AWS Cloud Practitioner

AWS Cloud Practitioner

1. History of Computing Infrastructure

   • The Past
     • Companies host their own data centers, which means they must pay operation costs and provision for their worst-case scenario, leading to over-provisioning
   • The Present
     • Infrastructure is provided as a service (IaaS) by cloud providers like AWS, GCP, or Azure
       • Pay as you go
       • Scale as necessary
       • New IT Resources on Demand
       • Benefit from Economies of Scale
       • Result: Deploy globally quickly and focus on what differentiates your application

2. Cloud Economics: Business Value

   • Cost Benefits
     • Compare cost of traditional IT setup to AWS
       • Identify end-to-end costs
         • Typically 20-40% less
     • Best Practices
       • Ensure releavnt shareholders are present
       • Use realistic estimates for usage
       • Assign cost / value to business value factors like agility, risk, automation
     • anti-practices
       • Apples to apples comparison. AWS Is more than just discounts
       • Forget operational costs like power, administration, rent, software, hardware
     • Lower costs with AWS
       • Consumption model: only pay for what you need
       • Average payback time of 6 months
       • AWS Flywheel
     • Migration
       • Return on Investment
         • Customers want to retire hardware that is fully deprecated.
         • Cost ( and effort ) of migration requires analysis
         • Cost savings / (sunk costs + migration costs ) = ROI
       • Cloud Readiness
         • Human factor
           • Team lacks the skills and experience to transform
           • May oppose cloud development for logical or illogical reasons
         • Application factor
           • Dependency and Tension
           • Identify easy and difficult applications
       • Process typically takes about a month
   • Staff Productivity
     • Less tactical ( undifferentiated ) more strategic (undifferentiated work)
     • Staff Productivity functions
       • Facilities
         • Disaster recovery, planning, power, cooling reduced / eliminated
       • Security
         • Security tool administration, set up new devices reduced
       • Network
         • Network maintenance, hardware, infrastructure eliminated or reduced
       • Server
         • Reduce patching, arranging hardware planning, repairing, purchasing, etc
       • Storage
         • System updates, installation reduced or eliminated
       • Application
         • Test code, execution, environment improved
   • Operational Resilience
     • important factors are up-time and security
     • Cornerstones
       • Operations
         • causes for failure: human errors, configuration errors, commonplace accidents
         • how AWS helps: leverages automation, monitors access, end-to-end servuce management
       • Security
         • Causes for failure : malware, network, poor authentication
         • How AWS Helps: Shared security model, built in DDOS protection, AWS IAM
       • Software
         • Causes for failure: Resources exhaustion, logic errors, bad monitoring, failed updates
         • How AWS helps: quick rollouts, auto CI, smaller deployments, collects resources
       • Infrastructure
         • causes for failure: hardware failure, natural disaster, power outage
         • How AWS helps: run applications over multiple AZs, multiple power sources
   • Business Agility
     • Delivering more value in less time
     • Measure using key performance indicators (KPI)
       • ex. new applications launched per year, features per release, etc.
     • Innovate by failing fast while reducing risk and cost

3. AWS Tools

   • Compute
     • Elastic Cloud Compute (EC2)
       • Elastic for auto-scaling
       • Multiple Instance-types
     • Elastic Auto-scaling
       • Monitors instance health
       • Replaces broken instances
       • Balance across AZ
       • Dynamic and predictive Scaling
     • EC2 Auto-Scaling
       • Types: Application load balancer, network load balancer, classic load balancer
       • Distributes application traffic across EC2 Instances
     • AWS Lambda
       • Serverless Computing
     • Container-Service
       • Scalable Container management
   • Storage
     • Elastic Block Storage (EBS)
       • Network-attached service like a hard drive
       • Independent of EC2, can be attached to an instance in same AZ
         • One EBS volume to one ec2 instance, one instance can have many EBS
     • Simple Storage Service (S3)
       • Stores data as objects in buckets
       • Common use cases: backup + storage, media hosting, software delivery, application hosting
       • Cheaper based on infrequency of access
   • Database
     • DynamoDB
       • No SQL
     • RDS / Aurora
       • Relational Database. Auroa: Postgres
     • Elasticache
       • Caching System
     • Host DB on EC2
       • Option to host DB on compute instance
         • Pros: More control. Cons: less automation
   • Network
     • Virtual Private Cloud (VPC)
       • Gives users full control over application network
         • Utilize subnets, which are ranges of IPs
     • Security Groups
       • Controls access to instances. Instances can be added to subsets of security groups
     • Network Access Control Lists (NACL)
       • Control Access to subnets
     • Route 53
       • Routes end users to internet application
   • Security
     • Identity and Access Management (IAM)
       • Fine-grained access control
         • Access Analysis and MFA
         • Integration with company directories
     • Shared Responsibility Model
       • AWS Responsible for infrastructure, user responsible for application
   • Management
     • Management Console (GUI)
     • CLI
     • SDKs

4. Well-Architected Infrastructure

   • Security
     • Encompasses infrastructure, data, incidence protection
     • Best Practices
       • Implement security at all levels
       • Enable traceability for all important actions
       • Principle of least privilege - Give people only as much privilege as they need
       • Focus on security your system - Shared Responsibility model
   • Reliability
     • Ability to recover from failure and meet demand
     • Best Practices
       • Test recovery procedures
       • Automatically Recover
         • ex. use auto-scale to remove broken instances
       • Scale horizontally, Scale automatically
         • use auto-scale to create more instances instead of better instances.
   • Cost Optimization
     • Pillars
       • Match supply with demand
         • leverage auto-scaling and elasticity
       • Increase expenditure awareness
       • Optimize over time
         • use reserved instances over on-demand
     • Best Practices
       • Pay only for what you use
       • Measure and attribute expenditure
   • Operational Excellence
     • Pillars
       • Manage and automate changes
       • Respond to events
       • Define Standards
   • Performance Efficiency
     • Pillars
       • Select customizable solutions
       • Review to continually innovate
         • Monitor Performance
         • Automate using CloudWatch
       • Monitor AWS Services
       • Consider trade-offs
     • Best Pratices
       • Democratize advanced technologies
         • Consider consuming as a service instead of learning a skill
       • Experiment and use the best tool for the job

5. Customer Strategies

   • Migration Strategies
     • Rehost
       • Recreate on-premises setup on AWS
       • Lift and Shift
     • Replatform
       • Lift, Tinker, and Shift
       • Retain core architecture, but make targeted AWS optimizations. ex migrate db to RDS
     • Refactor
       • Modernize by re-imagining how the application is architected and developed. ex monolith to services
     • Retire
       • Shutting off non-useful components
     • Retain / Revisit
       • Keeping certain applications on-premise, only migrate what you need
     • Repurchase
       • Moving workflows to software as a service
   • Customer facing discussion stages
     • Discovery
       • Understand the customer's challenges
       • Preparation
         • Research customer's industry. Understand trends , competitors, recent news, etc.
         • Research customer's business, understanding business mode, mission statements, products, leadership, and customer's relationship with AWS
       • Practices
         • Encourage detailed conversation with open-ended questions
         • 5 Whys. Dive deeper to discover the real desired outcome
         • Whiteboarding. Keep track of the conversation and illustrate workflows
     • Presenting the Solution
       • Present findings and propose AWS solution
       • Preparations
         • Collraborate, Illustrate, Iterate
         • Vary solutions by different factors
       • Pratices
         • Revise, gain buy-in from customer, propose PoC
         • Bring supplies, maintain eye contact, draw clearly, stand aside
     • Handling Objections
       • Goals
         • Connect - empathize. Make them feel heard and validated
         • Condense - identify the core concern
         • Continue - Take steps towards resolution
       • Practices
         • Data-driven approach
         • Use case studies
         • Have confidence in your response
       • Common Object Responses
         • Security
           • Top Priority, validated against third party frameworks
         • Cost or Cost savings
           • Reduced total cost of ownership
           • Optimized and predictable
         • Scalability and Response
           • Create Guardrails around security, availability, resilience, compliance
           • Amazon Watch Tower
       • Customer Meeting Practices
         • Keys to consistent results
           • prepare - understand industry, terminology
           • Anticipate, differentiate, stay on message
         • Anti-practices
           • Promises- don't use guaranteed, definitely, never
           • Use acronyms or jargon - spell it out
           • Focus on technology - focus on the customer
           • Focus on the short-term: focus on long-term
     • Delivering a PoC
       • PoC and Purpose
         • a PoC is a small-scale practical version of the proposed solution that runs the customer's application
         • Purpose is to help customer evaluate solution and compare to existing environment
           • Don't just demonstrate technology, cater to their busienss
       • Build a PoC
         • Take the feedback from the solution proposal
         • Collect networking and security, application code, databases, data, etc.
       • Utilize AWS online resources

6. Cloud Financial Management

   • Measurement and Accountability
     • Help customers understand their spending
     • Measuring and monitoring tools
       • Generate automated reporting and recommendations
         • Reports should have cost and usage data, optimization, recommendations
       • Options
         • Build a custom dashboard
         • APN tool
         • AWS Cost Explorer
   • Cost Optimization
     • Right-sizing Instances
       • Identify instances that can be downsized
         • Reserved Instances
           • up to 75% savings
           • Use for stuff that needs to run for most of the time
           • Convertible to other instance types
         • Spot Instances
           • Spare EC2 capacity, can be reclaimed at 2 minute notice
           • Use for fault-tolerant, flexible, stateless workloads
             • ex. big data, CI/ CD, high performance computing
       • utilize CloudWatch metrics
   • Cloud Financial Operations
     • Establish a Cloud Center of excellence
       • goals: align incentives, automate report, set up KPI
       • Cost Management Questions
         • How much of our workloads are steady state? Can we make them reserved instances, etc.
       • Establish clear metrics and targets
   • Migration Portfolio Assessment
     • Use to validate business plan, plan migration, and estimate cost of migration
     • Business case
       • Discovery
         • Evaluate total cost of ownership or cost savings
         • Estimate comparisons between on-premises and AWS
         • Estimate migration costs
       • Directional sense without getting too deep
       • Streamlines, automates, and scales financial analysis
     • MPA Data Import using Configuration Management Database extract
     • MPA Cost Comparison