Cloud Foundations, Characteristics, Vendors and Responsibility

cloud computing

Cloud computing provides network-accessible, metered computing resources from a shared pool that can be provisioned and released through an API with little.

on-demand self-service

The ability of an authorized consumer to provision computing resources through an interface or API without manual provider action.

resource pooling

Shared provider capacity assigned dynamically among tenants, with allocation controls and isolation limiting interference.

elasticity

Elasticity adjusts allocated capacity as demand changes, whereas scalability is the broader ability of a design to handle growth.

measured service

Measured service records resource consumption such as time, requests, bytes or capacity for billing, quota, allocation and optimization.

region and availability zone

A region is a geographic cloud area; an availability zone is an isolated location within it used to reduce correlated failure.

shared responsibility

The shared-responsibility model divides security and operational duties between the provider and customer according to the selected cloud service.

cloud governance

Policies and automated controls for identity, security, resource organization, compliance, cost and lifecycle management.

AWS Azure and GCP

Major public-cloud providers offering on-demand compute, storage, networking, databases, analytics and managed services.

Cloud Computing Fundamentals

What is Cloud Computing?

Cloud computing is the delivery of computing services (servers, storage, databases, networking, software) over the internet on a pay-as-you-go basis.

Key Concepts

Cloud Service Models

Cloud Providers

Virtualization

Batch vs Stream Processing

Batch Processing

Batch processing processes data in discrete chunks at scheduled intervals.

12:00 AM daily → Collect all transactions → Process → Load to DWH

Characteristics:

Tools: Apache Hadoop MapReduce, Apache Spark (batch mode), AWS EMR

Stream Processing

Stream processing processes data continuously as it arrives.

Event happens → Process immediately → Real-time result
(Latency: milliseconds to seconds)

Characteristics:

Tools: Apache Kafka, Apache Flink, Apache Spark Streaming, AWS Kinesis

Batch vs Stream Comparison

Distributed Storage

Why Distributed Storage?

• Single machine cannot handle PB-scale data
• Traditional storage: 1 machine; limited capacity
• Distributed storage: Spread data across hundreds or thousands of machines

HDFS — Hadoop Distributed File System

HDFS is the distributed storage system designed for Big Data (part of Hadoop ecosystem).

File: sales.csv (500MB)
### Cloud Storage Services
## Cloud Characteristics and Responsibility

Cloud computing provides network access to a pool of configurable resources that can be allocated and released through standardized interfaces. On-demand self-service lets authorized users provision without manual provider action. Broad network access exposes services through common protocols. Resource pooling shares physical capacity with logical isolation. Rapid elasticity adjusts capacity with demand. Measured service records consumption for control and billing.

Virtualization divides physical hardware into isolated machines while containers isolate processes on a shared kernel. Neither mechanism alone creates a cloud. Cloud operation also requires automated provisioning, APIs, metering and pooled capacity. Regions contain separate availability zones. Designs use zones for local resilience and regions when recovery must survive a regional failure.

The shared-responsibility boundary depends on the service. A provider secures facilities, physical hardware and the managed layers it operates. Customers remain responsible for identity, data, access policy and configuration. In IaaS they also patch guest operating systems and applications. In managed platforms the provider assumes more runtime work. Misconfigured public access remains a customer risk even when the storage service itself operates correctly.

Major vendors expose similar categories such as compute, object storage, databases, networking and identity but names and detailed behavior differ. Selection should compare failure model, portability, regional presence, skills, governance and total cost rather than only headline compute price.