original file

Module 6 Section 1: Compute Services Overview

i came late to this class, he was already talking about CT Templates under Amazon Compute Services

Amazon Compute Services

• there are many services to support different user needs

note: managed vs unmanaged

• managed (parts are managed by the cloud provider)
• unmanaged (everything is managed by you)

bolded - covered in the module

1. Amazon EC2 - VMs
2. Amazon EC2 Autoscaling
3. Amazon Elastic Container Registry (Amazon ECR) - Container Templates, repository of many different kinds of containers, fully managed service that takes care of different kinds of containers
4. Amazon Elastic Container Service (Amazon ECS) - Containers, fully managed orchestration service, helps run and manage containerized applications
5. VMware Cloud on AWS
6. AWS Elastic Beanstalk - Orchestration
7. AWS Lambda - Serverless compute (just run your code)
8. Amazon Elastic Kubernetes Services (Amazon EKS) - Open source Container Product, Fargate is a product from AWS, EKS lets you use the open source service already
9. Amazon Lightsail - Web app
10. AWS Batch
11. AWS Fargate - managed containers, serverless compute engine for containers, we just manage the container itself nalang
12. AWS Outposts - on premises
13. AWS Serverless Application Repository

Categorizing Compute Services

• services are "options" for you to choose from with each one having its pros and cons
• depending on the skillset level and use-case

1. VMs
2. Serverless
3. Containers
4. Orchestration

Choosing the Optimal Compute Service

• there are many services to choose from, so choose the option that best fits your needs
• the optimal compute service or services that you will use will depend on your use case
• some aspects to consider:
  • what is your application design?
  • what are your usage patterns?
  • which configuration settings will you want to manage?
• selecting the wrong compute solution for an architecture can lead to lower performance efficiency
  • a good starting place - understand the available compute options

if you're happy using cloud → you use more

Module 6 Section 2: Amazon EC2 (VMs)

11:03

Amazon Elastic Compute Cloud (Amazon EC2)

• VMs are the closest to traditional services
• Example uses:
  • application server, web server, database server
  • game server, mail server, media server, catalog server, file server
  • computing server, proxy server

Amazon Elastic Compute Cloud (Amazon EC2) Overview

• VMs have full control, but also have more responsibilities
• provides virtual machines (EC2 Instances) in the cloud
• gives you full control over the guest operating system (windows or linux) on each instance
• you can launch instances of any size into an Availability Zone anywhere in the world
  • launch instances from Amazon Machine Images (AMIs)
  • launch instances with a few clicks or a line of code, and they are ready in minutes → speed and agility
• you can control traffic to and from instances via security groups

EC2/VMs: Selecting an Instance Type

• instance types are machine categories (how will the EC2 instance you create be used?)
• if you have specific needs, you can choose a category that fits your use case
• the instance type that you choose determines:
  • memory (RAM)
  • processing power (CPU)
  • disk space and disk type (storage)
  • network performance
• instance type categories (for AWS, categories depend on the service provider):
  • general purpose
  • compute optimized
  • memory optimized
  • storage optimized
  • accelerated computing
• instance types offer family, generation, and size

EC2/VMs Instance type naming and sizes

• refers to how much resources
• example: t3.large
  • t is the family name
  • 3 is the generation number
  • large i the size
    • how many vCPU cores, how many GB of memory

Select instance type based on use case

• based on use case → what your generation needs

EC2/VMs: User Data Script (Optional)

• you can run scripts on create such as download apache and git pull your web code
• optionally specify a user data script at instance launch
  • great for initial setup
• use user data scripts to customize the runtime environment of your instance
  • script runs the first time the instance starts
• can be used strategically

EC2/VMs: Specify Storage

• configure the root volume (like C:/)
  • where the guest operating system is installed
• attach additional storage volumes (like D:/) (optional)
  • AMI might already include more than one volume
• for each volume, specify:
  • the size of the disk (in GB)
  • the volume type
    • different types of SSDs and HDDs are available
  • if the volume will be deleted if the instance is terminated
  • if encryption should be used
    • encryption is a choice: so if you choose not to have encryption and something bad happens, that's on you; user choice means user responsibility

EC2/VMs: storage options

there are two options for the root disk storage option

• Amazon Elastic Block Store (Amazon EBS) - separate cost from your VM
  • durable, block-level storage volumes (like a hard disk or hard drive, overwrites blocks of storage)
  • you can stop the instance and start it again, and the data will still be there (non-volatile but with payment)
• Amazon EC2 Instance Store - already part of a VM so no added cost
  • ephemeral/temporary storage (volatile and good for temporary instances) is provided on disks that are attached to the host computer where the EC2 instance is running
  • if the instance stops, data stored here is deleted
• Other options for storage (not for the root volume)
  • mount an Amazon Elastic File System (Amazon EFS) file system - network storage (good for scaling and shared data)
  • connect to Amazon Simple Storage Service (Amazon S3) - object storage for unstructured data

EC2/VMs: Example Storage Options

• Instance 1 characteristics:
  • has an Amazon EBS root volume type for the operating system
  • what will happen if the instance is restarted?
    • data will retain
• Instance 2 characteristics:
  • has an Instance Store root volume type for the operating system
  • what will happen if the instance stops because of a user error or system malfunction?
    • data will not retain

EC2/VMs: Console view of running EC2 instance (web GUI aka management console)

EC2/VMs: launching an EC2 instance with the AWS Command Line Interface

• the web GUI/management console is good for 1-2 VMs but if you need thousands it's hard to run the wizard so many times, so use the CLI instead
• EC2 instances can be created programmatically
• generates all the needed instances for you without needing to do it manually

EC2/VMs: Amazon EC2 instance lifecycle

• machine states
  
• only VMs with block storage can be stopped
• when you turn ephemeral storage ones it's deleted

EC2/VMs: Check your understanding

• managed service → managed by the cloud provider
  • you have less control but it's easier on you because you have less responsibility
• what's the best option? → the answer is always that it depends
• managed service is like outsourcing. a VM is normally cheaper if you already have manpower and technical skills

Module 6 Section 3: Amazon EC2 Cost Optimization

30:26

• TLDR: cloud is expensive
• you have to optimize your cost so you can maximize what you pay for

Amazon EC2 Pricing Models

• there are many options for different use-cases

• On-demand Instances

  • pay by the hour
  • no long-term commitments
  • eligible for AWS free tier
  • stop guessing capacity but very expensive

• Dedicated Hosts

  • a physical server with EC2 instance capacity fully dedicated to your use
  • good for compliance, rent the whole host server

• Dedicated Instances

  • instances that run in a VPC on hardware that is dedicated to a single customer
  • good for isolation

• Reserved Instances

  • full, partial, or no upfront payment for instance you reserve
  • discount on hourly charge for that instance
  • 1-year or 3-year term (long term with discounts)

• Scheduled Reserved Instances

  • purchase a capacity reservation that is always available on a recurring schedule you specify
  • 1-year term
  • reservation but with schedules
    • if you don't need it 24/7

• Spot Instances

  • instances run as long as they are available and your bid is above the Spot Instance Price
  • they can be interrupted by AWS with a 2-minute notification
  • interruption options include terminated, stopped, or hibernated
  • prices can be significantly less expensive compared to On-Demand Instances
  • good choice when you have flexibility in when your applications can run
    • if you need high compute with the lowest cost and you are not in a hurry

Per second Billing is available for On-Demand Instances, Reserved Instances, and Spot Instances that run Amazon Linux or Ubuntu

Amazon EC2 Pricing Models: Benefits and Use Cases

The 4 pillars of cost optimization

practices to help reduce cost

1. Right size
2. Increase elasticity
3. Optimal Pricing Model
4. Optimize storage choices

Pillar 1: Right Size

• determine your baseline by monitoring and observing

• right size means resources are not wasted and the cost is manageable

• provision instances to match the need

  • CPU, memory, storage, and network throughput
  • select appropriate instance types for your use

• use Amazon CloudWatch (monitoring) metrics

  • how idle are instances and when?
  • downsize instances

• best practice: right size, then reserve (since Reserved Instances are cheaper)

Pillar 2: Increase Elasticity

• scaling will support the changes or sudden resource spikes

• stop or don't use the machines you don't really need right now

• stop or hibernate Amazon EBS-backend instances that are not actively in use

  • ex. non-production development or test instances

• use automatic scaling to match needs based on usage

  • automated and time-based elasticity

Pillar 3: Optimal Pricing Model

• you can use multiple pricing models at the same time

• leverage the right pricing model for your use case

  • consider your usage patterns

• optimize and combine purchase types

examples

• use On-Demand Instance and Spot Instances for variable workloads
• use Reserved Instances for predictable workloads
• consider serverless solutions like AWS Lambda

Pillar 4: Optimize Storage Choices

• storage is not cheap

• reduce costs while maintaining storage performance and availability

• resize EBS volumes

• change EBS volume types

  • can you meet performance requirements with less expensive storage?
  • ex. Amazon EBS Throughput Optimized HDD (st1) storage typically costs half as much as the default General purpose SSD (gp2) storage option

• delete EBS snapshots that are no longer needed

• identify the most appropriate destination for specific types of data

  • does the application need the instance to reside on an Amazon EBS?
  • Amazon S3 storage options with lifecycle policies can reduce costs

Cost Estimation Summary: Measure, monitor, improve

• cost estimation a continuous, ongoing process
• define and enforce cost allocation tagging
• define metrics, set targets, and review regularly
• encourage teams to architect for cost
• assign the responsibility of optimization to an individual or to a team

recording ended here

• rest of the slides are more in-depth about compute types