Cloud Computing:
The way of accessing on-demand IT resources like computing, storage, networking,
database, virtualization and other resources, provided by the cloud service
provider, over the internet with pay-as-you-go basis is called cloud computing.
Instead of buying, installing and maintaining our own servers or data centers, we
used to get the servers virtually from the cloud service providers with some
additional services and we're gonna pay for what we use.
cloud service providers are those, who provide their infrastructure or other
services to us for our usage or building our applications etc.,
Let's see what is cloud and why cloud?
Cloud: The physical location of the infrastructure of those cloud service
providers, whom services we're using.
Why cloud?
Answer is simple, as already said, instead of buying, installing and maintaining
our own servers or data centers, we use the services offered y cloud service
providers and pay for what-we-use.
To get in depth detail of why cloud, we need to know what are the issues of having
our own data center, which is commonly referred as Traditional datacenters or On-
premises data center.
If we need a traditional or on-premises data center, we need
a) huge area of land to setup the data center.
b) to buy our own servers, storage devices, networking components, databases, more
hardware and software components.
c) to transport them to our local site, where data center needs to set up.
d) to install them on our own, in our local site.
e) to do maintenance of the servers and other components whenever needed.
f) to run them regularly to avoid this maintenance, even if there is load or no
load.
g) more servers when the load is increased and more storage to store the daily logs
etc.,
h) backup storage and servers to work during the unwelcomed events like outages and
need an immediate assistance for disaster recovery.
i) to have additional IT team to help us during such inconvenience.
j) human power or human resources to take care of all these.
k) additional security to be purchased to secure the data inside servers.
l) to work from the same data center every time, we cannot shift them to another
place as it costs additional charges for the company.
And with our own data centers, it
m) is difficult to upgrade the servers or components regularly as per the fast
growing world.
n) had complex connections between the resources.
l) takes much time to launch or install one fully working server to serve for our
application purpose.
Characteristics of on-premises data centers:
Characteristics are those that all above stated points are relatively brought
together and gives a meaningful structurally formed name to it.
1) Infrastructure: As said, We need huge area to setup our own data center, which
is comprised of many hardware and software components like servers, storage racks,
databases etc., which had complex connections between the resources.
2) Scalability: We never know, how the load gets varied, we might see a sudden
spike or might see a drop in the load. So, to meet them, we need to increase the
number of servers running during spike and need to reduce the running servers
during the load fall. It is more complicated in case of on-premises data center to
meet the expectations on-time. Similarly storage, databases and other resources
should act accordingly, which is impossible in on-premises case.
3) Accessibility: There are certain limitations in case of on-premises data
centers, as we need to use only the existing resources and platforms and cannot
wait for new ones and limitations in the working environment, we cannot work from
the other data centers or we cannot access the data of other data centers, even if
they are located nearby.
4) Security: Needs separate security systems based on the hardware and software
resources. For example, Servers need separate security when compared to databases
and also to control the incoming and outgoing traffic, there needs some secured
connections and an additional firewall for instant security patches and to avoid
the risks from malware or suspicious activities by the unknown resources. Also
needs a complete secured platforms to operate on-premises data centers.
5) Disaster recovery and backup: When there are any unwelcomed events occurred in
or near on-premises data centers like outages, cyclones, thunderstorms etc., there
are high chances of data loss and components failure, in such cases, there needs
immediate assistance and need to redirect the traffic to new servers, databases or
storage devices and with new connections, which is time taking process.
6) Human resources: In order to install or maintain servers or to work with the
load variations and in case of security to on-premises data centers, we need
expertise human resources available all the time. And also need an additional IT-
team to work or deal with clients or the users of our data centers.
7) Cost model: As all above things from Owning a land to the complete
infrastructure setup is involved with cost like buying, transporting, installing,
maintenance and owning additional resources is all linked with the money or cost.
So, to have a on-premise data center, there needs a huge investment. And also needs
operational expenditure of the data center like human resources, disaster recovery,
upgrading to new technology etc.,
Challenges faced with the on-premises data centers:
All the characteristics above said is most important for any data center and each
one is a challenge for on-premise data center. For example, If we say,
infrastructure, we need to look for the servers, storage and databases that costs
less and best performed.
And if we say, scalability, they must be scalable and be ready to face the
situations with the sudden variations in the load. And coming to accessibility,
they must have limitations to avoid any data concurrency and inconsistency and also
they have no chance of getting new components as it may leads to additional cost.
And if we say, security, they must be aware of all types of attacks that cause
damage to their functionality and that cause data loss. And coming to disaster
recovery and backup, they need to be ready with additional measures in case of the
un welcomed events.
And if we say, human resources and cost model, obviously every event of the on-
premises data center needs an human resource availability and all the events are
related to the expenditure.
Apart this, there are some more challenges to be faced by the on-premises data
centers. They are:
1) Globalization: As per the changing world and growing technology, the data
centers need to shape their platform accordingly to provide their services all over
the world. For this, they need to setup new data centers across the world.
2) Aging data center: As we said, the rapid change in technology, we couldn't rely
on old data centers to perform well in the growing market. There must be new data
centers to come in with upgraded machinery and upgraded technology to provide
better service to their clients and users.
3) Storage growth: As usage of the application and operating the data center keeps
on increasing, the storage also gets increased and to get rid of this, they need to
compress the old existing data or delete them or to buy new storage racks or
devices.
4) Application explosion: As new trend in users makes the old data centers search
for new platforms and new software to upgrade the existing ones to offer new
services and also need to search for new clients in order to expand their
businesses by exposing their new services offered.
5) change of ownership and cost of ownership: No one can maintain the data center
for ages and ages, there will be new owners coming in to maintain and operate the
data centers and change the existing system. In order to do so, the new owners buy
the data centers from the old owners as a direct dealership from bidding or
auctions or in indirect way like the second or third party ownership. For example,
two vendors invest together on same data center or government owns data center and
other people operates it and so on.
6) Acquisitions: Acquisition refers to the asset that one have or the learning or
developing skill that one have. As per the growing technology, there must be a
quality of developing their skills in the resources available in the data centers
and then only they become the assets of them or else, the human resources become
burden for them with the old technology and make it as a liability.
-----------------------------------------------------------------------------------
--------
Now, let's see in depth detail of why cloud?
Since, we are using the services, required for our application or software to run,
provided by the cloud service provider, we need not
a) require the huge area to setup a data center.
b) No data center, no requirement of buying servers or hardware and software
components of complex connections and no need to transport them and no need of
installing them.
c) to run the servers continuously and no need of maintenance.
d) to worry about varying loads, as you can scale-up or scale-down the resources as
per requirement.
e) to worry about the backup and disaster recovery during unwelcomed events as
cloud service provider take care of it.
f) to have additional IT team and need not require more human power or human
resources.
g) to worry about any additional security, it is also taken care of service
provider.
h) to access from the same place every time, as we can access it globally from
anywhere.
when we use cloud, it
a) is easy to launch any service offered by service provider, just with few clicks
and in less time.
b) offers many services to work with and at better pricing.
c) can be accessed from anywhere in the world and can connect to any data center,
available globally via internet.
Characteristics of cloud computing:
a) On-Demand self service: We need not any additional help from another person to
use the services offered by the cloud service providers, need not to wait for the
provision or allocation of resources, you can create them on your own. If we have
internet, we can create an account in some of the cloud service provider platforms
and use the services offered by them and can use any service with few clicks and
can launch the applications.
b) Broad Network access: With few clicks, we can use any service, offering by the
service provider over the internet from anywhere and from any device, connected to
the internet, over the world and from any service provider.
c) Resource pooling: cloud providers share their physical and virtual resources
among number of cloud users and so, the resources are dynamically allotted to the
customers, based on the demand.
d) Rapid Elasticity: As per the future requirement, we can scale-up or scale down
the resources, whenever needed and we can use the auto scaling property to automate
this scaling process.
e) Measured service: No prior payment is required to use any service, you can pay
for what you use and can start and stop the usage whenever you prefer. You can
measure the services that you used and what's the remaining and can work
accordingly or can scale it accordingly.
cloud v/s on-premises:
|--------------------------------------------|-------------------------------------
-------|
| Cloud | On-premises |
|--------------------------------------------|-------------------------------------
-------|
| No Physical infrastructure | Need to buy, install and maintain the
physi|
| | -cal infrastructure. |
|--------------------------------------------|-------------------------------------
-------|
| No capital expenditure and spends only on | High capital and operational
expenditure. | | operational expenditure. |
|
|--------------------------------------------|-------------------------------------
-------|
| Accessible over the internet and from anywh| Not accessible via internet and must
work | | -ere in the world. | within the premises of data
center. |
|--------------------------------------------|-------------------------------------
-------|
| Measured services, you can frequently check| services cannot be measured
frequently as |
| the usage and act accordingly. | it takes much time and cost. |
|--------------------------------------------|-------------------------------------
-------|
| No prior payment and payment is done after | Prior payment is required and will
have to |
| usage and will pay only for the services yo| pay for all the services in on-
premises. |
| u use and for the time you used. | |
|--------------------------------------------|-------------------------------------
-------|
| Rapid elasticity, can scale-up or scale dow| It is not easy to scale up or scale
down |
| -n the resources easily, whenever required.| the resources whenever needed.
|
|--------------------------------------------|-------------------------------------
-------|
| Speed and agile, can provision the resource| Slow and not agile, you need to wait
until |
| -s whenever required and of any capacity. | you get the required resource
provisioned |
| | i.e., It get released by the other user. |
|--------------------------------------------|-------------------------------------
-------|
| No additional security is required. | You have to look up for additional
security|
| | to protect your data. |
|--------------------------------------------|-------------------------------------
-------|
| It is easy to take backup and easy for disa| It is not that much easy to take
backup, as| | -ster recovery. | they require additional storage
and databas| | | -es. And it's not easy in case of
disaster | | | recovery. |
|--------------------------------------------|-------------------------------------
-------|
| Focus only on innovations and application | Not easy to focus on innovations and
applic|
| development. | -ation development. |
|--------------------------------------------|-------------------------------------
-------|
Cloud Service Models:
______________________
Cloud service models refers to the different types of services offered through the
cloud and are classified into three main models. They are:
1) Infrastructure as a service,
2) Platform as a service,
3) Software as a service.
Commonly they are known as Iaas, Paas, Saas.
Infrastructure as a Service: Cloud service providers provide their computing
infrastructure like virtualization, computing servers, storage and networking as
services over the internet for hosting our websites or enterprise applications.
For example, Amazon EC2 is a computing server offered by AWS.
Characteristics of Iaas:
Scalability: We can scale up or scale down the resources as per our requirement.
Cost effective: Since we are using virtualization technique, we need not to spend
more on the physical infrastructure.
Control: We will be having control over our resources and over deployed or
deploying applications.
Platform as a Service: Cloud service providers provide a platform and computing
resources allowing customers to focus on developing, testing and deploying their
applications without worrying about the underlying infrastructure.
AWS Elastic Beanstalk, Microsoft Azure services are the best example of Paas.
Characteristics of Paas:
Scalability: Built in scalability to handle varying loads.
Abstracted Infrastructure: Customers focus on their developing applications without
underlying on the physical infrastructure.
Developing tools: Paas offers many development tools for developers to build and
deploy their applications.
Software as a Service: Cloud service providers deliver software applications over
the internet on subscription basis eliminating the need of installation and
maintenance of the software.
For example, Microsoft's office suite application, Google's workspace.
Characteristics of Saas:
Accessibility: Accessible from any device from anywhere in the world.
Managed Services: Cloud service provider manage the underlying infrastructure
including software updates, security and maintenance.
Subscription based: Typically offered on subscription model at predictable costs.
___________________________________________________________________________________
________
Cloud Deployment Models:
Deployment models refer to how cloud services are made accessible to end users.
Deployment models are commonly known as Types of cloud. Mainly, there are 3
categories of deployment models.
a) Public cloud,
b) Private cloud,
c) Hybrid cloud.
a) Public cloud: The cloud Infrastructure is owned and operated by the large
companies like AWS, Google etc., but is made available to all general users or
public to access them over the internet. It offers massive scalability and a pay-
as-you-go model.
For example, The infrastructure and the services provided by AWS as well as Google
are used by various organizations and also the general users. And so, AWS and
Google cloud's are public clouds.
Since public cloud is available to general users as well, there will be some
security concerns but, the cloud service provides can provide the additional
security measures with additional costs.
b) Private cloud: The cloud Infrastructure is owned and operated entirely for an
organization and it is managed by the organization on their own and accessible only
to the members of organization. It offers high performance and high security level,
not allowing to access it by the public users.
c) Hybrid cloud: It is a combination of both public and private clouds allowing
data to shared among the applications hosted in the two types of cloud.
This type of cloud offers a flexibility of sharing data and applications and also
provides high security by storing sensitive data in private cloud and formal data
in public cloud. And also allows scalability with the help of public clouds.
We do hear two more types of cloud namely community cloud and multi cloud.
community cloud refers to the cloud which shares infrastructure to a community of
users or organizations with common goals or needs.
Multi cloud refers to the strategy, where an organization can get cloud services
from multiple providers at a time or later. But, for every cloud service purchased
by the organization will have a lock-in period (period of time, the services can be
used) and also had a choice to select the best provider among the multiple cloud
service providers.
___________________________________________________________________________________
________
Cloud Security: Security is a major concern for a public user or an organization,
while using the cloud. So, every cloud must follow some security measures and
practices to protect data, applications and infrastructure.
Here are the top 3 practices of cloud computing.
Data Encryption,
Access Control,
Multi-Factor Authentication.
Data Encryption: Encryption refers to converting the data or information into non-
readable code or information. This is a best practice for storing data in the cloud
and had various methods to do this encryption.
With the help of data encryption, the data gets unreadable for the
unauthorized users. For example, In Amazon EC2 instances, we use a key pair, which
is a combination of public and private keys. Here, the generated key is in
encrypted format.
2. Access Control: Cloud service providers must implement strict access
control, as per the users level, which ensures only authorized users can perform
any actions like accessing sensitive data and resources in the cloud.
3. Multi-factor Authentication (MFA): MFA adds an extra layer of security to
the cloud, which allows users to gain access to the cloud services only if they
verify through the predefined passwords and any other security factors like
biometric authentication or by using security tokens etc.,
___________________________________________________________________________________
________
AWS Cloud: AWS is a leading cloud service provider by the Amazon that deals with
almost all aspects of cloud services.
AWS offers wide variety of cloud based services like computing power, storage,
networking, databases, machine learning, Analytics, mobile and web applications,
security and more.
Benefits of AWS:
Security: AWS has robust features for data protection including a comprehensive
suite of tools for protecting infrastructure, data, and accounts, such as Identity
and Access Management (IAM), network security (VPC security groups, AWS Shield for
DDoS), data protection (encryption for data at rest and in transit), and threat
detection and incident response services like Amazon GuardDuty and AWS Security
Hub.
Reliability: AWS services are highly available and also had services for disaster
recovery and backup. Also had notification services when the threshold value
reaches. So that we can regularly have a look at them to focus on the usage.
Scalability: We can scale up or scale down the resources whenever required and also
can preset the number of resources required.
Pay-as-you-go pricing: Amazon offers some of the services free for new users for a
certain time period and then, we pay for the services for what we use. And we have
different pricing models for different services.