Introducing the Cloud

Chapter 1
Episode 1.01
Setting up Your Cloud Accounts
• A credit card is used during cloud account setup to verify
identity and process charges if required
• You may be required to use a cell phone and process a security
code that is texted to you
• Some cloud providers stop service before charges incur and
others do not, be careful
Episode 1.02
Defining the Cloud
 Definition
• The NIST definition of cloud
computing:
Cloud computing is a model
for enabling ubiquitous,
convenient, on-demand
network access to shared
pools of configurable
computing resources that can
be rapidly provisioned and
released with minimal
management effort or
service provider interaction.
 Ubiquitous

• Available anywhere…
- …there is an Internet
connection
 Convenient

• No server room
required
• No power provisioning
required
• No hardware
installation required
• Sometimes, no
software installation
required
 On-Demand

• Available when
required
• Scheduled availability
• Created in minutes –
not hours or days
 Shared Pool

• Multi-tenant
- More than one entity
using a shared server
• Automatic
prioritization
 Computing
Resources
• Processing
• Memory
• Storage
• Networking
• Special hardware
Rapidly Provisioned

• Cloud provisioning is
simplified through:
- Single-click launch
- Template-based launch
- Solution-based launch
 Minimal
Management
• Automatic
updates/patch
management
• Integrated monitoring
and reporting
• Automatic scaling
• Scheduled availability
• Cloud computing is ubiquitous and available anywhere Internet
access is available
• Computing resources include processing, memory, storage,
networking, and specialized hardware
• Cloud solutions provide shared pools of resources with multiple
tenants
Episode 1.03
Deployment Models
Cloud Deployment
Models
• Deployed by:
- Others
(public/community)
- Self (private)
- Both (hybrid)
 Selecting a
Deployment Model
• Constraints and
requirements drive
selection
- Examples:
§ Business policies
§ Functional (tasks)

§ Security (compliance)
• Public and community clouds are hosted online by a service
provider
• Private clouds are hosted internally within a company
• Hybrid clouds are some combination of public/community and
private cloud
Episode 1.04
Public Cloud Demo
 Public Cloud

• Used by public
consumers
• Pay-as-you-go/pay-as-
you-grow
• Examples
- Microsoft Azure
- Amazon AWS
- Google Cloud Platform
(GCP)
 DEMO

• Show interfaces for

- Azure
- AWS
- Google Cloud
• Pay-as-you-go/Pay-as-you-grow is provided by public cloud
providers
• AWS is the market leader in public cloud
• Azure and GCP are growing rapidly
• All public cloud providers offer a web-based and command line-
based management interface
Episode 1.05
Private Cloud Demo
 Private Cloud
• Implemented in-house
- Hardware and software
- Everything managed in-house
- Behind a firewall
- Possibly in a DMZ
• Based on virtualization
- Virtual machines to rapidly
deploy servers as needed
• Need extra physical hardware
- Rapid deployment
- Storage space
- Compute
- Services
 DEMO

• OpenStack
• Cloudify
• Private clouds implement a virtualization layer and a
management layer
• Private clouds provide storage, compute and services
• OpenStack and Cloudify are examples or cloud management
software
Episode 1.06
Community Cloud Demo
Community Cloud

• Built for specific groups

- Healthcare
- Science
- Education
- Systems management
(partner cloud)
• Sometimes blurred
with simple SaaS
solutions
 DEMO

• coconstruct.com –
SaaS/Community
Cloud
• Penta.com –
SaaS/Community
Cloud
• Community cloud and SaaS are similar, but SaaS is a superset
of community cloud
• Community clouds exist for healthcare, science, education,
government and more
• Partner clouds are in the category of community clouds as well
• CoConstruct.com and Penta.com are examples of community
clouds
Episode 1.07
Hybrid Cloud Demo
 Hybrid Cloud

• Mixture of public and

private cloud solutions
- Data in the private
cloud, processing in the
public cloud
- Processing in the
private cloud, data in
the public cloud
• Applications/APIs used
to integrate the two
 Hybrid Cloud

• Driving factors
- Cost
- Data
- Security
• Hybrid cloud solution
- Using applications in
the cloud to create data
stored in a private cloud
 DEMO

• CloudBolt.io
• A hybrid cloud is a mixture of public and private cloud solutions
• Applications and APIs are available to integrate public and
private clouds
• Cost, Data, and Security may be driving factors that assist in
selection of a hybrid cloud model
Episode 1.08
Additional Models
 Single Server

• A private cloud run by

a single powerful
server
• A single physical server
in the cloud
• A single virtual server
in the cloud
 Single Cloud

• One cloud provides all

services
• Small and medium
businesses
• Partial security
constraints are not in
place
 Multi-Cloud

• Multiple clouds
- Services
- Departments
- Divisions/companies
- Security constraints
• Orchestration
platforms help manage
multi-cloud
deployments
• Single server cloud is a cloud implementation that uses a
single, usually powerful, server
• Single cloud indicates that only one cloud account or service is
used
• Multi-cloud indicates that multiple accounts or services are used
Episode 1.09
Cloud Architectures
Cloud Architectures

• Software as a Service
(SaaS)
• Platform as a Service
(PaaS)
• Infrastructure as a
Service (IaaS)
• Everything as a Service
(XaaS)
 SaaS

• Software as a Service
- Provides the software in
the cloud
§ Google apps
§ Adobe cloud

§ Microsoft Office 365

§ Salesforce automation

§ Inventory management

§ Project management

- May be API as a Service

(AaaS)
 PaaS

• Platform as a Service
- Simple implementation
of platforms
- Runtimes, modules,
components
- Easy application
deployment
- Often used by
developers and
programmers
 IaaS

• Infrastructure as a
Service
- Complete solution from
hardware up
- VMs and operating
systems
- Network configuration
- Network services
- Ex: AWS, Azure, Google
Cloud Platform
 XaaS

• Everything as a Service
- Networking
- Analytics
- Artificial intelligence
- Device management
- Data extract, transform,
load (ETL) operations
- Ex: DBaaS (Database as
a Service), WSaaS (Web
Server as a Service)
• Examples of SaaS include Office 365 and Salesforce.com
• PaaS includes LAMP and other application development
environments
• IaaS includes networking and security filtering features
Episode 1.10
Advanced Cloud Services
Internet of Things
• Connecting “things” to the network
- May connect to the Internet
• Cloud Service Providers (CSPs) offerings:
- IoT communication protocols
- Application processing
- Database storage
- Aggregate IoT solutions
Artificial Intelligence
• Artificial Intelligence (AI): ability of a machine
to perform tasks that normally require human-
like perception, cognition, decision-making, and
action in a smart, efficient, and effective way
• Machine Learning (ML): subset of AI that
involves the development of algorithms and
statistical models that enable a system to learn
from data and improve its performance on a
task over time
AI Cloud Offerings
• Image recognition
• Text recognition and natural language processing
• Data analysis and prediction
• Generative language models
• Automated decision support
Episode 1.11
Capacity, Elasticity, and Support Agreements
 Capacity
• The workload capability
of a system
- Storage
§ Amount
§ Speed of read/write for I/O
operation concurrency
• IOPS – input/output
operations per second
- Networking
§ Speed of transfer/number
of users
- Processing
§ Speed of workload
processing
 Elasticity

• The ability to expand

and contract as
required
- CPU resources
- Storage
- Servers
- Threads/requests
 Service Model
Maintenance
• Metering mused to
measure cloud resource
consumption
• Chargeback can be made
to departments other
than IT
- Each department is
charged for their IT
resources
• Pay-as-you-grow
- Allows for low cost of
entry
 Service Model
Maintenance
• Demand-driven service
- Service to be
provisioned based on
current demand
 Service Model
Maintenance
• Responsibility
- You manage what you
put in the cloud
- Cloud provider manages
the cloud itself
- Support agreements
§ SLAs
§ Response time

§ Service contact methods

• Capacity is the workload capability of a system and applies to
storage, networking, and processing
• Processing may include CPU and memory
• Elasticity is the ability to expand and contract as required
Episode 1.12
Advanced Capacity Planning Concepts
Capacity Planning
• Define intentions
• Define requirements
• Create a design
• Test the design
• Implement the design
• Validate the design
Requirements Engineering
• Requirement types
- Business
- Stakeholder/User
- Technical
§ Functional
§ Non-Functional
• Business > Stakeholder > Technical (System)
• BRS > StRS > SyRS/SRS
Trend Analysis
• System use changes over time
• Trend analysis explores the trajectory of change
• Allows for prediction of future capacity needs
- Long-term predictions
- Short-term predictions
Standard Templates
• Cloud providers offer prebuilt templates
• Cloud engineers can create templates
• Templates may be deployed in different ways
- Manually
- Scripted
- Automated with scaling features
Licensing Requirements
• Know the license model
- Per user, per machine, per cpu
• Know the license duration
• Ensure scalability
Episode 1.13
Resource Balancing
Resource Balancing

• Provides for the

resources required at
optimal cost
• Several techniques can
be used
- On-demand provisioning
- Auto-scaling
- Hybrid clouds
- Serverless processing
Ideal Resources for
the Cloud
• Public resources
- Ex: your website
• Private scale resources
• Resources used for
distributed work
- Local access to
resources across regions
• Anything requiring
central processing
Poor Resources for
the Cloud
• Anything requiring offline
access
- Possible solution:
synchronize with the cloud
• Some security-related
resources
- Cloud provider might not be
able to adhere to strict
security requirements
• Low-latency demand
resources
- Edge processing may be
better
• Capacity is the workload capability of a system and applies to
storage, networking, and processing
• Processing may include CPU and memory
• Elasticity is the ability to expand and contract as required
Episode 1.14
Change Management
Change Management
Components
• Advisory board
- Usually an expert
approves/rejects changes
• Approval process
- When should something
go through official
approval process?
• Documentation
- Change Management
Database (CMDB)
- Spreadsheet
Change Approval
Process
• Submit a change request
- Description
- Positive reason for the
change
- Possible negative
consequences
• Change reviewed by
advisory board
• Approval/rejection
- Implement or alter
• Advisory boards assist in determining if a proposed change
should be implemented
• All changes should be documented for future reference
• Change approval processes include descriptions, change
justification, potential problems, and review