Module 3

Questions and Answer Marks

1. Differentiate between centralized and distributed computing 5

Ans:-
Aspect Centralized Computing Distributed Computing
Definition All processing is done at a single Processing is spread across multiple
location. systems.
Resource Resources are concentrated in a central Resources are distributed among
Allocation server or data center. multiple computers or nodes.
Scalability Limited scalability due to centralized Highly scalable as additional nodes
control. can be added.
Fault Single point of failure—if the main High fault tolerance—if one node
Tolerance server fails, everything is affected. fails, others can continue.
Response Can be slower due to congestion at the Faster response due to parallel
Time central system. processing.
Security Easier to secure because of centralized More complex security due to
control. multiple nodes.
Cost Lower initial costs but may require Higher initial setup cost but can be
Efficiency expensive maintenance. cost-effective in the long run.
Usage Used in traditional mainframe systems Used in cloud computing, peer-to-
Examples and banking servers. peer networks, and blockchain.
Data Storage All data is stored in a single centralized Data is replicated and distributed
database. across multiple locations.
Flexibility Less flexible due to centralized More flexible and adaptable to
management. changes.
2. List and explain types of clouds with a neat diagram. 10

Ans:- Types of Clouds:

1. Public
2. Private
3. Hybrid

1. 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.

2. Private Cloud:

o Dedicated cloud infrastructure for a single organization, either hosted on premises 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.

3. 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,MicrosoftAzure Stack.

3. Explain structure of data center networking with a neat diagram 10

Ans:- Data-Center Networking Structure

 The core of a cloud is the server cluster (or VM cluster). Cluster nodes are used as compute nodes.
 A few control nodes are used to manage and monitor cloud activities.
 The scheduling of user jobs requires that you assign work to virtual clusters created for users.
 The gateway nodes provide the access points of the service from the outside world. These gateway
nodes can be also used for security control of the entire cloud platform.
 In physical clusters and traditional grids, users expect static demand of resources. Clouds are
designed to handle fluctuating workloads, and thus demand variable resources dynamically.
 Private clouds will satisfy this demand if properly designed and managed.
 Data-center server clusters are typically built with large number of servers, ranging from thousands
to millions of servers (nodes). For example, Microsoft has a data center in the Chicago area that
has 100,000 eight-core servers, housed in 50 containers.
 In supercomputers, a separate data farm is used, while a data center uses disks on server nodes
plus memory cache and databases.
 Data centers and supercomputers also differ in networking requirements, as illustrated in Figure.
Supercomputers use custom-designed high-bandwidth networks such as fat trees or 3D torus
networks. Data-center networks are mostly IP-based commodity networks, such as the 10 Gbps
Ethernet network, which is optimized for Internet access.
 Figure shows a multilayer structure for accessing the Internet.
 The server racks are at the bottom Layer 2, and they are connected through fast switches (S) as
the hardware core.
 The data center is connected to the Internet at Layer 3 with many access routers (ARs) and border
routers (BRs).
4. Explain objectives of cloud computing 5

Ans:- Six design objectives for cloud computing:

• Shifting computing from desktops to data centers Computer processing, storage, and software
delivery is shifted away from desktops and local servers and toward data centers over the Internet.
• Service provisioning and cloud economics Providers supply cloud services by signing SLAs with
consumers and end users. The services must be efficient in terms of computing, storage, and power
consumption. Pricing is based on a pay-as-you-go policy.
• Scalability in performance The cloud platforms and software and infrastructure services must be able
to scale in performance as the number of users increases.
• Data privacy protection Can you trust data centers to handle your private data and records? This
concern must be addressed to make clouds successful as trusted services.
• High quality of cloud services The QoS of cloud computing must be standardized to make clouds
interoperable among multiple providers.
• New standards and interfaces This refers to solving the data lock-in problem associated with data
centers or cloud providers. Universally accepted APIs and access protocols are needed to provide high
portability and flexibility of virtualized applications.
5. Explain the cost model with a neat diagram 5

Ans:- Cost Model :

 In traditional IT computing, users must acquire their own computer and peripheral equipment as
capital expenses.
 In addition, they have to face operational expenditures in operating and maintaining the computer
systems, including personnel and service costs.
 Figure 4.3(a) shows the addition of variable operational costs on top of fixed capital investments
in traditional IT.
 The fixed cost is the main cost, and that it could be reduced slightly as the number of users
increases. However, the operational costs may increase sharply with a larger number of users.
 Therefore, the total cost escalates quickly with massive numbers of users.
 On the other hand, cloud computing applies a pay-per-use business model, in which user jobs are
outsourced to data centers.
 To use the cloud, one has no up-front cost in hardware acquisitions. Only variable costs are
experienced by cloud users, as demonstrated in Figure 4.3(b).
 Overall, cloud computing will reduce computing costs significantly for both small users and large
enterprises.
 Computing economics does show a big gap between traditional IT users and cloud users.
 The savings in acquiring expensive computers up front releases a lot of burden for startup
companies.
 The fact that cloud users only pay for operational expenses and do not have to invest in permanent
equipment is especially attractive to massive numbers of small users.
 This is a major driving force for cloud computing to become appealing to most enterprises and
heavy computer users.
 In fact, any IT users whose capital expenses are under more pressure than their operational
expenses should consider sending their overflow work to utility computing or cloud service
providers.
6. Explain cloud ecosystem with a neat diagram. 10

Ans:- Cloud Ecosystem Overview:

 Components: The emergence of Internet clouds has created an ecosystem of providers, users, and
technologies.

 Public Cloud Influence: The cloud ecosystem has evolved mainly around public clouds.

 Private & Hybrid Clouds: They are not exclusive, as public clouds play a role in both cloud types.

 Remote Access: Private/hybrid clouds allow access via web service interfaces like Amazon EC2.

Sotomayor's Four Levels of Cloud Ecosystem Development:

 User Level: Consumers need a flexible computing platform.

 Cloud Management Level: The cloud manager offers virtualized resources using an IaaS platform.

 Virtual Infrastructure (VI) Management Level: The VI manager allocates VMs across multiple
server clusters.

 VM Management Level: VM managers control VM operations on individual host machines.

Cloud Tools & Interfaces:

 VI Management Tools: oVirt, vSphere/4, VM Orchestrator for managing resources dynamically.

 Cloud Virtualization Tools: Eucalyptus, Globus Nimbus for virtualizing cloud infrastructure.

 Cloud Access Interfaces: Amazon EC2WS, Nimbus WSRF, ElasticHost REST for accessing cloud
services.
  VM Management & Generation: OpenNebula, VMware vSphere for handling Xen, KVM, and
VMware tools.

7. Explain different service models with a neat diagram 10

Ans:- Different service models

1. Infrastructure-as-a-Service (IaaS)

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.

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.

2. Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS)

o Provides a cloud-based environment for application development and deployment.

o Includes integrated tools such as databases, runtime environments, and development frameworks.

Examples:
o Google App Engine: Scalable application hosting.

o AWS Elastic Beanstalk: Automated deployment and scaling of web applications.

o Microsoft Azure App Services: PaaS offering for .NET, Java, Python, and more.

Benefits:

o Reduces the complexity of software development.

o Allows developers to focus on writing code instead of managing infrastructure.

o Scalable and flexible resource allocation.

Challenges:

o Limited control over the underlying infrastructure.

o Vendor lock-in concerns.

3. Software-as-a-Service (SaaS):

o Delivers software applications over the internet without requiring installation on local machines.

o Users access software via web browsers on a subscription basis.

Examples:

o Google Workspace (Docs, Sheets, Gmail): Cloud-based productivity tools.

o Microsoft 365: Office applications and collaboration tools.

o Salesforce: Customer relationship management (CRM) software.

Benefits:

o Easy access from anywhere with an internet connection.

o Automatic updates and maintenance handled by the provider.

o Lower upfront costs compared to traditional software licensing.

Challenges:

o Data security and privacy concerns.

o Dependence on internet connectivity.

o Limited customization compared toon-premises solutions

8. Explain Data center interconnection networks 10

Ans:- Data-Center Interconnection Networks

A critical core design of a data center is the interconnection network among all servers in the data center
cluster. This network design must meet five special requirements: low latency, high band width, low cost,
message-passing interface (MPI) communication support, and fault tolerance. Specific design
considerations are :

1. Application Traffic Support

 The network topology must support MPI communication patterns.

 Both point-to-point and collective communications must be handled.

 High bisection bandwidth is necessary.

 One-to-many communications support distributed file access.

 Metadata master servers communicate with slave nodes in the cluster.

 The network must efficiently perform MapReduce functions.

2. Network Expandability

 The interconnection must be expandable for thousands of server nodes.

 Network topology should support future growth.

 Load balancing and data movement optimization are essential.

 Fat-tree and crossbar networks can be implemented with Ethernet switches.

 Data-center containers serve as modular building blocks.

 Server containers make installation efficient, needing only power supply, network link, and
cooling.

3. Fault Tolerance & Graceful Degradation

 Redundant paths should be available for fault tolerance.

 Servers achieve reliability via data replication.

 Software & hardware redundancy should mitigate failures.

 Network support software must detect and avoid faulty links.

 Hot-swappable components ensure continuous operation.

 Hierarchical network structure enhances fault resistance.

 4. Switch-Centric Data-Center Design

 Switch-Centric: Uses switches to connect servers, requiring no server modifications.

 Server-Centric: Requires special drivers on servers for data traffic relaying.

 Switches need to be well-structured for optimal connections.

9. Explain data center management issues 5

Ans:- Data-Center Management Issues

• 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.
10. Explain different architectural design of compute and storage clouds 15

Ans:- Different architectural design of compute and storage clouds

1. A Generic Cloud Architecture Design

Cloud Platform Design Goals

 Scalability: Handle increasing loads by adding nodes.

 Virtualization: Abstract physical resources into logical units.

 Efficiency: Optimize resource utilization and power consumption.

 Reliability: Ensure high availability and fault tolerance.

Enabling Technologies

 Broadband internet, multicore CPUs, disk arrays.

 Virtualization for servers, storage, and networks.

 SOA (Service-Oriented Architecture), SaaS, and Web 2.0.

Generic Architecture

 Comprised of clusters of virtual and physical servers.

 Includes storage systems, firewalls, SANs, and security devices.

 Users interact via web interfaces; provisioning tools dynamically allocate resources.
2. Layered Cloud Architectural Development

Three-Layer Architecture:

 Infrastructure Layer (IaaS): Virtualized compute, storage, and network resources.

 Platform Layer (PaaS): Runtime environments, tools for development and testing.

 Application Layer (SaaS): End-user applications (CRM, ERP, office suites).

Market-Oriented Cloud Architecture

 Includes service request examiner, admission control, and autonomic resource manager.

 Dynamically adjusts resources based on market demand and SLAs.

Quality of Service (QoS) Factors

 SLAs enforce requirements like availability, latency, and security.

3. Virtualization Support and Disaster Recovery

Hardware Virtualization

 Uses software to simulate hardware (e.g., Xen, VMware).

 Allows legacy OS and apps to run without modification.

Virtualization in Public Clouds

 AWS: Xen VMs with elastic scaling.

 Azure: Runtime VMs with declarative configuration.
 GAE: Limited to predefined frameworks.

Storage Virtualization

 Green data centers reduce power through consolidated virtual storage.

 Improves scalability and energy efficiency.

IaaS Virtualization

 Users gain full VM access.

 Infrastructure includes virtual disks

Architectural design of compute and storage clouds involves layered architecture, virtualized
infrastructure, dynamic provisioning, and resilient storage systems. By combining these aspects, cloud
platforms deliver scalable, reliable, and cost-efficient services.
11. Explain the cloud computing Architectural Design Challenges 10

Ans:- Architectural Design Challenges in cloud computing:

Challenge 1—Service Availability & Data Lock-in

 Single company management leads to single points of failure.

 Using multiple cloud providers can improve reliability.

 DDoS attacks threaten SaaS availability; scaling up defenses is a solution.

 Proprietary APIs prevent easy data migration; standardized APIs enable interoperability.

Challenge 2—Data Privacy & Security Concerns

 Public cloud exposes systems to attacks like malware, DoS, and hypervisor hijacking.

 Encryption, firewalls, and virtual LANs can improve security.

 Laws require data localization, making compliance a challenge.

Challenge 3—Unpredictable Performance & Bottlenecks

 I/O interference slows VM performance.

 Cloud apps are becoming data-intensive, complicating data placement.

 Bottleneck removal, wider links, and improved architecture can help optimize performance.

Challenge 4—Distributed Storage & Software Bugs

 Databases constantly grow, requiring scalable storage systems.

 Data consistency checks in SAN-connected clouds are critical.

 Debugging in large-scale systems is challenging; VM monitoring and simulators help capture

errors.

Challenge 5—Cloud Scalability, Interoperability & Standardization

 Pay-as-you-go model varies in computation, storage, and bandwidth costs.

 Open Virtualization Format (OVF) supports secure and portable VM distribution.

 Hypervisor-agnostic VMs and cross-platform migration are necessary for seamless

interoperability.

Challenge 6—Software Licensing & Reputation Sharing

 Open-source licensing dominates, but commercial licensing must evolve.

 Bad customer behavior can harm cloud reputation (e.g., EC2 IP blacklisting).
  Reputation-guarding services and legal liability solutions are needed at the SLA level.

12. Explain Architecture of GAE with a neat diagram 10

Ans:- Google App Engine (GAE) Architecture

1. User Interaction Layer

o Users interact with Google applications via web interfaces.

o Third-party providers use GAE to build cloud applications.

2. Google Cloud Infrastructure

o Applications run in Google-managed data centers with thousands of servers forming

clusters.

o Different cloud components work together to process and store data.

3. Core Building Blocks

o Google File System (GFS): Stores large-scale data efficiently.

o MapReduce: Supports large data processing for application development.

o Chubby: Manages distributed locking for resource coordination.

o BigTable: Enables structured and semi-structured data access.

4. GAE Execution Framework

o Scheduler: Manages task execution across cloud servers.

o BigTable Server: Handles large-scale data storage.

o MapReduce Jobs: Processes and analyzes big datasets.

o Scheduler Slave: Assists in executing scheduled tasks efficiently.

o GFS Chunkserver: Stores data chunks managed by the GFS master.

o Linux OS: Provides a stable environment for running cloud applications.

5. GAE as a Platform for Developers

o Runs third-party applications without requiring server maintenance.

o Provides an application framework similar to ASP, J2EE, and JSP.

o Supports Python and Java programming environments.

 o Acts as a dynamic web-serving infrastructure for cloud applications.

13. Explain Architecture of AWS with a neat diagram 10

Ans:- Architecture of AWS (Amazon Web Services)

AWS is a leading cloud service platform that follows the Infrastructure-as-a-Service (IaaS) model to
provide scalable, reliable, and cost-effective computing resources over the internet. Its architecture is
designed to deliver services such as computation, storage, database, networking, and messaging to
developers and businesses.

Key Components of AWS Architecture:

1. EC2 (Elastic Compute Cloud):

o Provides resizable compute capacity.

 o Supports launching virtual machines (instances) on-demand.

o Different instance types are available for various workloads.

2. EBS (Elastic Block Store):

o Block-level storage for use with EC2.

o Persistent and allows backup through snapshots.

3. S3 (Simple Storage Service):

o Object storage designed to store and retrieve any amount of data.

o Provides high durability and availability.

4. SQS (Simple Queue Service):

o A messaging service that ensures reliable communication between distributed components.

5. SimpleDB:

o A NoSQL database to store structured data.

6. CloudWatch:

o Monitoring tool that provides metrics like CPU usage, memory, network traffic, etc.

7. Elastic Load Balancer (ELB):

o Automatically distributes incoming application traffic across multiple EC2 instances.

8. Auto Scaling:

o Automatically adjusts the number of EC2 instances according to traffic demand.

9. Amazon RDS (Relational Database Service):

o Manages SQL-based databases (e.g., MySQL, PostgreSQL).

10. Amazon VPC (Virtual Private Cloud):

o Allows users to provision logically isolated sections of the AWS cloud.

14. Explain Architecture of Microsoft Windows Azure platform with a neat diagram 10

Ans:- Architecture of Microsoft Windows Azure platform

Microsoft Windows Azure (now Microsoft Azure) is a cloud computing platform introduced in 2008. It
provides a range of services to develop, host, and manage applications through Microsoft-managed data
centers. Azure supports IaaS, PaaS, and SaaS models.

Key Components of Azure Architecture:

1. Windows Azure OS:

o Controls and manages hardware resources.

o Handles hosting, load balancing, and scaling.

2. Windows Azure Fabric Controller:

o Acts as the kernel of Azure OS.

o Manages resources like compute, storage, and networking.

o Assigns VMs, monitors health, and recovers from failures.

3. Compute Services:

o Virtual machines are deployed using Web Roles (for web hosting) and Worker Roles (for
background processing).

4. Storage Services:

o Offers structured and unstructured storage (Blobs, Tables, Queues).

o SQL Azure allows access to relational data using familiar SQL Server tools.

5. Development Environment:
 o Supports .NET and other programming frameworks.

o Developers can use Visual Studio and SDKs.

6. Application Services:

o Includes Live Services, SharePoint, Dynamics CRM, Office Live, etc.

7. Communication Protocols:

o Azure uses SOAP and REST to integrate with other platforms and services.

15. Explain Resource Provisioning methods with relevant graphs 10

Ans:- Resource Provisioning methods with relevant graphs :

Resource provisioning in cloud computing refers to allocating computing resources (like VMs, CPU,
memory) based on user demand. There are three main static methods, illustrated in Figure 4.24

1. Peak Load Provisioning (Overprovisioning)

 Description: Resources are provisioned based on the highest possible demand.

 Pros: No user request is denied; full QoS.

 Cons: Heavy resource waste during low usage times.

 Graph: Shows a flat high capacity line with demand curve below it, leading to shaded waste area.

2. Underprovisioning 1

 Description: Resources are provisioned just along average or slightly above-expected demand.
  Pros: Better resource utilization than peak provisioning.

 Cons: Some user demands are unfulfilled, leading to dissatisfaction and provider loss.

 Graph: Capacity line matches middle of demand curve, with shaded area above capacity (lost
demand).

3. Underprovisioning 2 (Fixed Capacity with Declining Demand)

 Description: Constant provisioning of resources regardless of changing demand.

 Pros: Simple to manage.

 Cons: Severe waste when demand declines; users may abandon service.

 Graph: Capacity line is constant, but demand curve declines, creating a large waste area below
capacity.

These static methods are inefficient without elasticity. Both the cloud provider and the user may suffer
from lost revenue and poor experience. Dynamic, adaptive provisioning (like auto-scaling) is preferred in
real-world scenarios.

16. Explain Virtual machine creation and management with a neat diagram 10

Ans:- Virtual machine creation and management

Virtual Machine (VM) creation and management are crucial aspects of cloud computing that enable
flexible, scalable, and efficient resource usage. The VM manager interacts with physical infrastructure or
IaaS platforms to provision, start, stop, or monitor VMs.

1 Independent Service Management

 Services are loosely coupled and execute independently.

 APIs are provided for interacting with services.

 Example: Amazon SQS (Simple Queue Service) supports message-based communication between
services.

 Services can run independently of each other yet communicate via queues.

2 Running Third-Party Applications

 Cloud platforms support applications built by third-party developers.

 APIs are often web service-based (REST, SOAP).

 Examples:

o Google App Engine (GAE) and Microsoft Azure provide their own APIs.

o IBM WebSphere for Java applications on IBM’s Blue Cloud.

o Amazon EC2 runs application engines in VM instances.

3 Virtual Machine Manager

 Acts as a middleware between the cloud gateway and physical/cloud infrastructure.

 Controls lifecycle: create, deploy, start, stop, monitor VMs.

 Supports multiple environments like:

o OpenNebula

o Amazon EC2

o French Grid’5000

 Example: Melbourne group developed such VM managers for various platforms.

4 Virtual Machine Templates

 VM configurations are defined in reusable templates.

 Parameters include OS, CPU, memory size, etc.

  Example templates:

o Ubuntu – 1 core, 128 MB

o Fedora – 2 cores, 256 MB

 Simplifies rapid provisioning and automation.

5 Distributed VM Management

 VM managers can operate across geographically distributed clusters.

 Ensures uniform management across various sites and infrastructures.

 Helps with fault tolerance, scaling, and global cloud management.

17. Explain Inter cloud exchange of cloud resources through brokering with a neat diagram 10

Ans:- Inter cloud exchange of cloud resources through brokering

Cloud Exchange (CEx)

 Acts as a marketplace connecting service consumers and cloud providers.

  Aggregates demand from application brokers.

 Matches them with offers published by cloud coordinators.

Application Broker

 Represents cloud customers.

 Negotiates with cloud coordinators to lease resources based on SLA (Service Level Agreement).

Cloud Coordinator

 Represents cloud infrastructure providers.

 Publishes available resources to the Cloud Exchange.

Enterprise Resource Manager (Proxy)

 Handles resource requests on behalf of users.

Bank and Auctioneer

 Bank handles secure financial transactions.

 Auctioneer enables price negotiation via market models (e.g., auction or commodity market).

Working Mechanism :-

 The Enterprise client sends a resource request.

 The broker interacts with CEx to find suitable offers.

 The auctioneer matches offers with bids.

 Upon successful negotiation, resources are allocated and SLAs enforced.

 Bank system ensures secure payment handling.

Advantages

 Efficient use of global cloud resources.

 Market-driven pricing.

 SLA-based dynamic resource allocation.

 Seamless workload migration and application scaling.