AWS Cloud Practitioner AWS has over 200+ cloud services
Evolution of Computing
Cloud Concepts
● Dedicated - physical server wholly
➢ Cloud Computing - using a network of utilized by a single customer. A lot of
remote servers hosted on the internet to Wasted Space and a lot of cons.
store, manage, and process data ● Virtual Machines - running virtual
instead of doing it locally. machines on one machine.
○ Most popular in the Compute.
Evolution of cloud hosting Uses a Hypervisor to run the
➢ Dedicated server - On premise. VMs.
expensive, high maintenance, and ○ Still underutilized and has
secured. conflicts in resource sharing.
➢ Virtual Private Server - uses ○ Easy to vertical and horizontal
virtualization, and isolation of resources. scale.
➢ Shared Hosting - cheap, poor isolation, ● Containers - virtual machine running
limited functions. multiple containers
➢ Cloud hosting - multiple physical ○ Docker daemon, the name of the
machines act as one system. Flexible, software layer that lets you run
scalable, secure, and cost effective. multiple containers.
○ Maximize utilization of the
Cloud service Provider does: available capacity, cost effective.
○ More efficient than using multiple
➢ Provides cloud services VMs
➢ Has cloud services chained together ○ Multiple apps run side by side
which creates a cloud architecture. without being limited by OS
➢ Access cloud services through a Single requirements.
Unified API (AWS API) ○ A lot more work to maintain
➢ Utilizes metered billing containers.
➢ Built in monitoring (AWS Cloud Trail) ● Functions - are managed VMs running
➢ Offers Infrastructure as a Service managed containers.
➢ Offers automation via Infrastructure as ○ Known as Serverless Compute.
Code (IaC)
4 Most Common Cloud Services for IaaS: Types of Cloud Computing:
➢ Compute (e.g. EC2 Virtual Machines) - ● Software as a Service (For Customers) -
virtual computer that can run product run and managed by the service
applications, programs, and code. provider (Gmail, Office 365).
➢ Networking (e.g. EBS Virtual Hard ● Platform as a Service (For Developers) -
Drives) - virtual network defining internet Focused on deployment and
connections or network isolations management of apps. Don’t worry about
between services or outbound to the configuring or understanding
internet. hardware/OS.
➢ Storage (e.g. RDS SQL Databases) - ● Infrastructure as a Service (For admins)
virtual hard-drive that can store files. - the building blocks of cloud IT.
➢ Databases (e.g. VPC Private Cloud Networking features, Vms, and data
Network) - virtual database for storing storage. Don’t worry about IT staff and
reporting data or a database for general hardware (Azure, Aws).
purpose web-application.
Cloud Deployment Models:
, ● Public Cloud (Cloud-native/Cloud first) - 1. Trade capital expense for variable
Everything is built on the CSP. expense - Pay on-demand, no upfront
○ Commonly used by companies cost.
starting out. (e.g. Startups, SaaS 2. Benefit from massive economies of
Offerings, New companies) scaling - You are technically sharing the
● Private Cloud (On Premise) - Everything cost with other customers. Thousands of
is built on the company's data centers. customers utilize a fraction of a server.
(e.g. OpenStack). 3. Stop guessing capacity - Scale up and
○ Government, hospitals, insurance down. Launch and destroy services
companies whenever.
● Hybrid Cloud - using both On premise 4. Increase speed and agility - launch
and a CSP. resources within a few clicks in minutes.
○ Companies that started with their 5. Stop spending money on running and
own data center (e.g. Banks, maintaining data centers - allows you to
Large Service Providers, Legacy focus on other important things.
on-premise) 6. Go global in minutes - deploy your app
● Cross-Cloud (Multi-cloud) - using in multiple regions around the world.
multiple cloud providers (E.g. Azure Lower latency.
Arc).
Follow along segment:
Seven Advantages of Cloud - Andrew
● Setting up IAM User, 47 min brown
● Understanding regions and region
exclusive services, 57 min 1. Cost effective - pay for what you
● Brief story about billing consume, no upfront cost.
2. Global - launch workloads anywhere in
Follow along segment: the world, just choose a region.
3. Secure - Physical security is taken care
● AWS Budgets - Protect yourself from of. You can always have control of
accidental over use of services. security to the granular level.
● AWS Budgets - set custom budgets 4. Reliable - Data backup, disaster
that alert you when you exceed recovery, data replication, and fault
budgeted thresholds. tolerance.
● AWS Free tier 5. Scalable - increase or decrease
● AWS Cloud watch - a collection of resources based on demand.
services 6. Elastic - Automate scaling during spikes
● Alarms > Billing - helps monitor charges and drops in demand.
on your AWS bills 7. Current - Hardware and software is
● Setup MFA options patched, upgraded, and replaced by the
CSP without interrupting you.
Computing Power AWS services:
AWS Global Infrastructure - Globally
1. General Compute - Elastic Compute
distributed hardware and data centers that are
Cloud EC2
physically networked together as one large
2. GPU Compute - AWS Inferentiare (Inf1)
resource for the end customer.
3. Quantum Compute - AWS Braket
Made up of:
Six Advantages of Cloud, original
description of cloud benefits by AWS ● 25 launched regions
● 81 availability zones
, ● 108 Direct Connection locations
● 275 Points of Presence
● 11 Local Zone Availability Zones - physical location made up
● 17 Wavelength Zones of one or more data center
● Data center - secured building that
contains hundreds of thousands of
Regions - geographically distinct locations computers.
made up of one or more Availability zones. ● A Region generally contains 3 AZs.
● Data centers within a region are isolated
● Physically isolated from each other and from each other (diff buildings) but close
independent in terms of location, power, enough to provide low latency.
water supply. ● It’s common practice to run workloads
● Generally has three availability zones across at least 3 AZs in case 1 or 2 data
● New services almost always become centers fail, this is called high
available first in US East 1 availability.
● Not all services are available in all ● AZs are represented by a region code,
regions followed by a letter identifier (e.g. us-
● All billing information appears in US east-1a).
East ● A subnet is associated with an AZ.
● Service cost vary per region ● You do not choose the AZ when
launching resources, instead, you
Factors you need to consider when
choose the subnet associated with the
choosing a region:
AZ. Most of the time, this is already
● Regulatory compliance in the region managed by AWS.
● Cost of AWS Services in the region ● US-East-1 region has 6 AZs.
● Services available in the region
Visualization:
● Distance or latency of the region to my
end users
Regional services
● AWS scopes their management console
on a selected region
● Determines where an AWS service will
be launched and what will be seen
within an AWS service’s console
● You don’t explicitly set the region for a
service at the time of creation.
Global Services
● Some services operate across multiple
regions, will be fixed to “global”
● No concept of region for services
available globally.
● In some services (S3 Bucket), a single
region must be explicitly chosen. Follow along segment:
● In some services, you can choose a
group of regions (Globally, North Getting to know services that allow you to
America, etc using CloudFront select AZs, regions, and some geographical
Distribution). locations only.
, ● AZs are redundantly connected to
multiple tier-1 transit providers.
Fault Tolerance (Global Infrastructure)
Multi-AZ for high availability:
What is a Fault Domain?
● When an application is partitioned
● Section of a network that is vulnerable across AZs, companies are better
to damage if a critical device or system isolated and protected from issues such
fails. as power outages, lightning strikes,
● The purpose is that if a failure occurs, it tornadoes, earthquakes, and more.
will not cascade outside that domain,
which limits the damage possible.
● You can have fault domains nested
inside fault domains. AWS Global Network:
Fault Level - Collection of fault domains ● Represents the interconnections
between AWS global infrastructure.
● AWS region would be a fault level ● Commonly referred to as the backbone
● An AZ would be a fault domain. of AWS.
● A private expressway where things can
The scope of a fault domain could be: move very fast between datacenters, in
comparison to traversing the internet
● Specific servers in a rack
instead.
● An entire rack in a data center
● An entire room in a data center Edge Locations - Utilized to get data in and
● The entire data center building out of AWS very quickly.
● CSP is responsible for defining the
boundaries of a domain (except Azure). ● Act as on and off ramps to the AWS
global network.
Each AWS Region is completely isolated from
other regions. This allows great fault tolerance Services that uses Edge locations as an on-
and stability. ramp to quickly reach AWS resources in other
regions by traversing the AWS Global network:
Each AZ is isolated but connected through low
latency links. ● AWS Global accelerator
● AWS S3 transfer acceleration
● Each AZ is designed as an independent
failure zone. Amazon Cloudfront - uses EL as an off-ramp
● A failure zone is AWS describing a fault to provide at the edge storage and compute
domain. near the end user.
Failure Zone: VPC Endpoints - Ensures your resources stay
within the AWS network and do not traverse
● AZs are separated within a typical over the public internet.
metropolitan region and are located in
lower risk flood plains.
● Discrete uninterruptible power supply
(UPS) and onsite backup generation Points of Presence:
facilities.
an intermediate location between an AWS
● Data centers located in different AZs are
region and the end user, this can be a data
designed to be supplied by independent
center or collection of hardware.
substations to reduce the risk of an
event on the power grid impacting more For AWS, a PoP is a data center owned by
than one AZ. AWS or a trusted partner which is utilized by
Evolution of Computing
Cloud Concepts
● Dedicated - physical server wholly
➢ Cloud Computing - using a network of utilized by a single customer. A lot of
remote servers hosted on the internet to Wasted Space and a lot of cons.
store, manage, and process data ● Virtual Machines - running virtual
instead of doing it locally. machines on one machine.
○ Most popular in the Compute.
Evolution of cloud hosting Uses a Hypervisor to run the
➢ Dedicated server - On premise. VMs.
expensive, high maintenance, and ○ Still underutilized and has
secured. conflicts in resource sharing.
➢ Virtual Private Server - uses ○ Easy to vertical and horizontal
virtualization, and isolation of resources. scale.
➢ Shared Hosting - cheap, poor isolation, ● Containers - virtual machine running
limited functions. multiple containers
➢ Cloud hosting - multiple physical ○ Docker daemon, the name of the
machines act as one system. Flexible, software layer that lets you run
scalable, secure, and cost effective. multiple containers.
○ Maximize utilization of the
Cloud service Provider does: available capacity, cost effective.
○ More efficient than using multiple
➢ Provides cloud services VMs
➢ Has cloud services chained together ○ Multiple apps run side by side
which creates a cloud architecture. without being limited by OS
➢ Access cloud services through a Single requirements.
Unified API (AWS API) ○ A lot more work to maintain
➢ Utilizes metered billing containers.
➢ Built in monitoring (AWS Cloud Trail) ● Functions - are managed VMs running
➢ Offers Infrastructure as a Service managed containers.
➢ Offers automation via Infrastructure as ○ Known as Serverless Compute.
Code (IaC)
4 Most Common Cloud Services for IaaS: Types of Cloud Computing:
➢ Compute (e.g. EC2 Virtual Machines) - ● Software as a Service (For Customers) -
virtual computer that can run product run and managed by the service
applications, programs, and code. provider (Gmail, Office 365).
➢ Networking (e.g. EBS Virtual Hard ● Platform as a Service (For Developers) -
Drives) - virtual network defining internet Focused on deployment and
connections or network isolations management of apps. Don’t worry about
between services or outbound to the configuring or understanding
internet. hardware/OS.
➢ Storage (e.g. RDS SQL Databases) - ● Infrastructure as a Service (For admins)
virtual hard-drive that can store files. - the building blocks of cloud IT.
➢ Databases (e.g. VPC Private Cloud Networking features, Vms, and data
Network) - virtual database for storing storage. Don’t worry about IT staff and
reporting data or a database for general hardware (Azure, Aws).
purpose web-application.
Cloud Deployment Models:
, ● Public Cloud (Cloud-native/Cloud first) - 1. Trade capital expense for variable
Everything is built on the CSP. expense - Pay on-demand, no upfront
○ Commonly used by companies cost.
starting out. (e.g. Startups, SaaS 2. Benefit from massive economies of
Offerings, New companies) scaling - You are technically sharing the
● Private Cloud (On Premise) - Everything cost with other customers. Thousands of
is built on the company's data centers. customers utilize a fraction of a server.
(e.g. OpenStack). 3. Stop guessing capacity - Scale up and
○ Government, hospitals, insurance down. Launch and destroy services
companies whenever.
● Hybrid Cloud - using both On premise 4. Increase speed and agility - launch
and a CSP. resources within a few clicks in minutes.
○ Companies that started with their 5. Stop spending money on running and
own data center (e.g. Banks, maintaining data centers - allows you to
Large Service Providers, Legacy focus on other important things.
on-premise) 6. Go global in minutes - deploy your app
● Cross-Cloud (Multi-cloud) - using in multiple regions around the world.
multiple cloud providers (E.g. Azure Lower latency.
Arc).
Follow along segment:
Seven Advantages of Cloud - Andrew
● Setting up IAM User, 47 min brown
● Understanding regions and region
exclusive services, 57 min 1. Cost effective - pay for what you
● Brief story about billing consume, no upfront cost.
2. Global - launch workloads anywhere in
Follow along segment: the world, just choose a region.
3. Secure - Physical security is taken care
● AWS Budgets - Protect yourself from of. You can always have control of
accidental over use of services. security to the granular level.
● AWS Budgets - set custom budgets 4. Reliable - Data backup, disaster
that alert you when you exceed recovery, data replication, and fault
budgeted thresholds. tolerance.
● AWS Free tier 5. Scalable - increase or decrease
● AWS Cloud watch - a collection of resources based on demand.
services 6. Elastic - Automate scaling during spikes
● Alarms > Billing - helps monitor charges and drops in demand.
on your AWS bills 7. Current - Hardware and software is
● Setup MFA options patched, upgraded, and replaced by the
CSP without interrupting you.
Computing Power AWS services:
AWS Global Infrastructure - Globally
1. General Compute - Elastic Compute
distributed hardware and data centers that are
Cloud EC2
physically networked together as one large
2. GPU Compute - AWS Inferentiare (Inf1)
resource for the end customer.
3. Quantum Compute - AWS Braket
Made up of:
Six Advantages of Cloud, original
description of cloud benefits by AWS ● 25 launched regions
● 81 availability zones
, ● 108 Direct Connection locations
● 275 Points of Presence
● 11 Local Zone Availability Zones - physical location made up
● 17 Wavelength Zones of one or more data center
● Data center - secured building that
contains hundreds of thousands of
Regions - geographically distinct locations computers.
made up of one or more Availability zones. ● A Region generally contains 3 AZs.
● Data centers within a region are isolated
● Physically isolated from each other and from each other (diff buildings) but close
independent in terms of location, power, enough to provide low latency.
water supply. ● It’s common practice to run workloads
● Generally has three availability zones across at least 3 AZs in case 1 or 2 data
● New services almost always become centers fail, this is called high
available first in US East 1 availability.
● Not all services are available in all ● AZs are represented by a region code,
regions followed by a letter identifier (e.g. us-
● All billing information appears in US east-1a).
East ● A subnet is associated with an AZ.
● Service cost vary per region ● You do not choose the AZ when
launching resources, instead, you
Factors you need to consider when
choose the subnet associated with the
choosing a region:
AZ. Most of the time, this is already
● Regulatory compliance in the region managed by AWS.
● Cost of AWS Services in the region ● US-East-1 region has 6 AZs.
● Services available in the region
Visualization:
● Distance or latency of the region to my
end users
Regional services
● AWS scopes their management console
on a selected region
● Determines where an AWS service will
be launched and what will be seen
within an AWS service’s console
● You don’t explicitly set the region for a
service at the time of creation.
Global Services
● Some services operate across multiple
regions, will be fixed to “global”
● No concept of region for services
available globally.
● In some services (S3 Bucket), a single
region must be explicitly chosen. Follow along segment:
● In some services, you can choose a
group of regions (Globally, North Getting to know services that allow you to
America, etc using CloudFront select AZs, regions, and some geographical
Distribution). locations only.
, ● AZs are redundantly connected to
multiple tier-1 transit providers.
Fault Tolerance (Global Infrastructure)
Multi-AZ for high availability:
What is a Fault Domain?
● When an application is partitioned
● Section of a network that is vulnerable across AZs, companies are better
to damage if a critical device or system isolated and protected from issues such
fails. as power outages, lightning strikes,
● The purpose is that if a failure occurs, it tornadoes, earthquakes, and more.
will not cascade outside that domain,
which limits the damage possible.
● You can have fault domains nested
inside fault domains. AWS Global Network:
Fault Level - Collection of fault domains ● Represents the interconnections
between AWS global infrastructure.
● AWS region would be a fault level ● Commonly referred to as the backbone
● An AZ would be a fault domain. of AWS.
● A private expressway where things can
The scope of a fault domain could be: move very fast between datacenters, in
comparison to traversing the internet
● Specific servers in a rack
instead.
● An entire rack in a data center
● An entire room in a data center Edge Locations - Utilized to get data in and
● The entire data center building out of AWS very quickly.
● CSP is responsible for defining the
boundaries of a domain (except Azure). ● Act as on and off ramps to the AWS
global network.
Each AWS Region is completely isolated from
other regions. This allows great fault tolerance Services that uses Edge locations as an on-
and stability. ramp to quickly reach AWS resources in other
regions by traversing the AWS Global network:
Each AZ is isolated but connected through low
latency links. ● AWS Global accelerator
● AWS S3 transfer acceleration
● Each AZ is designed as an independent
failure zone. Amazon Cloudfront - uses EL as an off-ramp
● A failure zone is AWS describing a fault to provide at the edge storage and compute
domain. near the end user.
Failure Zone: VPC Endpoints - Ensures your resources stay
within the AWS network and do not traverse
● AZs are separated within a typical over the public internet.
metropolitan region and are located in
lower risk flood plains.
● Discrete uninterruptible power supply
(UPS) and onsite backup generation Points of Presence:
facilities.
an intermediate location between an AWS
● Data centers located in different AZs are
region and the end user, this can be a data
designed to be supplied by independent
center or collection of hardware.
substations to reduce the risk of an
event on the power grid impacting more For AWS, a PoP is a data center owned by
than one AZ. AWS or a trusted partner which is utilized by
