MODULE 5

1.Explain the architecture and core components of Google app engine.

Google AppEngine is a Platform-as-a-Service (PaaS) offering that facilitates the development

and hosting of scalable web applications. It leverages Google's distributed infrastructure to
handle high traffic, allocating resources dynamically to meet the demand. It supports multiple
programming languages, including Java, Python, and Go. Here's a breakdown of its
architecture and core concepts:
1. Infrastructure
• Purpose: AppEngine hosts web applications and efficiently handles user requests.
• Functionality:
o AppEngine uses Google’s data centers and many servers to process requests.
o When a request is made, AppEngine identifies the servers that process it,
evaluates their load, and, if necessary, redirects or allocates additional
resources.
o The system doesn't expect state maintenance between requests, simplifying
load balancing.
2. Runtime Environment
• Purpose: The runtime environment provides the execution context for hosted
applications.
• Sandboxing:
o Apps are isolated in a sandbox environment to ensure security.
 o It restricts potentially harmful operations like file system access or long-
running processes.
• Supported Languages:
o Java: Java 6 and Java 5 (via SDK), supports JSP and servlets.
o Python: Uses Python 2.5.2, with an optimized environment.
o Go: Supports Go 1.3, allowing applications to run within AppEngine.
3. Storage
• Types of Storage:
o Static File Servers: For non-dynamic files like CSS, HTML, images, etc.
o DataStore: A scalable object database for semi-structured data, built on
Bigtable, where entities are stored with a key and properties. DataStore is
optimized for fast access and supports transactions, though with some
limitations for scalability.
o MemCache: An in-memory cache for frequently accessed data, reducing
access times for common objects.
4. Application Services
• UrlFetch: Allows applications to fetch resources from external HTTP/HTTPS
endpoints, both synchronously and asynchronously.
• MemCache: A distributed in-memory cache for objects frequently accessed,
enhancing application performance.
• Mail and XMPP:
o Mail allows sending emails and attachments asynchronously.
o XMPP enables chat integration with services like Google Talk.
• Account Management: Allows integration with Google Accounts for user
authentication and profile management.
• Image Manipulation: Supports lightweight image operations like resizing, rotating,
and enhancing images.
5. Compute Services
• Task Queues:
o Enables long-running tasks that can't be handled in the response time of a
request.
o Tasks are submitted for execution at a later time, with automatic retry if failed.
• Cron Jobs:
 o Allows scheduling tasks to run at specific times, such as maintenance tasks or
periodic notifications.

2.Discuss in detail the following media applications of cloud computing technologies

i) Animoto

• A popular cloud-based platform for creating videos using images, music, and video
fragments.
• Users upload photos and videos, select themes, and the service’s AI engine
automatically applies animation and transition effects.
• It uses Amazon Web Services (AWS) infrastructure:
➢ EC2 for web front-end and worker nodes.
➢ S3 for storage of media files.
➢ SQS for managing rendering tasks through a queue system.
• It uses auto-scaling capabilities managed by Rightscale to dynamically scale based on
demand.
• Handles up to 4,000 servers during peak times, ensuring scalability and reliability.

ii) Maya Rendering with Aneka

• Used in engineering and movie production for rendering 3D models.

 • The GoFront Group in China Southern Railway uses a private cloud solution for 3D
rendering of train designs.
• Aneka manages the private cloud network, turning desktops into a distributed
computing system.
• Rendering tasks are distributed across available machines, allowing the Maya
renderer to execute and gather results efficiently.
• The system reduces rendering time from days to hours by utilizing off-peak desktop
hours.
• The system handles tasks like the number of frames and cameras required for
rendering, optimizing computational resources.

iii)Video encoding on cloud.

• Encoding.com provides cloud-based video transcoding services, converting videos

into various formats suitable for different devices.
• It integrates with Amazon Web Services (AWS) (EC2, S3, CloudFront) and
Rackspace for computing power and storage.
• Users can upload videos, specify the destination format, and receive the converted
video.
• Additional features include adding watermarks, logos, and audio/image conversion.
• Provides different pricing models: monthly, pay-as-you-go, and high-volume rates.
• Used by over 2,000 customers and has processed more than 10 million videos,
showing its scalability and success.
3.Explain in detail about the application of cloud computing in
i)Healthcare: ECG analysis in the cloud

• Cloud computing is used in healthcare for diagnostics, monitoring, and analysis.

• ECG analysis on the cloud allows efficient and remote health monitoring.
• Wearable devices with ECG sensors monitor the patient’s heart activity.
• Data is transmitted to a mobile device and then to a cloud-hosted web service.
• The cloud uses SaaS (web service for data storage), PaaS (runtime platform for
processing), and IaaS (infrastructure for task execution).
• Heartbeat data is processed to extract waveforms and detect anomalies by comparing
them with a reference waveform.
• Doctors and first-aid personnel are notified if anomalies are detected.
• Elasticity of cloud infrastructure reduces the need for large hospital investments.
• Cloud services are accessible from any internet-connected device, enabling
integration with hospital systems.
• Pay-per-use and volume-based pricing models save costs.
• Enables continuous patient monitoring without frequent hospital visits.
• Provides faster diagnoses and immediate notifications in critical cases.
• Improves resource utilization and reduces operational costs for hospitals.
ii)Geoscience: satellite image processing

• Geoscience applications involve massive data collection, production, and analysis.

• Geographic Information Systems (GIS) manage geospatial data for advanced farming,
civil security, and resource management.
• Satellite remote sensing generates vast amounts of raw images requiring intensive
processing.
• Cloud computing supports the processing of satellite images by providing scalable
infrastructure.
• Images are transferred from local storage to cloud facilities for transformations and
corrections.
• A cloud-based implementation in India integrates services for geocoding, data
visualization, and image processing.
• SaaS provides tools for GIS tasks, while PaaS handles data importing and processing.
• Aneka and Xen private cloud enable dynamic provisioning of resources.
• Cloud computing reduces workload on local systems and provides elasticity for
computing needs.
• Supports efficient extraction of meaningful geospatial information for decision-
making.
iii) Biology: protein structure prediction

• Biology applications require high computational power and often operate on large
datasets.
• Protein structure prediction is a critical task in life sciences, especially for drug
design.
• It involves complex computations to identify the protein structure with minimal
energy, requiring exploration of vast state spaces.
• Cloud computing provides scalable computational power on demand, eliminating the
need for dedicated clusters or bureaucratic processes.
• Jeeva is a project utilizing cloud technology for protein structure prediction, offering a
web portal for scientists.
• The prediction uses machine learning (support vector machines) to classify protein
structures into secondary classes (E, H, C).
• The classification phase leverages parallel execution, significantly reducing
computation time.
• The task is translated into a task graph, submitted to Aneka cloud middleware for
processing.
• Results are made available through the web portal for visualization.
• Cloud advantages include scalability (grow/shrink on demand), pay-per-use pricing,
and ease of offering the service dynamically.
iv) Biology: gene expression data analysis for cancer diagnosis

• Gene expression profiling measures the expression levels of thousands of genes

simultaneously.
• It helps understand biological processes triggered by medical treatments at a cellular
level.
• It is a critical part of drug design, allowing scientists to analyze the effects of
treatments.
• It is used in cancer diagnosis and treatment, classifying tumors based on gene
expression data.
• Cancer involves uncontrolled cell growth caused by mutated genes, and profiling
identifies these mutations.
• Classifying gene expression data is challenging due to high dimensionality and
limited sample sizes.
• Classifiers like the eXtended Classifier System (XCS) help classify large
bioinformatics datasets.
• CoXCS, a variation of XCS, handles high-dimensional datasets by dividing the search
space into subdomains.
• CoXCS parallelizes computations, as classification in subdomains can occur
concurrently.
• Cloud-CoXCS implements CoXCS on the cloud using Aneka, solving classification
problems in parallel.
• Strategies in Cloud-CoXCS define how outcomes are composed and whether
iterations are required.
4.Explain Amazon web services(AWS) in detail.
AWS is a cloud computing platform that provides a broad range of scalable, flexible, and
cost-effective services. It enables users to build and manage applications with features like
infrastructure scalability, messaging, and data storage. Accessible through web interfaces
(SOAP/REST) and a user-friendly console, AWS operates on a pay-as-you-go model,
allowing businesses to efficiently manage resources and expenses. It supports raw compute
power, storage, networking, data management, application deployment, and advanced
services to cater to diverse application needs.

1.Compute Services
• Core component of cloud systems.
• Amazon EC2 is a key service offering IaaS, enabling deployment of virtual servers
(instances) using images.
• Users can configure instances (e.g., memory, CPU, storage) and access them
remotely.
Amazon Machine Images (AMIs)
• AMIs are templates for creating virtual machines.
• Stored in Amazon S3 with unique identifiers (e.g., ami-xxxxxx).
• Contain OS and predefined file system layouts.
 • Can be created from scratch or bundled from existing instances.
• Stored AMIs can be private or shared, with optional product code association for
revenue.
EC2 Instances
• Virtual machines created from AMIs with customizable configurations.
• Compute power is defined using EC2 Compute Units (ECUs), ensuring consistent
performance over hardware upgrades.
• Instance categories include:
o Standard Instances: General-purpose configurations.
o Micro Instances: Low resources, suitable for small applications with
occasional workload surges.
o High-Memory Instances: Large memory for high-traffic web apps.
o High-CPU Instances: Compute-intensive applications.
o Cluster Compute Instances: High CPU, memory, and I/O for HPC.
o Cluster GPU Instances: For graphics-heavy or GPU-compute tasks (e.g.,
rendering clusters).
EC2 Environment
• Provides essential services (e.g., address allocation, storage attachment, security).
• Instances have internal IPs for internal communication and Elastic IPs for external
accessibility.
• Elastic IPs support failover and remapping between instances.
• EC2 instances have domain names based on IP and availability zone.
Advanced Compute Services
• AWS CloudFormation:
o Facilitates complex deployments using JSON templates to define resource
dependencies.
o Integrates EC2 with other AWS services like S3, Route 53, etc.
• AWS Elastic Beanstalk:
o Simplifies deployment and management of web applications.
o Automates provisioning while allowing control over underlying EC2
infrastructure.
 o Focuses on application deployment, unlike CloudFormation, which handles
infrastructure setup.
• Amazon Elastic MapReduce (EMR):
o Cloud platform for running Hadoop-based MapReduce applications.
o Utilizes EC2 for computing and S3 for storage.
o Supports tools like Pig and Hive, and provides elasticity for cluster
management.

2. Storage Services
• Amazon S3 (Simple Storage Service):
o Object-based storage system with components like buckets (containers) and
objects (stored data).
o Supports REST APIs for operations like PUT, GET, and DELETE.
o Features include metadata tagging, access control policies (ACPs), and
advanced features like logging and BitTorrent integration.
o Immutability and eventual consistency for stored objects.
• Amazon EBS (Elastic Block Store):
o Block storage for EC2 instances with persistent data storage.
o Provides features like snapshots, cloning, resizing, and cross-zone
connectivity.
o Pricing based on allocated storage and I/O requests.
• Amazon Glacier:
o Low-cost archival storage for infrequent data access.
o Retrieval modes: Expedited, Standard, and Bulk, with varying speeds and
costs.
• Amazon ElastiCache:
o In-memory caching service compatible with Memcached for fast data access.
o Offers dynamic cluster resizing, automation, and compatibility with EC2
instances.
• Structured Storage Services:
o Preconfigured EC2 AMIs: Custom DBMS configurations like MySQL,
Oracle, and PostgreSQL.
 o Amazon RDS: Managed relational database with features like backups, Multi-
AZ deployment, and read replicas.
o Amazon SimpleDB: Semi-structured data storage with flexible querying and
eventual consistency.
• Amazon CloudFront:
o Content Delivery Network (CDN) for global distribution of static and
streaming content.
o Uses globally distributed edge servers for low-latency delivery.
o Supports protocol-based restrictions and content invalidation.

3. Communication Services
• Amazon VPC (Virtual Private Cloud):
o Enables the creation of isolated virtual networks within AWS.
o Supports public, private, or hybrid networking setups.
• Amazon Direct Connect:
o Provides dedicated, high-bandwidth connectivity between on-premises systems
and AWS.
• Amazon Route 53:
o Dynamic DNS service to map domain names to AWS resources with high
reliability.
o Supports hosted zones and query management.
• Messaging Services:
o Amazon SQS (Simple Queue Service): For decoupling application
communication via message queues.
o Amazon SNS (Simple Notification Service): Publish-subscribe system for real-
time notifications.
o Amazon SES (Simple Email Service): Scalable email service for reliable email
delivery.

4. Additional Services
• Amazon CloudWatch:
o Monitoring service providing performance statistics for AWS resources.
 o Assists in application optimization and cost management.
• Amazon FPS (Flexible Payment Service):
o Payment infrastructure for selling goods and services, supporting one-time,
periodic, and aggregated payments.

5.Explain EC2 environment

• EC2 instances run in a virtual environment that allocates resources like addresses,
storage, and security settings for hosting applications.
• By default, EC2 instances are assigned internal IP addresses, enabling communication
within the EC2 network and allowing them to access the Internet as clients.
• Instances can be associated with an Elastic IP (EIP). EIPs are static IP addresses that
can be remapped to different instances, enabling high availability and failover
capabilities.
• Each EC2 instance receives an external IP address and a domain name, typically in
the format ec2-xxx.xxx.xxx.compute-x.amazonaws.com, where the domain name
includes the instance's external IP and availability zone information.
• EC2 instances can be deployed in different availability zones, with options in regions
like the United States (Virginia, Northern California), Europe (Ireland), and Asia
Pacific (Singapore, Tokyo). Pricing may vary by zone.
• Instance owners can control where to deploy instances and configure the security of
the instances, ensuring access and network availability according to their needs.
• A key pair (public and private keys) can be associated with an instance during
creation. This allows the owner to securely connect to the instance and access it with
root privileges.
• EC2 uses a basic firewall configuration to control accessibility. Security groups,
which can be attached to instances, allow specifying rules for source addresses, ports,
and protocols (TCP, UDP, ICMP). Instances can be part of multiple security groups.
• While security groups manage external access, internal security configurations within
the instance itself are also essential for comprehensive protection.

6.Expalin Computing services

1.Compute Services
• Core component of cloud systems.
• Amazon EC2 is a key service offering IaaS, enabling deployment of virtual servers
(instances) using images.
• Users can configure instances (e.g., memory, CPU, storage) and access them
remotely.
Amazon Machine Images (AMIs)
• AMIs are templates for creating virtual machines.
• Stored in Amazon S3 with unique identifiers (e.g., ami-xxxxxx).
• Contain OS and predefined file system layouts.
• Can be created from scratch or bundled from existing instances.
• Stored AMIs can be private or shared, with optional product code association for
revenue.
EC2 Instances
• Virtual machines created from AMIs with customizable configurations.
• Compute power is defined using EC2 Compute Units (ECUs), ensuring consistent
performance over hardware upgrades.
• Instance categories include:
o Standard Instances: General-purpose configurations.
o Micro Instances: Low resources, suitable for small applications with
occasional workload surges.
o High-Memory Instances: Large memory for high-traffic web apps.
o High-CPU Instances: Compute-intensive applications.
o Cluster Compute Instances: High CPU, memory, and I/O for HPC.
o Cluster GPU Instances: For graphics-heavy or GPU-compute tasks (e.g.,
rendering clusters).
EC2 Environment
• Provides essential services (e.g., address allocation, storage attachment, security).
• Instances have internal IPs for internal communication and Elastic IPs for external
accessibility.
• Elastic IPs support failover and remapping between instances.
• EC2 instances have domain names based on IP and availability zone.
Advanced Compute Services
• AWS CloudFormation:
o Facilitates complex deployments using JSON templates to define resource
dependencies.
o Integrates EC2 with other AWS services like S3, Route 53, etc.
 • AWS Elastic Beanstalk:
o Simplifies deployment and management of web applications.
o Automates provisioning while allowing control over underlying EC2
infrastructure.
o Focuses on application deployment, unlike CloudFormation, which handles
infrastructure setup.
• Amazon Elastic MapReduce (EMR):
o Cloud platform for running Hadoop-based MapReduce applications.
o Utilizes EC2 for computing and S3 for storage.
o Supports tools like Pig and Hive, and provides elasticity for cluster
management.

7.Explain S3 key concepts

• S3 Design: S3 is designed to offer simple storage through a REST interface,
resembling a distributed file system but with key differences.
• Storage Organization: S3 uses a two-level hierarchy: storage is organized in buckets
that cannot be nested or partitioned, though logical groupings can be simulated by
naming objects appropriately.
• Object Immutability: Objects cannot be renamed, modified, or relocated once
stored. To change an object, it must be removed and added again.
• Eventual Consistency: S3 provides an eventually consistent data store, meaning
changes (especially large ones) may not be instantly reflected across the global
network.
• Request Failures: Requests may occasionally fail due to the large distributed nature
of S3's infrastructure.
• Access: Access is granted through HTTP requests (GET, PUT, DELETE, HEAD, and
POST) depending on the operation, with PUT/POST adding content, GET/HEAD
retrieving content, and DELETE removing elements.
• Naming Buckets and Objects:
o Buckets can be referenced in three ways: canonical form, subdomain form,
and virtual hosting form.
o Objects are referenced by URI paths, and naming conventions allow for
logical groupings, even without a file system structure.
 • Buckets: Buckets are top-level containers for objects. They are created in specific
geographic locations, and objects in a bucket are stored in the same availability zone.
Once created, buckets cannot be renamed or relocated.
• Objects: Objects store the content and are identified by a unique name within the
bucket. Objects cannot be modified after creation; the maximum size is 5 GB.
• Metadata: Objects can have user-defined metadata, which are stored as key-value
pairs with up to 2 KB of data.
• Access Control and Security: Access is controlled through Access Control Policies
(ACPs) and permissions such as READ, WRITE, and FULL_CONTROL. Grantees
can be specific users or groups, with the ability to modify the ACP using GET/PUT
requests.
• Advanced Features:
o Server Access Logging: Tracks detailed information about requests made to
the bucket and its objects.
o BitTorrent Integration: S3 objects can be exposed to the BitTorrent network
for file sharing.
8.Explain Storage services
• AWS provides a collection of services for data storage and information management.
• The core service in this area is represented by Amazon Simple Storage Service (S3).
• This is a distributed object store that allows users to store information in different formats.
• Amazon S3 (Simple Storage Service):
o Object-based storage system with components like buckets (containers) and
objects (stored data).
o Supports REST APIs for operations like PUT, GET, and DELETE.
o Features include metadata tagging, access control policies (ACPs), and
advanced features like logging and BitTorrent integration.
o Immutability and eventual consistency for stored objects.
• Amazon EBS (Elastic Block Store):
o Block storage for EC2 instances with persistent data storage.
o Provides features like snapshots, cloning, resizing, and cross-zone
connectivity.
o Pricing based on allocated storage and I/O requests.
• Amazon Glacier:
 o Low-cost archival storage for infrequent data access.
o Retrieval modes: Expedited, Standard, and Bulk, with varying speeds and
costs.
• Amazon ElastiCache:
o In-memory caching service compatible with Memcached for fast data access.
o Offers dynamic cluster resizing, automation, and compatibility with EC2
instances.
• Structured Storage Services:
o Preconfigured EC2 AMIs: Custom DBMS configurations like MySQL,
Oracle, and PostgreSQL.
o Amazon RDS: Managed relational database with features like backups, Multi-
AZ deployment, and read replicas.
o Amazon SimpleDB: Semi-structured data storage with flexible querying and
eventual consistency.
• Amazon CloudFront:
o Content Delivery Network (CDN) for global distribution of static and
streaming content.
o Uses globally distributed edge servers for low-latency delivery.
o Supports protocol-based restrictions and content invalidation.

9.Explain Communication services

• Amazon VPC (Virtual Private Cloud):
o Enables the creation of isolated virtual networks within AWS.
o Supports public, private, or hybrid networking setups.
• Amazon Direct Connect:
o Provides dedicated, high-bandwidth connectivity between on-premises systems
and AWS.
• Amazon Route 53:
o Dynamic DNS service to map domain names to AWS resources with high
reliability.
o Supports hosted zones and query management.
• Messaging Services:
 o Amazon SQS (Simple Queue Service): For decoupling application
communication via message queues.
o Amazon SNS (Simple Notification Service): Publish-subscribe system for real-
time notifications.
o Amazon SES (Simple Email Service): Scalable email service for reliable email
delivery.
10.Explain the Application Life Cycle
• Development and Testing:
o Developers start building applications on a local development server.
o The server simulates the AppEngine runtime and provides a mock
implementation of services like DataStore, MemCache, and UrlFetch.
o It helps developers profile the application behavior, especially with regards to
DataStore service queries.
o The server traces queries during testing to generate information on required
indexes.
• Java SDK:
o Supports building apps with Java 5 and Java 6 runtime environments.
o Can be used with Eclipse by installing the AppEngine plugin.
o Helps develop applications using the servlet abstraction, which includes other
features like creating web apps with Eclipse Web Platform.
• Python SDK:
o Used for building web applications with Python 2.5.
o Offers the GoogleAppEngineLauncher tool for managing and deploying apps
locally.
o Provides access to services like logs, the SDK console, and application
dashboard.
o Integrates webapp framework for easier development.
o Command-line tools offer more advanced operations for development.
• Application Deployment and Management:
o Applications are deployed after development and testing using simple tools or
command-line operations.
o Developers must create an application identifier to uniquely identify the app
(e.g., http://application-id.appspot.com).
 o After deployment, the app is fully managed by AppEngine, with developers
using the administrative console for monitoring resource usage and managing
versions.
o It’s possible to map custom DNS names to the application for commercial use.
11.Business and Consumer Applications
Business and Consumer Sector Benefits from Cloud Computing:
• Cloud computing helps transform capital costs into operational costs, making it
attractive for IT-centric enterprises.
• The ubiquity of cloud access to data and services benefits end users as well.
• Elastic nature of Cloud technologies allows quick scaling of ideas into products and
services without huge upfront investments.
Cloud Computing Applications:
• Cloud computing is used for a variety of applications such as CRM (Customer
Relationship Management), ERP (Enterprise Resource Planning), productivity, and
social networking apps.
CRM and ERP in the Cloud:
• CRM Applications:
o Cloud CRM applications help small enterprises and start-ups with functional
CRM software without high upfront costs through subscription models.
o CRM is easily moved to the Cloud due to its less specific needs and ability to
access business and customer data anywhere.
• ERP Applications:
o ERP systems integrate various enterprise functions like finance, HR,
manufacturing, and supply chain management.
o Cloud ERP solutions are less mature but growing, with competition from well-
established in-house solutions.
Salesforce.com:
• Salesforce.com is a popular Cloud CRM solution with over 100,000 customers.
• Offers customizable CRM solutions integrated with third-party features, based on the
Force.com Cloud development platform.
• Features include metadata cache, bulk processing, query optimizers, full-text search
engine, and multi-tenant support.
Microsoft Dynamics CRM:
• Offered both on-premises and as an online solution with monthly per-user
subscription.
• Hosted in Microsoft’s global data centers with a 99.9% SLA.
• Supports marketing, sales, and advanced CRM functionalities, accessible via web
browsers or web services (SOAP/REST).
• Easily integrates with other Microsoft products and allows extension through plug-
ins.
• Leverages Windows Azure for additional development and integration.
NetSuite:
• Offers applications for managing various business aspects: ERP, CRM, and e-
commerce.
• Powered by two data centers with redundant links, ensuring 99.5% uptime.
• Provides an integrated all-in-one solution (NetSuite One World) and infrastructure for
developing customized applications.
 • The NetSuite Business Operating System (NS-BOS) supports building SaaS business
applications.
• SuiteFlex enables custom integration of features into new web applications, which are
then distributed via SuiteBundler.
12.Multiplayer Online Gaming
• Online Multiplayer Gaming:
o Attracts millions of global gamers who interact in a shared virtual
environment.
o Unlike traditional LAN, multiplayer online games support hundreds of players
in the same session.
o Architecture based on game log processing facilitates interactions:
▪ Players send updates to the game server.
▪ The server integrates updates into a log accessible to all players via
TCP port.
▪ Game clients read the log and update the local user interface with
actions of other players.
• Game Log Processing:
o Used for tracking players' actions and building statistics (e.g., player ranks).
o Provides additional value to online gaming portals and attracts more players.
o Game log processing can be compute-intensive, especially with many players
and games.
o Web-based gaming portals face "spiky" user behavior, leading to volatile
workloads.
• Cloud Computing for Online Gaming:
o Cloud computing provides elasticity to handle fluctuating workloads, ensuring
seamless game log processing.
o Cloud infrastructure can scale as required based on the number of players and
games.
o Titan Inc. (now Xfire) implemented a cloud-based game log processing
prototype:
▪ Used a private cloud deployment to offload game log processing.
▪ Enabled concurrent processing of multiple logs and supported a larger
number of users.