CLOUD

NETWORKING
 Meet / House Keeping

O Lecturer, Students

O TEACHING / LEARNING STRATERGY: Slides (for

Presentation & Lectures)
O Practical Sessions
O Discussions
O Assignments

2
 Overview
O Basic concepts, terminologies of
O Cloud Computing
O Edge Computing
O Distributed cloud Network

3
 Virtualisation
O Fairly old technology but its very relevant to cloud computing technology
O Its the process of creating a software (virtual) based version of sometime – network , Computer.
O It needs the piece of *software* called hypervisor that runs above the physical server/host
and pulls the resources from the physical server and allocates to the Virtual environments
O Types :
O Type 1 (Bare-metal ) more secure directly on the physical server, lower latency) eg
VMWare, Microsoft HyperV
O Tyep 2 (Hosted) over a host OS higher latency eg Oracle, Virtualbox
Spin up different Virtual Machines can have different OS, “flexibility” and portable (can be
moved)
O Benefits
O Cost reduction / savings (multiple servers on one, maintence and electricity)
O Speed and agility (virtualization makes it faster and quicker to deploy a machine for a test
scenario )
O Lower downtime with a backup.
O Draw-backs
O Resource scaling (static allocation)
O Auto-provision (high traffic)
O Multi-tenancy (single machine, single user)
4
Cloud
Computing

5
 What is Cloud Computing?
O Basically any computing model/paradigm that provides convenient, on-demand network access
to a shared pool of configurable resources (networks, servers, storage, applications) with
minimal management effort or interation from service provider

O Compute services offered may vary but must have the following essential / basic
Characteristics:
O On-demand self-serice
O Broad network access: mechanisms that allow any kind of clients (eg mobile phones,
tablets, laptops or workstations)
O Resource pooling
O Rapid Elasticity
O Measured Service: Leveraging on metering, resource usuage can be monitored, controlled
& reported this also provides transparency for both provider and consumer.

O The cloud (service) provider is responsible for

O The physical hardware required
O Keeping it up to date
O Secure (both at site and through access)

O The goal of any cloud services is to make business easier and more efficient no matter the
size or needs. 6
 Components
O self- service
O on-demand
O sharing / multi-
tenancy

Very importantly CC
is self-service,
flexible and cost -
efficient.
7
 Cloud Stakeholders
O End-User: Usually the end-user is the consumer of the
reliable and available cloud services and only pays for
services that have been used.
O Cloud Service Provider: This is usually the provider of
the virtualisation, resource cloud services who
provides services through
pooling and service level agreement enforcement to
multi-clients.

O Cloud Application Vendor: Usually sells the on-demand

cloud applications to end-user.
O C l o u d Tool P r o v i d e r : Provides c l o u d s u p p o r t a n d
manageability tools to do accounting, monitoring and
usage reporting. 8
Cloud Computing Compute Resources
O Two main Compute Resources
O Virtual Machine : emulator of a computer,
Operating System and H/W that appears
Physical but you don’t buy the h/w or O/S,
it shared securely and isolated
O Storage : Devices and applications read
and/or write data (eg buying tickets,
leaving voicemail, etc) cloud allows to scale
up or down to meet needs. Only pay for
what you use
9
 Benefits of Cloud Computing
O Elasticity: service capacities can change
rapidly
O Scalability : resources grow on users’
demand
O Reduced Cost: only pay for services rendered
or used
O Availability / Internationalization: clients are
sure of access to resources all the time
(24/7).

10
 Benefits of Cloud Computing II

O Portability
O Level Playing Ground
O Resource Pooling and Auto-Provision
O Multi-tenancy
O Security
O Compliance / Standardization
O Its Current : Patches and Upgrades
O Its Reliable: offer data backup, disaster
recovery, and data replication
11
 Service Delivery Model
There are different types of
cloud computing models
through which s e r v i c e s are
delivered but the common
ones are:

O Infrastructure as a Service (IaaS)

O Platform as a Service (PaaS)
O Software/Application as a Service
(SaaS/AaaS)
O “X’ as a Service (XaaS)

12
Cost and Ownership

13
Management Responsibilty

14
 Deployment Model
Private Cloud : resources are hosted privately to
Cloud computing
O
serve just one client
deployment models O Public Cloud : resources are hosted are usually hosted
are based on by a commercial service provider and it serves the
general public
location.
O Community Cloud: resources hosted here are
usually deployed and shared by communities
with the same mission, needs and requirements
O Hybrid Cloud: a combination of two or more
cloud deployments.

15
 Private Cloud

Advantages
O Support any scenario or legacy
application
O Control and responsible for security
O Meet strict security, compliance ore
legal requirements

Disadvantages
• Initail setup and maintenance cost
• Reasonal scaling would cost more
• IT skills and expertise that are hard to
come by

16
 Public Cloud

Advantages
O High scalability at minimal cost
O Pay as you go pricing
O Not responsible for maintenance and ugrades
O Minimal technical knowledge, leverages on the skills of the provider

Disadvantages
• Specific security requirements that may/can not be met by the
provider
• Govt policies, industry standards or legal requirements that may
not be met
• Cannot manage the hardware or services as you may want
• Unique business requirement may be hard to meet – legacy 17
application
 Hybrid Cloud

Advantages
O Legacy systems and application can still run locally
O Flexibility what goes to the public and what remains on site locally
O Supplement with cloud providers where cheaper and when its not
use own equipment and resources
O Meet security and legal compliance with own equipment providing
control of the environment

Disadvantages
• Complicated set-up and management
• Maybe more expensive – set up cost is there

18
 Challenges of Cloud Computing

O Legal Issues
O Security
O Cloud Governance and Control
O Standards / Interoperability
O User Interface and Access
O Lack of National Data Usage / Protection Policies
O National Infrastructire, Political and Legal
Problems
O Cost and availability of infrastructure / bandwidth

19
 Compliance Terms and Requirements

Some questions to ask about a potential provider

include:
O How compliant is the cloud provider when it comes to
handling sensitive data?
O How compliant are the services offered by the cloud
provider?
O How can I deploy my own cloud-based solutions to
scenarios that have accreditation or compliance
requirements?
O What terms are part of the privacy statement for
the provider?

20
Comparison of Cloud Service Providers
O Interface: this describes the type of interface, if it’s a web or
command interface. How easy can users interact with the system
through the interface and can it be customized?
O Performance / Scaling Latency: this can measure the latency
between time of booting through to provision of instance
requested.
O Service: users have different needs so the cloud services would be
according to their different needs. For example, is it cloud
infrastructure, cloud software?
O Cost: What would be paid for and what is free? How is pricing done?
Support / Agreements: what kind of support is available and how
often,
O is it flexible in the packages? What type of agreement is available?
O Data: where is it hosted? How is it backed up?
O Security / Control: how much security is in place and is there 21
some access control implemented according to roles?
 Assignment 1

O In no more than 500 words

discuss the following
• Would Cloud Computing affect
how data centres are built?
Submission : Dec 18th 2023
11pm GMT +1
22
Edge
Computing

23
 Edge Computing
Bringing enterprise applications closer to
data sources

By NoMore201 - Own work, CC BY-SA 4.0, 31

https://commons.wikimedia.org/w/index.php?curid=82034067
 Edge Computing – Definition

1. According to Wikipedia
Any computer program that delivers low latency
nearer to the requests

2. According to Karim Arabi

All computing outside the cloud happening at the edge
of the network, and more specifically in applications
where real-time processing of data is required

25
 Why Edge Computing

O Unprecedented scale and complexity of Data because of IoT

O Bandwidth and latency issues (Improved response times and saves

bandwidth)

With edge Computing Data is processed and analyzed closer to source

O Faster response times and improved customer experiences

O Untapped data potential can be unlock

26
 With Edge Computing

O Lower latency as volumes of data, traffic and distance is reduced

O Better computation on real time application

O Better Algoritm over time with frequent updates

O Better Analysis on Big Data

O Security breaches can compromise sensitive data and put trade

secrets or other proprietary corporate information at risk are
minimised.

Edge computing lets you 'bring the math to the data’…….(IBM)

27
 Distributed Wireless Cloud

public cloud services dispersed across various

geographic locations that are chosen to meet
application requirements, but the operations and
governance of those services are still centrally
controlled by a single public cloud provider………(IBM)

28
 Distributed Cloud Computing

O Though public cloud is seen as reliable, secure and cost-

effective for some scenarios, some non-tech industires such
as health and banking have slowly adopted the cloud.

O Reasons
O Data locality requirement (healthcare and banking)
O Low latency requirement (media and telecommunications)

O Distributed cloud enables a geographically distributed but

centrally managed distribution of public cloud services to
optimised for performance, compliance and edge
computing.

O Simply put Distributed cloud is a public cloud computing

service that lets you run a public cloud infrastructure in
multiple different locations. And allows the client manage
everything from a single control plane (User Interface – UI) 29
 Distributed Cloud Computing

Distributed cloud is a public cloud computing service that lets you run public
cloud infrastructure in multiple locations—your own cloud provider's data
centers, other cloud providers' data centers, third-party data centers or
colocation centers, and on-premises—manage everything from a single
control plane.

O Organisations may manage their resouces using hybrid cloud

(private/public, on-prem or edge)
O You would need to consider : expertise, locality, latency,etc
O Distributed cloud extends the capabilities of the public cloud by
allowing the client manage all these scenarios/environments using the
public cloud
O Distributed cloud aims to provide organisations/customers with
O Consistency in deploying clusters across different locations
O Efficiency in Devops with the option to locate their Databases across different locations
O Innovation with AI and ML capabilities / deployments
O Governance as it relates to data regulations, compliance, locality
O Basically tackles the issues that hybrid cloud may run into
30
 Distributed Cloud Computing

Clients can then deploy and run applications or individual

application components in a mix of cloud locations and
environments that best meets their requirements for
performance, regulatory compliance, and more………thereby
also helping companies surmount the challenges of complying
with country- or industry-specific data privacy regulations
(IBM) 31
 How Distributed Cloud Works

Basically it works by
• Distributing the public cloud compute resources of one
provider across various locations of the client’s choice
that may or may not belong to the provider

• Extends the provider’s centralized cloud geographically

while the provider still retains central control over
control, operations, security,etc

• The client /customer access everything wherever the

location as a single cloud and manges it from single control
plane.

Distributed Cloud makes edge computing management a lot

easier – from a single control plane client can make updates,
manage performance of multiple edge deployments.
32
 Benefits of Distributed Cloud

Distributed Cloud provide Organisations with the following

benefits:

O Reduced Latency / Faster Content Delivery: storing and

delivering content from locations closer to end-users
O Improved hybrid visibility and manageability: organisations have
greater control over its hybrid cloud infrastructure
O Efficiency, cost-effective : using same tools and personnel,
clients can develop and deploy anywhere.
O Easier localised regulatory compliance: processing data at source
would simplify data privacy as DC makes it easier to process
data at country of data location.
O IoT applications : DC and EC together would provide lower
latency to these applications.

33
 Execrise 1

IBM Cloud Satellite example

https://www.ibm.com/blog/understanding-distributed-cloud-
architecture-the-basics/
Further read as IBM Distributed Cloud compares to other vendors
https://www.ibm.com/blog/understanding-distributed-cloud-
architecture-differences-between-vendors/

Building an IBM cloud for free https://cloud.ibm.com/registration

34
 Fog Computing
Fog computing is the distribution of IT resources accross
network elements spanning from the cloud to the edge
[15]

35
 What Fog Computing?

A form of distributed computing that complements/extends cloud 36

computing
 Comparisons

https://www.scalecomputing.com/resources/edge-
computing-vs-fog-computing-vs-cloud-computing

For personal reading by the students

37
 Difference between Cloud, Fog
and Edge Computing
Location of Data Processing :
Cloud on a central server located far away from source of information
Edge directly on the devices to which the sensors are connected
Fog shifts edge computing to the processors connected on the LAN hardware

Processing Power and Storage Capabilities

Cloud provides superior and advanced processing capabilities
Less with Fog and Lesser in Edge computing

Purpose
Cloud more suited for long term indepth analysis of data
Fog and Edge more suited for quick analysis, real-time response.

Security
Where security is a major concern, Fog and Edge are preferable.

38
 Summary
Can cloud alone keep up with the high influx of data? Is it even
necessary to send everything to the cloud? Is there was a way of
selectively storing data on the cloud?

The growth of the I/IoT has increased the need for edge, fog, and
cloud platforms.

The I/IoT is composed of edge, fog and cloud architectural layers, such
that the edge and fog layers complement each other. Fog computing
uses a centralized system that interacts with industrial gateways and
embedded computer systems on a local area network, whereas edge
computing performs much of the processing on embedded computing
platforms directly interfacing to sensors and controllers.

39
Thank-You
 References
1. https://www.javatpoint.com/community-cloud
2. NIST, “The NIST Definition of Cloud Computing.” [Online]. Available:
https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication80
3. 0-145.pdf

4. Virtualisation Explained :
5. https://www.youtube.com/watch?v=FZR0rG3HKIk&feature=emb_rel_pause
6. https://www.youtube.com/watch?v=r3EDlvfQAsw
7.
8. Edge Comuting made easy
9. https://www.youtube.com/watch?v=kLYevrIn79s
10. https://www.ibm.com/cloud/what-is-edge-computing
https://www.intel.com/content/www/us/en/edge-computing/edge-cloud.html
11.
12. https://mediacenter.ibm.com/media/1_on1n95xj
https://www.ibm.com/cloud/blog/distributed-cloud-vs-hybrid-cloud-vs-multicloud-vs-
13.
edge-computing- part-2
14. https://www.intel.com/content/www/us/en/edge-
15. computing/overview.html
16. https://www.sdxcentral.com/cloud/definitions/what-is-distributed-cloud-
definition/

48