
o Examples: VMware, Intel, IBM, XenEnterprise.
2. Network as a Service (NaaS) o Manages network
connectivity and virtual LANs.
o Examples: AT&T, Qwest, AboveNet.
3. Location as a Service (LaaS) o Offers data center
collocation services (housing, power, security). o Sometimes
referred to as Security as a Service (SaaS).
o Examples: Savvis, Internap, Digital Realty Trust.
4. Infrastructure as a Service (IaaS) o Provides compute, storage,
and networking resources on demand.
o Examples: Amazon AWS, Microsoft Azure, Rackspace Cloud, IBM Ensembles.
5. Platform as a Service (PaaS) o Offers development
environments, frameworks, and APIs.
o Examples: Google App Engine, Force.com, Microsoft Azure, IBM BlueCloud.
6. Software as a Service (SaaS) o Delivers fully managed
applications over the internet.
o Examples: Salesforce, Webex, Concur, RightNOW, Teleo, Netsuite.
Roles of Cloud Service Players
Cloud Players IaaS Role PaaS Role SaaS Role

IT Administrators/ Cloud Monitor SLAs & enable Monitor SLAs &

Providers Monitor SLAs platforms
deploy software
Software Developers/ Deploy & store Enable platforms Develop & deploy
Vendors data via APIs software
Deploy & store Develop & test web
End Users/ Business Users data Use business software
applications
Trends in Cloud Services
• SaaS Expansion: Initially led by CRM applications, now extends to HR, finance, and
distributed collaboration.
• PaaS Adoption: Increasing with platforms like Google App Engine, Salesforce,
Facebook, Microsoft Azure.
• IaaS Growth: Driven by AWS, Azure, and Rackspace, providing scalable compute and
storage.
• Vertical Cloud Services: Multiple integrated services working together in a cloud
mashup.
Cloud Software Stack

pg. 26
BCS601 Cloud Computing Module-3

Cloud platforms are designed with high throughput, fault tolerance, and high availability (HA).
The software stack consists of:
1. Virtualization Layer – Flexible infrastructure using VMs.
2. Storage Layer – Manages large-scale data storage.
3. Database Layer – Supports structured and unstructured data.
4. Compute Layer – Executes cloud applications.
5. Application Layer – Provides user-facing software services.
Runtime Support & Cloud Scheduling
• Cluster Monitoring: Tracks the status of cloud resources.
• Job Management System: Schedules tasks efficiently across cloud nodes.
• MapReduce Scheduling: Specialized for big data processing in distributed cloud
environments.
Advantages of SaaS Model o No upfront
costs for hardware and software.
o Lower operational costs for providers compared to
traditional hosting.
o Cloud storage options (vendor-specific or public cloud). o
Scalability and flexibility for businesses and developers.
Summary of Resource Provisioning and Platform Deployment in Cloud Computing
The emergence of cloud computing has led to significant changes in software and hardware
architecture, emphasizing processor cores, VM instances, and parallelism at the cluster node
level. Effective resource provisioning and platform deployment are essential for optimizing
cloud infrastructure performance.
1. Provisioning of Compute Resources (VMs)
Cloud providers allocate resources through Service Level Agreements (SLAs), ensuring CPU,
memory, and bandwidth availability for a preset period. Balancing underprovisioning (risking
SLA violations) and overprovisioning (causing resource waste) is a key challenge. Efficient
provisioning involves VM installation, live migration, and failure recovery. Examples
include Amazon EC2, IBM’s Blue Cloud, and Microsoft Azure, all of which rely on
virtualization technologies.
2. Resource Provisioning Methods
Three primary resource provisioning methods exist:
• Demand-Driven Provisioning: Adjusts resources based on utilization thresholds (e.g.,
Amazon EC2 auto-scaling). It is simple but ineffective for abrupt workload changes.
• Event-Driven Provisioning: Allocates resources based on predicted workload spikes
during specific events (e.g., seasonal sales). This method minimizes Quality of Service
(QoS) loss if predictions are accurate.
• Popularity-Driven Provisioning: Allocates resources based on Internet search
trends. It anticipates traffic surges but may waste resources if popularity forecasts are
incorrect.

pg. 27
BCS601 Cloud Computing Module-3

3. Dynamic Resource Deployment

Dynamic provisioning allows cloud systems to scale across multiple resource sites. The
InterGrid infrastructure, developed by Melbourne University, enables cross-grid resource

pg. 28
BCS601 Cloud Computing Module-3

allocation via InterGrid Gateways (IGGs), allowing interaction between local clusters and
external cloud providers.

4. Provisioning of Storage Resources

Cloud storage is built on physical or virtual servers and is essential for scalable applications
like email systems and web search engines. Future storage solutions will likely integrate
solidstate drives (SSDs) for higher performance. Common cloud storage systems include:
• Google File System (GFS) – High bandwidth for continuous access.
• Hadoop Distributed File System (HDFS) – Open-source alternative to GFS.
• Amazon S3 & EBS – Remote data storage and virtual disk services.
Cloud databases (e.g., Google BigTable, Amazon SimpleDB, Microsoft Azure SQL Service)
enable structured and semi-structured data management, allowing developers to build
applications efficiently.
Conclusion
Efficient cloud resource provisioning is critical for balancing performance, cost, and scalability.
By leveraging various provisioning strategies and storage solutions, cloud platforms optimize
their computing environments while maintaining service quality and efficiency.

Virtual Machine Creation and Management

This section discusses key aspects of cloud infrastructure management, including resource
management for independent services, execution of third-party applications, and various
components involved in virtual machine (VM) management.

pg. 29
BCS601 Cloud Computing Module-3

1. Independent Service Management: Cloud infrastructures provide APIs for running

independent services. Amazon's Simple Queue Service (SQS) facilitates reliable
communication between service providers. This enables multiple cloud applications to
run simultaneously.
2. Running Third-Party Applications: Cloud platforms support third-party applications
through web services, allowing developers to build applications with APIs instead of
traditional runtime libraries. Examples include Google App Engine (GAE), Microsoft
Azure, and IBM’s WebSphere.
3. Virtual Machine Manager (VMM): The VMM links gateways to resources, managing
VMs using virtualization technologies. It supports multiple environments, including
OpenNebula, Amazon EC2, and Grid’5000. The manager enables remote VM
deployment using hypervisors like Xen.
4. Virtual Machine Templates: VM templates define configuration parameters such as
processor allocation, memory, disk image, OS kernel, and pricing. Administrators
maintain a repository of templates to ensure consistency across the infrastructure.
5. Distributed VM Management: The InterGrid platform uses distributed VM
management, allowing gateways to request, allocate, and redirect VM resources across
multiple sites. A peering policy ensures workload balancing across cloud sites.
6. InterCloud Resource Exchange: To meet global demand, cloud providers establish
data centers in different geographical locations. The Melbourne group’s InterCloud
architecture enables seamless integration of multiple cloud providers for dynamic
resource allocation and load balancing.
7. Cloud Exchange (CEx): The CEx acts as a marketplace for cloud resources, allowing
providers and consumers to negotiate service agreements based on economic models
such as auctions and commodity markets. It ensures QoS-driven service delivery and
secure financial transactions through SLA-based contracts.
This emphasizes the need for efficient VM management, workload balancing, and inter-cloud
resource sharing to optimize cloud computing performance and scalability.

pg. 30
BCS601 Cloud Computing Module-3

pg. 31