Video summary

Sesión 3

Main summary

Key takeaways

Educational

Course/session logistics & participation (Session 3: Introduction to Cloud Computing)

  • Participants do not need to log in or be online during the session (no requirement for internet presence to attend).
  • Synchronous sessions and links can be found in the TEGNM MB area.
  • A QR code is provided to access playlists.
  • The MOP TEGNM is not used for roll call; no attendance is recorded.
  • The forum is intended only for questions and contributions related to the thematic content of each module.
  • Encouraged task after the session:
    • Post in the forum the most important lesson learned and an idea you found significant.
  • A 15-minute Q&A is planned at the end; further questions will be answered later in the forum.
  • Reminder:
    • Check informational messages in each module for general course progress and the final exam.

Learning objectives of the session

The session aims to reinforce understanding of:

  • Cloud computing
  • Service models

It is organized into four topics:

  1. Basic concepts/definition of cloud computing
  2. Main characteristics and ecosystem components
  3. Service models (multiple categories)

1) Definition and purpose of cloud computing

Core definition

Cloud computing is presented as a model enabling:

  • Ubiquitous, on-demand network access to shared computing resources

Key resources accessed “as a service” via the network include:

  • Networks
  • Servers
  • Storage
  • Applications
  • Related services

Colloquial explanation: delivery of computing resources and services (e.g., storage, processing, databases) over the internet, without needing physical infrastructure owned by the user.

Historical/operational context

Before cloud adoption, organizations typically needed:

  • Physical infrastructure (servers, equipment in controlled environments)
  • Dedicated human teams for maintenance and 24/7 operation

Over time, usage shifted:

  • From local storage (PC/laptop)
  • To internet-based cloud storage services

Consumer cloud storage examples mentioned:

  • Google Drive
  • Dropbox
  • OneDrive

Why organizations adopt cloud computing (four reasons)

  1. Pay-per-use

    • Pay only for resources consumed.
    • Avoids large up-front costs of buying and maintaining infrastructure.

  2. On-demand service

    • Handles workload spikes (e.g., Black Friday / major sales events).
    • Scales up with demand and scales down when demand falls, reducing wasted capacity spending.

  3. Lower barrier to growth / ability to meet unpredictable demand

    • When demand is unpredictable, local infrastructure can limit growth and availability.

  4. Lower initial investment / faster migration rationale

    • Less need for large capital expenditures and ongoing infrastructure maintenance.
    • Includes an anecdote about how heavy IT infrastructure investment reduced profits and affected competitiveness, motivating migration toward newer approaches (cloud discussed as part of that motivation).

2) Cloud providers and market leaders

  • The session lists “major contenders” that hold over 65% of global market share (three main providers), with the remainder held by others.
  • Growth and expansion:
    • Asian companies are investing heavily and adding data centers outside their region (including Latin America).
  • Example provided:
    • Huawei and a data center in Querétaro, plus claims about database service pricing competitiveness.
  • Other provider mentioned:
    • IBM (mentioned as supporting technology development)

Market leader figures stated:

  • Amazon Web Services (AWS): ~ 31%
  • Microsoft Azure follows
  • Google Cloud Platform follows

3) Core cloud characteristics (NIST-based) + detailed explanations

The session presents five NIST-attributed characteristics, each with implications and examples:

1. On-demand self-service

  • Users can provision resources without human interaction with the provider.
  • Analogy: like a supermarket—you pick what you need.
  • Consequences:
    • Resources/services may be free or paid, so users must understand pricing and features.
    • Requires sufficient network connectivity (adequate bandwidth).
  • “Provisioning” explained:
    • In IT/cloud, provisioning means setting up virtual resources such as:
      • servers
      • operating systems
      • databases
      • applications
    • With cloud, provisioning can be automatic and occur in minutes (not days).
  • Example:
    • Creating a virtual machine quickly via the provider’s interface/control panel.

2. Access via the network

  • Resources are accessible through the internet from many device types.
  • Data center location can be abstracted:
    • Users (e.g., in Mexico) may access resources hosted in Asia.
  • Emphasis:
    • Location matters less thanks to global distribution and network connectivity.

3. Resource pooling / aggregation

  • Users typically do not know the exact underlying physical machine(s).
  • Data center redundancy and high availability:
    • Virtual machines have images/backups.
    • If a VM fails, backup can be activated automatically—potentially on different hardware.
  • Shared physical infrastructure across multiple tenants, with reliability mechanisms to reduce risk of data loss or interruptions.

4. Elasticity (rapid elasticity)

Elasticity includes:

  • Scaling up when demand rises
  • Scaling down when demand decreases

Mostly automated in modern services; can also be configured based on load.

  • Example scenario:
    • A server handling 100 connections becomes overloaded at 1000.
    • Elasticity expands capacity, then reduces it later.
  • Scaling types mentioned:
    • Horizontal scaling: add more instances
    • Vertical scaling: increase capacity of a component (e.g., more memory for a database instance)
  • Example:
    • Black Friday / e-commerce spikes.

5. Metered service (measurement/pay-as-you-go)

  • Cloud systems automatically control, measure, and report usage.
  • Billing/monitoring granularity:
    • Statistics can be recorded at very fine resolution (the presenter mentions “per second”).
  • Personal experience anecdote:
    • Experimentation with providers around 2008–2010
    • Using cloud database services during the pandemic
    • Mention of free credit periods (e.g., “6 months” or “3 months” depending on provider)
    • Importance of monitoring to understand real cost.

4) Cloud computing ecosystem: main components (“stack”)

Cloud computing is described as a stack of physical and logical layers.

A) Physical infrastructure

  • Servers
  • Secondary storage systems (including storage arrays connected via mechanisms)
  • Networking equipment:
    • routers, switches, firewalls
  • Load balancers
  • Data centers emphasized as critical infrastructure
  • Evolution described:
    • file storage systems → relational/non-relational databases → analytics → AI-driven decision-making

B) Virtualization layer

  • Hypervisors abstract physical resources.
  • Mentioned approach:
    • “Bare metal” hypervisors (hypervisor virtualizes hardware directly)
  • Providers typically supply their own hypervisors.

C) Software and management tools

Focus on tooling supporting:

  • APIs
  • Automation
  • Orchestration
  • Monitoring
  • Security tools

API explanation: interface/mechanism for communication between software components (e.g., database APIs, service APIs for web apps).

Automation/orchestration examples:

  • Terraform
  • Ansible (spelling varies in subtitles)

Container orchestration:

  • Docker
  • Kubernetes
  • Containers support distributed computing across many systems/organizations.

5) Cloud service models (major and additional models)

The “pyramid” of the three main service models

1. IaaS (Infrastructure as a Service)

  • Provides virtualized resources:
    • servers, storage, networks
  • User responsibilities:
    • install/manage operating systems and applications
    • manage virtualized hardware
  • Example providers mentioned:
    • AWS
    • Google Compute Engine
    • Azure virtual machines
  • Best for:
    • organizations needing flexibility and control over infrastructure
  • Requires:
    • specialized IT teams for OS/database/network configuration.

2. PaaS (Platform as a Service)

  • Provides an integrated development environment.
  • Infrastructure is mostly hidden.
  • Provider manages OS/infrastructure; user focuses on:
    • programming and deploying applications
  • Examples mentioned generally:
    • managed databases/services such as “Google Cloud SQL” and AWS managed databases
  • Best for:
    • development teams that need fast, efficient deployment/testing environments.

3. SaaS (Software as a Service)

  • Ready-to-use applications hosted in the cloud.
  • Access via web (no local installation).
  • Provider manages:
    • updates, versions, backups, and runtime operation
  • User needs less technical infrastructure knowledge (user acts as an application consumer).
  • Examples mentioned:
    • Microsoft Office 365
    • Google Docs
  • Best for:
    • emerging businesses needing quick operational capability without building applications from scratch.

Comparison axis mentioned (conceptual): user control, OS responsibility, application responsibility, user profile/technical expertise expectations, and example services.

Additional/other service models mentioned

4. FaaS (Function as a Service) / “serverless”

  • Applications are decomposed into small units (“functions”).
  • Functions run only when specific events occur.
  • Billing is based on execution time (emphasis on paying per milliseconds).
  • Provider manages server lifecycle (power-on/run/shutdown).
  • “Serverless” clarified:
    • servers exist, but users do not maintain them.
  • Languages mentioned for functions:
    • Python, Java, and “Char” (likely C# in the subtitles).

5. DaaS (Desktop as a Service)

  • Cloud-hosted virtual desktops accessible from any internet-connected device.
  • Supports remote usage patterns (examples mentioned included TV console/email/document-style scenarios).
  • Advantages:
    • facilitates secure remote work
    • reduces costs (e.g., avoiding local installs like Office)
  • Operation model:
    • files/apps run on servers, not on the user’s device.

6. DBaaS (Database as a Service)

  • Configure, operate, and scale databases in the cloud without installing software or managing local hardware.
  • Emphasizes high availability and reduced operational burden.

Cloud deployment models (how clouds can be organized)

  • Public cloud
    • Offered by providers like Amazon, Google, and Microsoft.
  • Private cloud
    • An organization runs its own cloud.
  • Hybrid cloud
    • Combination of public and private.
  • Community cloud
    • Shared by multiple organizations.

Benefits and considerations mentioned

Benefits highlighted

  • Cost savings
  • Scalability and flexibility
  • Faster time to market (provisioning in minutes)
  • Improved collaboration
  • Advanced security (stated as a general claim; later noted as having trade-offs)
  • Sustainability and energy efficiency

Advice on selecting a cloud service model

A suggested methodology:

  • Identify your need/problem first
    • Determine what you already have (applications, databases) versus what you lack.
  • Match the model to the problem
    • Infrastructure → IaaS
    • Application environment → PaaS
    • Ready-to-use applications → SaaS
    • Event-driven code execution → FaaS
    • Cloud desktops → DaaS
    • Managed databases → DBaaS
  • Avoid buying unnecessary technology
    • Historical example: dot-com era spending on hosting costs without value.
  • Think of it like a “supermarket”
    • Many options exist; choose based on what solves your problem.
  • Evaluate realistically
    • Don’t rely only on marketing claims.

Security and shared responsibility (lesson)

  • “100% secure” systems are unrealistic in modern connected environments.
  • Cloud security follows shared responsibility:
    • Provider handles parts related to infrastructure/platform security
    • User must handle responsibilities such as:
      • access controls
      • ensuring correct restrictions
  • Practical implications:
    • define a security scheme/policy
    • continuously review provider security and your own data protection practices

Challenges to manage

  • Shared responsibility governance (provider + customer)
  • Supplier dependency risk:
    • suggested approach: evaluate multiple providers/avoid “marrying” one
  • Privacy and regulatory compliance:
    • expanding regulations; proper handling of sensitive data
  • Cost management:
    • costs may seem small at first, but must be incorporated into accounting and measured for return
  • Emerging trends:
    • AI/machine learning adoption
    • specialized industry clouds
    • multicloud and integration across providers
    • improved orchestration/UX
    • cyber resilience and security
    • digital sovereignty debate:
      • who controls data and where it is located
      • jurisdiction and compliance rules

Conclusion / key takeaways

  • Cloud computing (simplified): on-demand access to computing resources via the internet.
  • Five emphasized characteristics:
    • Self-service
    • Network access
    • Resource aggregation/pooling
    • Elasticity
    • Measurement
  • Service models summarized:
    • IaaS, PaaS, SaaS, plus FaaS, DaaS, DBaaS
  • Key message:
    • Cloud is not only technology—it’s also a business model that changes how organizations consume and manage IT.

Speakers / sources featured

  • NIST (United States National Institute of Standards and Technology)
    • cited as the source of the five cloud characteristics
  • Course presenter/speaker
    • name not provided in subtitles
  • Examples of companies/providers mentioned:
    • AWS
    • Microsoft Azure
    • Google Cloud Platform
    • Huawei
    • IBM
    • Automation tools: Terraform / Ansible (spelling varies)
    • Docker and Kubernetes
    • Microsoft Office 365, Google Docs
    • Oracle; “Asure”/Azure references (as cloud services in subtitles)
    • “ODU” (mentioned as an example SaaS provider)

Original video