Unit 1

1. Cloud Computing: Definition and Components

 Definition:

o A model that provides on-demand network access to shared configurable computing

resources.

o Delivers computing as a utility (like water or electricity) on a pay-as-you-go basis.

 Key Components:

o Hardware Infrastructure: Servers, storage, and networking devices.

o Virtualization Layer: Hypervisors that create virtual machines.

o Management & Orchestration Software: Tools for resource allocation, monitoring,

and automation.

o Service Delivery Models: Frameworks for offering services (SaaS, PaaS, IaaS).

o Data Centers: Facilities housing the physical infrastructure.

 Diagram: Cloud Computing Components

+------------------+

| End Users |

+--------+---------+

Internet/Network

+--------v--------------+

| Cloud Frontend | <-- Portals, APIs, Web Interfaces

+--------+--------------+

+--------v------------------+

| Cloud Management | <-- Provisioning, Billing, Monitoring

+--------+------------------+

+--------v-----------+

| Infrastructure | <-- Virtualized Servers, Storage, Network

+---------------------+
2. Characteristics, Advantages, and Disadvantages of Cloud Computing

 Characteristics:

o On-Demand Self-Service: Users can provision resources without human

intervention.

o Broad Network Access: Accessible over the Internet from various devices.

o Resource Pooling: Shared multi-tenant model with dynamic resource allocation.

o Rapid Elasticity: Scalable resources that can quickly expand or contract.

o Measured Service: Resource usage is monitored and billed on a pay-per-use basis.

 Advantages:

o Cost Efficiency: Reduces capital expenditure (CapEx) by shifting to operational

expense (OpEx).

o Scalability: Easily adjust resources based on demand.

o Accessibility: Remote access to data and applications.

o Maintenance: Provider handles hardware/software maintenance and updates.

 Disadvantages:

o Security Concerns: Data privacy and compliance issues.

o Downtime: Reliance on internet connectivity can cause availability issues.

o Limited Control: Organizations may have less control over infrastructure.

o Vendor Lock-In: Dependency on specific providers’ technologies and policies.

3. Cloud Service Models: SaaS, PaaS, IaaS, and Storage

 Software as a Service (SaaS):

o Applications delivered over the Internet.

o Examples: Google Workspace, Salesforce.

o Key Points: No need for local installation, automatic updates, subscription-based.

 Platform as a Service (PaaS):

o Provides a platform to develop, test, and deploy applications.

o Examples: Google App Engine, Microsoft Azure App Service.

o Key Points: Offers development tools, middleware, and database management.

 Infrastructure as a Service (IaaS):

 o Virtualized computing resources over the Internet.

o Examples: Amazon EC2, Microsoft Azure Virtual Machines.

o Key Points: Provides storage, servers, and networking; users manage OS and
applications.

 Cloud Storage Services:

o Scalable storage solutions accessed over the Internet.

o Examples: Amazon S3, Google Cloud Storage.

o Key Points: Highly durable, redundant, accessible anytime; billed based on usage.

4. Cloud Computing Reference Model

 Layers of the Reference Model:

o Hardware/Infrastructure Layer: Physical servers, storage, network.

o Virtualization Layer: Abstracts hardware resources.

o Resource Management Layer: Manages provisioning, load balancing, and scaling.

o Service Layer: Offers SaaS, PaaS, and IaaS.

o User/Client Layer: Interfaces for end users (web portals, mobile apps).

 Diagram: Cloud Reference Model

+--------------------+

| User/Client | <-- End User Applications

+---------+----------+

+---------v----------+

| Service Layer | <-- SaaS, PaaS, IaaS

+---------+----------+

+---------v----------+

| Resource Management| <-- Provisioning, Monitoring

+---------+----------+

+---------v----------+

| Virtualization Layer| <-- Hypervisors, Containers

 +---------+----------+

+---------v------------------+

| Hardware/Infra Layer| <-- Servers, Storage, Networks

+-----------------------------+

5. Cloud Deployment Models

 Public Cloud:

o Description: Services offered over the public internet by third-party providers.

o Advantages: Economies of scale, cost-effective, high scalability.

o Examples: AWS, Microsoft Azure, Google Cloud.

 Private Cloud:

o Description: Cloud infrastructure dedicated to a single organization.

o Advantages: Greater control, enhanced security.

o Examples: On-premises data centers or hosted private clouds.

 Hybrid Cloud:

o Description: Combination of public and private clouds.

o Advantages: Flexibility, optimized resource usage, balancing security and scalability.

 Community Cloud:

o Description: Shared by several organizations with common requirements.

o Advantages: Cost-sharing, tailored security and compliance.

 Diagram: Cloud Deployment Models

+---------------------+

| Public Cloud |

+---------------------+

+---------------------+

| Hybrid Cloud | <-- Combines Public & Private

+---------------------+

^
 |

+---------------------+

| Private Cloud |

+-----------------------+

+----------------------------+

| Community Cloud |

+----------------------------+

6. Identity Management as a Service (IDaaS)

 Definition:

o Cloud-based solutions for identity and access management.

 Key Features:

o Single Sign-On (SSO): One login for multiple applications.

o Multi-Factor Authentication (MFA): Additional security layers.

o User Provisioning/De-provisioning: Automated account management.

o Federated Identity: Integration with external identity providers.

 Benefits:

o Reduced IT overhead, enhanced security, centralized management.

7. Cloud Storage Providers

 Major Providers:

o Amazon S3: Scalable object storage, high durability.

o Google Cloud Storage: Unified object storage, low latency.

o Microsoft Azure Blob Storage: Secure, cost-effective storage for unstructured data.

o IBM Cloud Object Storage: High availability and reliability.

 Common Attributes:

o Scalability, data replication, accessibility via APIs, pay-per-use pricing.

8. Cloud System Architecture

  Key Architectural Layers:

o Hardware Layer: Physical servers, storage, networking.

o Virtualization Layer: Abstracts physical resources into virtual machines or containers.

o Middleware/Management Layer: Orchestrates resource provisioning, load

balancing, and scaling.

o Application Layer: Provides cloud services (SaaS, PaaS, IaaS) to end users.

 Key Characteristics:

o Multi-Tenancy: Sharing resources among multiple users.

o Fault Tolerance: Data replication and failover mechanisms.

o Scalability: Dynamic resource allocation based on demand.

9. Cloud Economics

 Cost Efficiency:

o Shifts from capital expenditure (CapEx) to operational expenditure (OpEx).

o Pay-as-You-Go: Only pay for the resources you use.

 Return on Investment (ROI):

o Lower upfront costs and flexible scaling drive quicker ROI.

 Total Cost of Ownership (TCO):

o Reduced IT maintenance and hardware replacement costs.

 Economies of Scale:

o Providers benefit from large-scale operations and pass savings to customers.

 Challenges:

o Unpredictable costs if not properly managed; potential hidden fees.

10. Involvement of Cloud Computing in Organizations

 Integration Types:

o Core Business Applications: CRM, ERP, and SCM systems hosted in the cloud.

o Collaboration & Communication: Cloud-based email, file sharing, and conferencing.

o Development & Testing: PaaS/IaaS environments for rapid application development.

o Disaster Recovery & Backup: Cloud storage for data recovery.

 Benefits:
 o Enhanced agility, cost savings, and improved scalability.

 Considerations:

o Security policies, compliance regulations, and vendor lock-in risks.

11. Role of Networking in Cloud Computing

 Connectivity:

o Provides the backbone for accessing cloud services over the Internet.

 Performance:

o Load Balancing: Distributes traffic evenly.

o Latency Reduction: Optimizes data transfer routes.

 Security:

o Implements firewalls, VPNs, and encryption to protect data in transit.

 Virtual Networking:

o Software-Defined Networking (SDN) for dynamic resource allocation and network

management.

12. Seven-Step Model of Migration into the Cloud

1. Assessment:

o Evaluate existing infrastructure, applications, and business needs.

2. Planning:

o Define migration strategy, timelines, and risk management.

3. Pilot Testing:

o Migrate a small set of applications to validate the process.

4. Migration:

o Gradually transfer data and applications to the cloud.

5. Integration:

o Ensure seamless connectivity between on-premises and cloud systems.

6. Optimization:

o Fine-tune performance, cost, and security post-migration.

7. Governance & Management:

o Monitor, manage, and continuously improve cloud operations.

 Unit-2:- Cloud File Systems and Data Storage
1. Cloud File System with Architectures

 Definition:

o Distributed file systems designed to store, manage, and access large volumes of data
over clusters.

 Key Concepts:

o Distributed Metadata: Centralized management of file metadata.

o Fault Tolerance: Data replication and error recovery.

o Scalability: Horizontal scaling to support large data sets.

 Example Architectures:

o Google File System (GFS) / HDFS:

 Master/NameNode: Manages metadata.

 Chunk/Data Nodes: Store actual data blocks.

 Diagram: Simplified Cloud File System

+-------------+

| Client |

+------+------+

+------v-----------+

| Master/ | <-- Metadata Management

| NameNode |

+------+-----------+

+------v----------+

| Chunk/Data | <-- Data Storage Nodes

| Nodes |

+------------------+

2. HBase Data Model

 Overview:
 o A NoSQL, column-oriented database built on top of HDFS.

 Key Characteristics:

o Tables: Consist of rows and column families.

o Column Families: Group related columns stored together.

o Schema Flexibility: Sparse data storage; columns can vary per row.

o Versioning: Stores multiple versions of a cell’s data.

 Data Structure:

o Row Key: Unique identifier for rows.

o Column Qualifier: Specifies the column within a family.

o Timestamp: Used for version control.

3. Cloud File System GFS/HDFS

 Google File System (GFS):

o Architecture: Master-slave design; master handles metadata, slaves (chunk servers)

store data.

o Fault Tolerance: Data replicated across multiple servers.

 Hadoop Distributed File System (HDFS):

o Architecture: NameNode (metadata) and DataNodes (data blocks).

o Block Replication: Typically three copies for reliability.

 Diagram: GFS/HDFS Architecture

+----------------+

| Client |

+-------+--------+

+-------v------------------+

| Master/NameNode|

+-------+-------------------+

+---------v-----------------+

| Data/Chunk Nodes |

+----------------------------+
4. Working of Cloud Data Store

 Functionality:

o Provides scalable, high-availability storage for structured/unstructured data.

 Mechanisms:

o Replication: Data is copied across nodes.

o Sharding: Data is partitioned to distribute load.

o Consistency Models: Options range from strong to eventual consistency.

 Applications:

o Often used for NoSQL databases and large-scale data processing.

5. Differences: Cloud File System vs. Normal File System

6. AFS (Andrew File System) Architecture

 Overview:

o A distributed file system that supports large-scale file sharing.

 Key Features:

o Client-Server Model: Clients cache files locally.

 o Volumes: Logical grouping of files that can be moved between servers.

o Cache Consistency: Ensures local caches are updated.

o Security: Built-in authentication and access control.

7. Difference Between SAN and NAS

8. HDFS Architecture in Detail

 Components:

o NameNode: Centralized metadata manager.

o DataNodes: Store actual data blocks.

o Secondary NameNode: Assists in checkpointing metadata (not a backup for

NameNode).

 Key Features:

o Block Storage: Files broken into blocks (default 128 MB).

o Replication: Each block is replicated (default factor is 3).

o Fault Tolerance: Automatic re-replication of lost blocks.

 Diagram: HDFS Architecture

+----------------+

| Client |

+-------+--------+
 |

+-------v--------+

| NameNode| <-- Metadata

+-------+--------+

+-------v--------+

| DataNodes | <-- Data Blocks

+----------------+

9. Data Storage Types: EDS, DAS, SAN, and NAS

 Enterprise Data Storage (EDS):

o Overview: Centralized, enterprise-grade storage solutions.

o Characteristics: High capacity, performance, and reliability.

 Direct Attached Storage (DAS):

o Overview: Storage directly attached to a computer or server.

o Characteristics: Fast access but limited scalability and sharing.

 Storage Area Network (SAN):

o Overview: Dedicated high-speed network that provides block-level storage.

o Characteristics: High performance, supports mission-critical applications.

 Network Attached Storage (NAS):

o Overview: File-level storage accessed over standard networks.

o Characteristics: Easy file sharing, scalable file storage.

10. Data Intensive Technologies for Cloud Computing

 MapReduce Framework:

o Purpose: Processes large datasets in parallel.

o Key Elements: Map tasks (processing) and Reduce tasks (aggregation).

 Hadoop Ecosystem:

o Includes: HDFS, YARN, HBase, Spark.

o Purpose: Enables scalable data storage and processing.

 NoSQL Databases:
 o Examples: Cassandra, MongoDB, HBase.

o Purpose: Handle unstructured and semi-structured data with high throughput.

11. Distributed Data Storage

 Concept:

o Data is stored across multiple nodes/servers.

 Advantages:

o Scalability: Easily add more nodes.

o Fault Tolerance: Data remains available despite node failures.

o Performance: Parallel access improves data retrieval speeds.

 Techniques:

o Sharding: Partitioning data into smaller chunks.

o Replication: Duplicating data across different nodes.

o Consistency Models: Manage data accuracy across distributed systems.