Cloud Computing

 

Cloud Computing

The Historical Evolution of Computing

The shift toward modern cloud computing occurred across four major milestones:

1950s: Mainframe Computing

Centralized, expensive mainframes shared via "dumb terminals." Introduced the core concept of economically shared hardware resources.

1970s: The Birth of Virtualization

Introduction of Virtual Machines (VMs), enabling multiple isolated operating systems to run simultaneously on a single physical host.

1990s: Virtualized Networking

Telecommunications shifted from dedicated, expensive point-to-point lines to Virtual Private Networks (VPNs) sharing public physical infrastructure securely.

Modern Day: Cloud Computing

Present

On-demand delivery of compute, storage, and application services over the internet with an elastic pay-as-you-go financial model.

Core Mechanism: Virtualization

Virtualization abstracts physical hardware by inserting a digital hypervisor layer between the hardware and the operating system.

• Problems Solved:

• Underutilization: Eliminates idle resources by running multiple isolated applications or OS instances on one server at full capacity.

• Legacy Software Support: Allows older legacy systems to run side-by-side with modern operating systems on identical physical hardware.

• Key Benefits:

• Cost Efficiency: Reduces hardware footprint, power consumption, and physical space requirements.

• Operational Agility: Enables near-instantaneous server provisioning compared to physical hardware acquisition.

• Automated Disaster Recovery: Simplifies backup, migration, and automated failover by treating entire system states as portable software files.

Core Strategic & Business Advantages

Moving from traditional hosting to cloud architecture shifts both operational and financial models:

Financial Efficiency

• CapEx to OpEx Shift: Replaces large upfront capital investments in physical data centers with variable operational expenses.

• Pay-as-You-Go: Billing occurs strictly for consumed resources (per-second or per-hour), avoiding costs for idle infrastructure.

• Economies of Scale: Public cloud providers aggregate millions of workloads, passing reduced unit costs down to customers.

Operational Agility & Scalability

• Elasticity: Infrastructure automatically scales up or down dynamically based on live traffic demands without advance notice.

• Global Reach: Rapid deployment of assets into multiple global regions reduces user latency and ensures compliance.

• Total Cost of Ownership (TCO) Reduction: Offloads non-differentiating operational burdens (power, physical security, hardware repairs, OS patching) to the cloud vendor.

Cloud Deployment Models

Model	Primary Characteristics	Common Use Case
Public Cloud	Infrastructure owned and operated by a third-party vendor over the public internet.	Rapid scaling, dynamic web applications, startup platforms.
Private Cloud (On-Premises)	Virtualized infrastructure dedicated solely to one organization, hosted internally or in a private data center.	Strict regulatory compliance, legacy systems, high security requirements.
Hybrid Cloud	Integrated environment combining public cloud services with private/on-premises infrastructure.	Gradual cloud migration, sensitive data isolation with elastic bursting capabilities.

The Cloud Service Pyramid (IaaS vs. PaaS vs. SaaS)

       / \

       /   \       SaaS: Software-as-a-Service (End-user applications)

      / SaaS \      Examples: Webmail, Google Workspace, Salesforce

     /--------\

    /   PaaS   \    PaaS: Platform-as-a-Service (Developer environment & runtime)

   /------------\   Examples: App Engine, Managed Kubernetes, Heroku

  /     IaaS     \  IaaS: Infrastructure-as-a-Service (Virtual servers, networks, storage)

 /----------------\ Examples: Compute Engine, Virtual Machines, Cloud Storage

Google Cloud Platform (GCP) & Innovations

Google Cloud Platform opens Google's internal global computing infrastructure to external developers.

• Core Offerings:

• Google Compute Engine (IaaS): On-demand virtual machines with granular per-second pricing.

• Google Cloud Storage (Object Storage): Scalable storage for unstructured files and media.

• Google Cloud Datastore / Firestore (Managed NoSQL): Serverless, auto-scaling document databases.

• Foundational Whitepapers & Big Data Contributions:

• MapReduce: The research paper that directly inspired Apache Hadoop for distributed batch processing.

• Bigtable: The design foundation behind distributed NoSQL databases like Apache HBase.

• Spanner: Globally distributed SQL database providing ACID transactions at scale.

Storage, Networking, & Performance Architecture

Performance Trade-Offs

• Latency vs. Throughput:

• Latency: Network delay measured in milliseconds (ms). Essential for real-time responsiveness.

• Throughput: Total volume of data transferred over a given window (MB/s). Essential for bulk data processing.

• Edge Caching (CDN): Replicates static assets across global edge locations physically closer to users, minimizing round-trip latency.

Abstracted Reliability & Managed Data Services

Modern managed storage abstracts physical hardware failure entirely. Redundancy, cross-region replication, and automated backups create a system where physical drive failures do not result in data loss or service disruption.

Big Data Analytics Stack

Modern platforms log and analyze immense data volumes using specialized distributed frameworks:

• Apache Hadoop: Distributed storage (HDFS) and batch computation.

• Apache Spark: High-speed, in-memory analytics engine.

• Apache HBase: Scalable, low-latency random read/write access over massive datasets.