Ai Cloud Computing

Traditional Deployment Models

On-Premises Deployment Characteristics

  • All equipment is located in your building

  • All equipment is owned by you

  • Clear lines of demarcation: everything in the building is your responsibility

  • Equipment is CapEx (capital expenditure)

  • New equipment typically takes over a week to deploy

  • Equipment requires technology refreshes

  • Consideration for redundancy is necessary.

Colocation Facilities

  • A colocation center rents out space for external customers

  • Facility owner provides power, cooling, and security for customer equipment

  • User desktops remain in offices

  • Independent providers (e.g., Equinix) offer multiple network connectivity options

Colo Solutions Characteristics

  • Colo provider owns data center; provides power, cooling, and security

  • You own your server, storage, networking equipment in the facility

  • Connections to your offices are the provider's responsibility

  • Colo hosting fees are considered OpEx (operational expenditure)

  • New equipment deployment takes same time as on-premises

  • Need to consider redundancy for owned hardware.

Cloud Computing

Definition

  • Cloud computing: model for on-demand network access to a shared pool of configurable resources

Characteristics

  1. On-Demand Self Service: Consumers can provision capabilities automatically without human interaction.

  2. Rapid Elasticity: Resources can be scaled up or down dynamically based on needs.

  3. Broad Network Access: Services are available over the network via various client platforms.

  4. Resource Pooling: Provider's resources serve multiple consumers, optimizing costs.

  5. Measured Service: Resource use is monitored and optimized automatically.

Virtualization

Server Virtualization

  • Shares hardware among multiple customers.

  • Hypervisors allocate resources to virtual machines.

Hypervisors

  • Type 1 (Bare Metal): Run directly on hardware (e.g., VMware ESXi, Microsoft Hyper-V).

  • Type 2: Run on a host OS (e.g., VMware Workstation, VirtualBox).

Containers

  • Lightweight virtualization of software layers above the OS.

  • Portable, fast to provision; no OS included.

  • Docker is a well-known container engine.

Cloud Service Models

  • IaaS (Infrastructure as a Service): Provider manages hardware; customer manages OS and apps.

  • PaaS (Platform as a Service): Provider manages the platform; customer develops applications.

  • SaaS (Software as a Service): Software is delivered over the internet.

Cloud Deployment Models

  1. Public Cloud: Open use; managed by third-party providers.

  2. Private Cloud: Exclusive use by a single organization.

  3. Community Cloud: Shared resources among several organizations with common interests.

  4. Hybrid Cloud: Combination of multiple cloud environments with data/application portability.

Cloud Computing Advantages

  • Scaleability: Adjust resources based on current needs.

  • Cost Efficiency: Pay for what is needed; reduced CapEx.

  • Competitive Advantage: Lower infrastructure costs free up capital for innovation.

  • Productivity: IT staff focus on strategic projects instead of hardware management.

  • Availability & Reliability: Major providers have robust facilities with redundancy and security.