Introduction to Cloud Computing

Cloud Computing Definition: On-demand access to shared computing resources like servers, storage, databases, networking, software, analytics, and intelligence over the internet.
Cloud Types: Public (vendors like AWS, Azure, Google Cloud Platform), Private (owned, operated, and maintained within an organization), Hybrid (combines public and private clouds).
Cloud Service Models: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS)
Cloud Deployment Models: Public, Private, Hybrid, Community (shared by several organizations with common concerns).

Cloud Computing Architecture & Infrastructure

Cloud Reference Model: Defines the components of a cloud system:
- Users: Individuals or organizations accessing cloud services.
- Cloud Provider: Owns and manages the cloud infrastructure.
- Cloud Services: Services offered by the cloud provider, including IaaS, PaaS, and SaaS.
- Cloud Infrastructure: Hardware and software used to deliver cloud services.
Virtualization: Technology that allows multiple operating systems to run concurrently on a single physical computer.
- Compute Virtualization: Virtualizes the processing power, memory, and storage of a physical server.
- Network Virtualization: Allows for the creation of virtual networks independent of physical network limitations.
- Storage Virtualization: Pools physical storage devices into a single logical storage space, streamlining data access and management.

Types of Hypervisors

Type 1: A hypervisor that runs directly on the physical hardware, managing the resources of the host machine.
Type 2: A hypervisor that runs as a software application on a host operating system.

Cloud Platforms in Industry

Major Vendors:
- Amazon Web Services (AWS): Offers a wide range of cloud services, including EC2 (Elastic Compute Cloud), S3 (Simple Storage Service), and more.
- Microsoft Azure: Provides comprehensive cloud services for computing, storage, networking, and more.
- Google Cloud Platform: Offers services for data analytics, machine learning, artificial intelligence, and more.

Cloud Applications

Protein Structure Prediction: Using cloud computing to analyze large amounts of data and simulate protein folding.
Data Analysis: Cloud computing enables the scaling of data analytics workflows, processing vast datasets for insights.
Satellite Image Processing: Cloud platforms handle the intensive processing required for analyzing and visualizing satellite imagery.
CRM and ERP: Cloud-based Customer Relationship Management (CRM) and Enterprise Resource Planning (ERP) systems provide scalable and accessible solutions for businesses.
Social Networking: Cloud infrastructure supports the vast user bases and data storage requirements of social media platforms.
Scientific Application: Cloud computing allows researchers to process complex scientific data and run computationally intensive simulations.
Business Application: Cloud services have revolutionized how businesses operate, enabling them to access resources on demand and scale operations as needed.

Advance Topics in Cloud Computing

Cloud Security: Ensuring the confidentiality, integrity, and availability of data and resources stored and processed in the cloud.
Risks of Cloud Migration: Data breach, data loss, compliance violations, and disruptions in service.
Approaches to Migration: A well-defined strategy with careful planning, security assessments, and risk mitigation considerations.
Federated Cloud/Intercloud: A network of interconnected clouds, where resources and services can be shared between different cloud providers.
Third Party Cloud Services: Services offered by external providers, such as identity and access management, data backup, and disaster recovery.
Business Continuity & Disaster Recovery: Strategies for ensuring continuous operations and data recovery in the event of a disruption to cloud services.
Service Level Agreement (SLA): A contract between a cloud provider and its customer that defines the level of service expected, including performance, availability, and support.
Dynamic Resource Provisioning & Management: Automatically scaling resources up or down in response to changing demands, optimizing resource utilization and cost.
Server Consolidation & Placement Policies: Strategies for optimizing server utilization and placement within a data center to improve efficiency and performance.
Energy Efficiency in Data Centers: Adopting energy-saving technologies and practices to reduce the energy consumption of data centers.
Elastic Load Balancing & Auto Scaling: Distributing traffic load across multiple servers to ensure high availability and performance.

Storage Network Design

Architecture of Storage: The design of storage networks is crucial for performance, scalability, and reliability.
Analysis and Planning: Understanding storage requirements, data access patterns, and performance goals.
Design Considerations: Network bandwidth, latency, security, and scalability.
Network Access Storage (NAS): File-based storage that allows multiple clients to access data over a network.
Fibre Channel (FC) Storage Area Network (SAN): A high-speed network designed for high-performance storage access.
Hybrid Storage Networking Technologies: Combining different storage technologies to optimize performance and cost.
- iSCSI: Internet Small Computer System Interface (iSCSI) allows standard Ethernet networks to be used for storage connectivity.
- FCIP: Fibre Channel over IP (FCIP) enables FC SANs to be extended over IP networks.
- FCoE: Fibre Channel over Ethernet (FCoE) encapsulates Fibre Channel traffic over Ethernet networks.

Storage Virtualization in Cloud Computing

Pooling Storage Resources: Virtualizing storage enables cloud providers to efficiently manage and allocate storage resources across multiple users.

Host System Design Considerations

Hardware and Software Selection: Optimizing the hardware and software configuration of host systems to meet the demands of cloud workloads.

Techniques for Big Data Processing

Google File System (GFS): A distributed file system designed for storing and accessing large datasets across multiple servers.
BigTable: A distributed database optimized for handling massive amounts of structured data.
MapReduce: A programming model that allows large-scale data processing to be divided into smaller, independent tasks that can be run on multiple machines.
Hadoop Distributed File System (HDFS): A distributed file system developed as part of the Apache Hadoop framework for storing and accessing large data sets.
HIVE: A data warehouse software that provides a SQL-like interface for querying data stored in HDFS.

Consensus in Cloud Computing

Issues in Consensus: Reaching agreement among multiple entities in a distributed system, where communication can be unreliable and nodes can fail.
Synchronous Systems: All nodes have access to a global clock and know the maximum time it takes for messages to be delivered.
Asynchronous Systems: Nodes do not share a common clock, and message delivery times are uncertain.

Byzantine Agreement

Agreement: A state where all non-faulty nodes in a distributed system agree on a common value.
Faults: Errors or malfunctions in a distributed system, such as node crashes or malicious behavior.
Tolerance: The ability of a system to continue operating correctly despite the presence of faults.
Measuring Reliability and Performance: Evaluating the performance and reliability of a distributed system based on factors like availability, latency, and throughput.

SLIs, SLOs, SLAs, TLAs

Service Level Indicator (SLI): A metric that measures the performance or availability of a service.
Service Level Objective (SLO): A target level of performance or availability for a service.
Service Level Agreement (SLA): A contract between a service provider and its customers that defines the expected level of service.
Target Level Agreement (TLA): A specific target level of service that a service provider commits to achieve.

### Byzantine Failure

A type of failure in a distributed system where a node may behave maliciously, sending inconsistent or misleading information to other nodes.

### Byzantine Generals Problem

A classic problem in distributed computing that illustrates the challenges of reaching consensus in the presence of Byzantine failures.

### Failures & Recovery Approaches in Distributed Systems

Fault Tolerance: Designing systems to handle failures gracefully.
Checkpointing: Regularly saving the state of a system to enable rollbacks and recovery in case of failures.