AWS Theory

Cloud Concepts

AWS Well-Architected Framework

• AWS Well-Architected Framework are the best practices for designing and using AWS systems optimally.
• AWS Trusted Advisor makes sure you’re following these guidelines.
• 6 pillars of AWS Well-Architected Framework
  • Security - Recommends using correct security settings for all AWS services
  • Reliability / Fault Tolerance - Recommends increasing redundancy with Availability Zones to handle outages.
  • Performance - Recommends optimizing speed and responsiveness of your apps to avoid users timing out.
  • Cost optimization - Recommends reducing bill by getting rid of unused resources and using the best Pricing Plan for your usage.
  • Sustainability / Service Usage Limits - Recommends making adjustments if you’re using over 80% of your service usage limits (high traffic) to avoid users timing out
  • Operational excellence - Recommends making small incremental changes to production and doing monthly game days to maintain daily operation of applications without issues.
• Other good design principles

  • Loose coupling / Decouple components

    • Make controllers independent by using queues (asynchronous messaging) instead of having them directly call each other
    • If one component fails, failure won’t cascade to other components.
    • AWS SNS, SQS, EventBridge help with this.

    

  • Think parallel

  • Implement elasticity

  • Design for failure

  • Services > Servers

    • Stray away from provisioning servers on hardware to run jobs.
    • Instead, switch cloud (serverless) computing using AWS services.

AWS Shared Responsibility Model

• AWS handles the security OF the cloud
  • Security of underlying infrastructure of AWS resources - Storage, Databases, Network, and Computational Engines (Lambda, EC2)
  • Security of Global Infrastructure setup - Availability Zones, Regions, Edge Locations etc.
  • Anything evolving physical data centers.
• Customers handle the security IN the cloud
  • Security of web applications
  • Security of user data (e.g. credit card info) stored inside AWS services
  • Encryption (Server-side, Client-side)
  • Manually configuring AWS security services - Firewalls, IAMs, Security Groups, Network ACLs,
• Some responsibilities are shared between both, depending on circumstance.
  • Configuration Management - AWS configures their global infrastructure. Customer configures their apps, DBs, and guest OS's.
  • Patch Management - AWS handles host OS’s, Customer handles guest OS’s
  • Training employees.
• Some AWS controls are inherited by the customer
  • Physical and Environmental Controls: AWS usually handles these, but the customer also has optional control.

Cloud Adoption Framework

• AWS Cloud Adoption Framework (CAF) - Service/Tool used to plan migrations to AWS Cloud

  • Separates migration strategy into 6 alternative perspectives

  Perspective	Focus	Stakeholders
  Business	Focuses on achieving leadership's desired business outcomes.	C-level Executives (CEO, CTO, CFO, COO, etc.)
  People	Focuses on building a bridge between tech and business, and prioritizing employee needs and growth during migration.	C-level Executives apart from CEO
  Governance	Focuses on maximizing organizational benefit and minimizing risk during migration.	C-level executives apart from CEO
  Platform	Focuses on enabling the tech teams implementing the migration to the cloud platform	Tech Leaders (IT, Engineers, Architects)
  Security	Focuses on ensuring compliance and data security during migration.	Tech and Security Leaders
  Operations	Focuses on enabling IT, SRE, and DevOps teams that will maintain daily operation of production applications and batch jobs	IT and SRE Leaders

  • Separates cloud transformation into 4 phases
    1. Envision: Planning. Measuring business outcomes.
    2. Align: Identify potential problems
    3. Scale: Expand existing cloud infrastructure
    4. Launch: Initial production launch. Evaluate business results.

Advantages of AWS & Cloud Computing Over Physical Servers

• Agility
  • Don't have to dedicate resources to managing infrastructure and on-prem data centers.
  • Can dedicate more resources to project management and application development.
  • Speeds up business and project management results (Agile)
• Deploy globally in minutes
  • Can quickly setup multiple AWS Regions and deploy servers in them.
  • Don't need to build physical servers around the globe.
• Elasticity
  • You only ever need to provision the resources that you need right now.
  • Can scale down AWS resources (EC2 instances, RDS clusters, etc.) when you don’t need them.
  • Easier than decommissioning on-prem server equipment.
  • Cheaper than overprovisioning on-prem server equipment to stay safe.