Cloud Computing (AL-604 (A)) - Unit 1 Short Notes

1.0 Introduction & Core Concepts

1.1 Definition and Evolution of Cloud Computing

• Definition (NIST): A model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.

• Core Characteristics (NIST):

  1. On-Demand Self-Service: Provision resources automatically without human interaction.

  2. Broad Network Access: Available over the network via standard mechanisms.

  3. Resource Pooling: Multi-tenant model with pooled physical/virtual resources.

  4. Rapid Elasticity: Resources can be scaled outward/inward quickly.

  5. Measured Service: Resource usage monitored, controlled, and billed.

• Evolution:

  • Cluster Computing: Group of loosely coupled computers working together.

  • Grid Computing: Distributed computing for large-scale, collaborative tasks (e.g., SETI@home).

  • Utility Computing: Pay-per-use model for computing resources (precursor to cloud billing).

  • Cloud Computing: Integrated service model (IaaS/PaaS/SaaS) with full abstraction and management.

[!TIP] Exam Focus: "Computing on Demand" is synonymous with On-Demand Self-Service. Be ready to explain how it enables dynamic provisioning via APIs/portals.

1.2 Grid Computing vs. Cloud Computing

Similarities: Both use distributed resources, aim for high utilization, and support parallel processing.

[!TIP] Common Pitfall: Grid focuses on collaborative problem-solving across organizations; Cloud focuses on service delivery to end-users/applications.

2.0 Cloud Service Models (SPI Model)

2.1 Infrastructure as a Service (IaaS)

• Definition: Provides fundamental computing resources—processing, storage, networks—over the internet. Users install OS and applications.

• Core Offerings: Virtual Machines (VMs), Virtual Networks, Block Storage (EBS), Object Storage.

• Examples: Amazon EC2, Google Compute Engine, Microsoft Azure VMs.

• User Control: OS, middleware, runtime, applications, data.

2.2 Platform as a Service (PaaS)

• Definition: Provides a platform allowing customers to develop, run, and manage applications without infrastructure complexity.

• Core Offerings: Runtime environment, middleware, development tools (SDKs, databases), OS.

• Examples: Google App Engine, Heroku, Microsoft Azure App Services, AWS Elastic Beanstalk.

• User Control: Applications and data only.

2.3 Software as a Service (SaaS)

• Definition: Delivers complete, ready-to-use software applications over the internet, typically via a web browser.

• Core Offerings: Complete application, data, configuration.

• Examples: Gmail, Salesforce, Microsoft 365, Dropbox.

• User Control: Minimal configuration and data.

2.4 Comparative Analysis of SPI Models

Control vs. Convenience Spectrum:

IaaS  <---(More Control, Less Convenience)---  PaaS  <---(Less Control, More Convenience)---  SaaS

Shared Responsibility Model:

[!TIP] Exam Question: "Define SaaS, PaaS, and IaaS models and explain their differences." Use the spectrum diagram and responsibility table in your answer.

3.0 Cloud Deployment Models

Selecting a Deployment Model:

Decision Framework: Evaluate based on Cost (CapEx vs OpEx), Control & Customization needs, Security & Compliance requirements (e.g., GDPR, HIPAA), Scalability needs, and Organizational Constraints (skills, legacy systems).

[!TIP] Exam Focus: "Public vs. Private Clouds" and "Hybrid Cloud benefits/challenges" are recurring 7-mark questions. Use the table for structured answers.

4.0 Virtualization: The Foundational Technology

4.1 Introduction to Virtualization

• Definition: The creation of a virtual (rather than actual) version of something, including virtual hardware platforms, storage devices, and network resources.

• Enables: Efficient hardware utilization (consolidation), multi-tenancy, isolation, and rapid provisioning—core to cloud computing.

4.2 Types of Virtualization

• Server Virtualization: Multiple OS instances on a single physical server.

  • Full Virtualization: Complete hardware simulation (VMware, Hyper-V).

  • Para-virtualization: Guest OS modified to use hypervisor API for better perf (Xen).

  • Hardware-Assisted Virtualization (HVM): CPU extensions (Intel VT-x, AMD-V) for efficient virtualization.

• Storage Virtualization:

  • Block-level (SAN): Presents logical disks (LUNs) from a storage pool.

  • File-level (NAS): Presents shared file systems (NFS, CIFS).

  • Concept: Abstracts physical storage into a single logical pool.

• Network Virtualization: Combines physical network resources into a virtual network.

  • VLANs: Segregate broadcast domains.

  • SDN: Separates control plane (controller) from data plane (switches).

4.3 Hypervisors (Virtual Machine Monitors - VMM)

Functions: Resource allocation (CPU, memory, I/O), isolation between VMs, hardware emulation.

4.4 Virtual Machine (VM) Concepts

• Lifecycle: Create → Start/Stop → Suspend/Resume → Migrate (live) → Destroy.

• VM Image/Template: Pre-configured OS + apps, used for rapid provisioning.

• Security Risks:

  • VM Escape: Breaking out of VM to host.

  • VM Sprawl: Uncontrolled VM proliferation.

  • Image Security: Malicious/outdated images, insecure credentials.

4.5 Logical Partitioning (LPAR)

• Definition: Firmware-level partitioning of a physical server into independent logical partitions (LPARs), each with dedicated resources (CPU, memory, I/O).

• Advantages: Strong isolation, predictable performance, security, flexible resource allocation.

• Disadvantages: Complexity, potential resource fragmentation, less flexible than full virtualization (no VMotion).

• vs. Full Virtualization: LPAR uses hardware partitioning (no hypervisor overhead), but lacks features like live migration and dynamic resource reallocation.

4.6 Virtualized Data Center Architecture

[[DIAGRAM: CANVAS: A layered diagram showing:

1. Physical Layer: Rack servers, SAN/NAS storage, physical switches/routers.

2. Virtualization Layer: Hypervisors on servers, virtual switches (vSwitch), virtual storage (vSAN).

3. Management Layer: vCenter/System Center for orchestration.

4. Cloud Portal/API Layer: End-user interface for self-service.

]] Components:

• Virtualized Servers (Compute): Hypervisor hosts running VMs.

• Virtualized Storage: Storage pools presented as virtual disks.

• Virtualized Networking: Virtual switches, VLANs, virtual firewalls.

• Management Layer: Central console (e.g., VMware vCenter) for provisioning, monitoring, automation. Benefits: Server consolidation, agility, automated provisioning, improved disaster recovery.

[!TIP] Exam Question: "Explain how virtualization enables efficient utilization..." Focus on consolidation (multiple VMs on one server), isolation, and dynamic resource allocation.

5.0 Cloud Architecture & Design Patterns

5.1 Cloud Computing Reference Model

[[DIAGRAM: CANVAS: A layered model from top to bottom:

1. Client Layer ( browsers, mobile apps)

2. Application Layer ( SaaS apps)

3. Platform Layer (PaaS runtime, middleware)

4. Infrastructure Layer (IaaS: VMs, storage, network)

5. Management Layer (orchestration, billing, security) - spans all layers.

]]

• Layers: Client → Application → Platform → Infrastructure → Management.

• Management Layer is cross-cutting, providing orchestration, provisioning, metering, security, and compliance across all service layers.

5.2 Service-Oriented Architecture (SOA) in Cloud

• Core Principles: Loose coupling, reusability, interoperability, contract-based interactions.

• Role in Cloud: Enables composition of cloud services (from different providers) into business processes. Facilitates integration between on-premise and cloud apps.

• Technologies: Web Services (SOAP/XML for enterprise, REST/JSON for web/mobile), APIs (public/private), Enterprise Service Bus (ESB).

• Example: A retail app using AWS (compute), Salesforce (CRM), and a payment gateway API—all integrated via SOA/REST.

5.3 Cloud Stack

• Conceptual Layered Model: The SPI stack built on virtualization.

  
  SaaS (Applications)
  
  PaaS (Runtime, Middleware)
  
  IaaS (VMs, Storage, Network)
  
  Virtualization Layer
  
  Physical Hardware
  
  

• Example Stacks:

  • OpenStack: Open-source IaaS cloud operating system (Nova, Swift, Neutron).

  • Cloud Foundry: Open-source PaaS (buildpack-based deployment).

  • Proprietary Stack: AWS (IaaS/PaaS) + SaaS apps.

[!TIP] Exam Focus: "Cloud Stack" and "SOA in Cloud" are direct past questions. Draw the stack diagram and explain SOA's role with a concrete example.

6.0 Cloud Security

6.1 Importance and Unique Challenges

• Shared Responsibility Model: Provider secures the cloud (infrastructure). Customer secures in the cloud (data, apps, access).

• Challenges: Multi-tenancy risks (data leakage), data residency/sovereignty, compliance (GDPR, HIPAA), API vulnerabilities, account hijacking, insider threats.

6.2 Core Security Aspects

6.3 Secure Execution Environments & Communications

• Secure Bootstrapping: Verified boot process, measured boot (TPM).

• Trusted Platform Module (TPM): Hardware-based security for key storage/attestation (cloud TPMs available).

• Encrypted VM Disks & Memory: VM-level encryption (e.g., AWS EBS encryption), confidential computing (encrypted memory).

6.4 Securing Virtual Machines (VMs)

1. Use Hardened/Trusted Images: CIS Benchmarks, minimal OS, patched.

2. Network Segmentation (Micro-segmentation): Apply security groups at NIC/VM level.

3. VM Activity Monitoring & Logging: Centralized logging (CloudWatch, Azure Monitor), audit trails.

4. VM Sprawl Management: Automated de-provisioning, tagging, lifecycle policies.

[!TIP] Exam Question: "Discuss the importance of information security..." Structure: 1) Shared Responsibility, 2) Multi-tenancy risks, 3) Cover Data, Network, IAM aspects, 4) Compliance. Use RBAC and encryption as key examples.

7.0 Supporting Technologies & Concepts

7.1 Utility Computing

• Pay-per-use model: Resources metered and billed based on consumption (compute hours, GB stored, network I/O).

• Analogy: Like electricity—plug in, use, pay bill.

7.2 Quality of Service (QoS)

• Definition: Metrics defining service performance: Availability (uptime %), Performance (throughput, latency), Reliability (MTBF).

• Key Issues:

  • Network Latency: Delay in data transmission (critical for real-time apps).

  • Bandwidth Guarantees: Reserved bandwidth for critical apps.

  • Service Level Agreements (SLAs): Formal contracts specifying QoS metrics and penalties.

  • Noisy Neighbor: One tenant's resource hogging affecting others.

7.3 Storage Clouds

• Object Storage (S3): Unstructured data (blobs), REST API, massive scale, eventual consistency (e.g., AWS S3).

• Block Storage (EBS): Raw block devices for VMs, high IOPS, persistent.

• File Storage (EFS/FSx): Shared file systems (NFS/SMB), hierarchical.

• STaaS: Storage delivered as a service with pay-per-use.

7.4 Cloud Infrastructure Benchmarks

• Purpose: Compare performance, scalability, cost of cloud platforms/vendors.

• Common Metrics:

  • Compute: vCPUs, GHz, benchmarks (SPECint).

  • Storage: IOPS (Input/Output Operations Per Second), throughput (MB/s), latency (ms).

  • Network: Bandwidth (Gbps), latency, jitter.

  • Cost: $/VM-hour, $/GB-month, total cost of ownership (TCO).

7.5 Independent Software Vendors (ISVs)

• Role: Develop enterprise/business applications (e.g., SAP, Oracle) that run on cloud platforms.

• Cloud Adaptation: Re-architect for cloud (multi-tenancy, elasticity), create cloud-native versions.

• Certification Programs: AWS ISV Accelerate, Azure Marketplace—test and certify apps for cloud deployment, ensuring compatibility and best practices.

[!TIP] Exam Focus: "QoS" and "Storage Clouds" are direct past questions. Define QoS, list latency, bandwidth, SLA as key issues. For storage, differentiate Object vs Block vs File with one example each.

8.0 Open-Source & Proprietary Cloud Platforms (Briefing)

8.1 OpenNebula

• Overview: Open-source cloud management platform for IaaS.

• Key Features: VM orchestration, hybrid cloud support (AWS, Azure, vCenter), simple architecture, focuses on stability and interoperability.

• Use Cases: Enterprise private clouds, research, education.

8.2 Nimbus

• Overview: Open-source IaaS toolkit for cloud computing, originally from scientific community.

• Key Features: Contextualization (injecting user data/config into VMs), supports EC2-compatible APIs, lightweight.

• Use Cases: Scientific clouds, testbeds, research infrastructure.

8.3 Comparative Context

[!TIP] Exam Focus: "How OpenNebula used in cloud computing?" Explain its role as an IaaS manager for building private/hybrid clouds, contrasting with public cloud providers. Mention EC2 compatibility and contextualization for Nimbus.