Skip to content

Latest commit

 

History

History
111 lines (95 loc) · 5.74 KB

1. Introduction.md

File metadata and controls

111 lines (95 loc) · 5.74 KB
Raw

Introduction

AWS Regions

  • AWS has regions all around the world, it is a separate geographic area. e.g. us-east-1 eu-west-1 ap-northeast-1
  • Pricing
  • Each region has at least two availability zones
    • E.g. us-east-1a, us-east-1b
    • Availability zone is one or more physical data centers.
      • They're separate from one another. They're isolated deliberately to survive any unforseen circumstances like disasters.
    • Name of AZ can correspond to different data centers in one account than in other.
      • E.g. same data center can be us-east-1a for one account while us-east-1b for another, this is purposefully done by aws to distribute resources across AZ.
  • Each AWS console are region scoped (except IAM, Route53 and S3{Anyhow buckets inside are regional})
    • I.e. changes, actions are specific to a region
  • See where the regions are and how many AZ's within them
  • There are also many Edge Locations for caching content for CloudFront (CDN).

Basic Terminology

  • Availability
    • Percentage uptime
      • Downtime can be due to faults or other such as planned maintenance.
    • E.g. the storage system is operational and can deliver data upon request
    • 📝High availability in AWS = Multiple AZ
    • Resiliency
      • Being able to adapt under stress or faults in order to avoid failure
        • Self-recover & remain functional to customers
      • Often provided by redundancies and automatic re-routing of operations within a system
    • Durability
      • Long-term data protection
      • E.g. the stored data does not suffer from bit rot, degradation or other corruption
    • Redundancy
      • The inclusion of extra components in a system so it will continue to function in case of a component failure.
      • E.g. automatic back-ups to Azure in case of AWS failure
    • Throughput
      • Measures of how many units of information a system can process in a given amount of time
  • Consistency
    • Application is expected to remain stable
    • Reliability
      • Percentage uptime, considering the downtime due only to faults
      • Measures consistency
      • Ability of a system to recover from infrastructure or service disruptions.
      • 💡 Kill unhealthy & spin up new => Quicker than human debugging
    • Fault-tolerance
      • App ability to self-detect and correct all types of problems in its environment
      • Requires extra redundancy to work in inconsistent environments.
      • 📝 Elastic Load Balancer, Route 53 with Auto Scaling Groups
    • Robustness
      • App ability to self-detect and correct all types of problems from its inputs.
      • Can work with inconsistent data.
  • 🤗Trivia: In distributed data store, CAP Theorem means you cannot reach more than two of: Consistency & Availability & Partition tolerance (packages being dropping/delayed between nodes).
  • Billing: You can see your bills in -> My billing dashboard -> bills

Basic Services

  • AWS Health: Provides ongoing visibility into the state of your AWS resources, services, and accounts.
    • Health events
      • Maintainance events that'll affect your services
      • Can set-up CloudWatch alarms.

AWS Support

  • Support types

    Basic Developer Business Enterprise
    Cost Free $29/mo $100/mo $15.000/mo
    Use case Account and billing questions only Experimenting Production use Mission-critical use
    Tech support NO Business hour via e-mail 24x7 via email, chat & phone 24x7 via email, chat & phone
    SLA ✕ 12-24 hrs at local business hours 1 hr response to urgent support cases 15 min to critical support cases w/ priority
    TAM & Support Concierge ✕ ✕ ✕ ✓
    Support cases ✕ 1 Person, unlimited cases Unlimited contacts / cases Unlimited contacts / cases

  • AWS Trusted Advisor

    • High level AWS account assessment, very useful to know whether our account is sticking to AWS' best practices.
    • Serverless global service
    • Pricing: Most checks requires different support plans (developer, only 7).
    • Analyzes your AWS accounts and provides AWS-defined recommendations based on core checks:
      • Cost Optimization: e.g. Low utilization Amazon EC2 instances, Idle Load Balancers, unassociated Elastic IP address...
      • Performance: e.g. High Utilization Amazon EC2 Instances, Large Number of Rules in an EC2 security group
      • Security: e.g. "specific ports unrestricted", "Amazon EBS public snapshots"
      • Fault Tolerance: e.g. "VPN Tunnel Redundancy", "Age of EBS snapshots", "EC2 AZ Balance".
      • Service Limits
        • Free to 📝check if you're getting close to limits
        • E.g. "EC2 On-Demand instances", "Cloud formation stacks"
    • Can enable weekly email notification from the console

AWS Pricing

AWS Pricing Calculator

  • Provides an estimate of AWS fees and charges for different cases/services.
  • ❗ Doesn't include any taxes
  • Uses AWS Price List API
  • Quick estimates
    • Ballpark estimate while requiring minimal information and parameters
  • Advanced estimates
    • Allows you to fine-tune the estimate
  • See calculator.aws

Data transfer charges

  • Applies globally to all AWS services
  • Ingress: Free
  • Egress
    • Free within same region, costs extra for Internet and other regions.
    • 💡 Always double check, not so cheap