Scribd
Upload
19 views4 pages

Cloud Computing

Uploaded by

sumedhagandhi123
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
19 views4 pages

Cloud Computing

Uploaded by

sumedhagandhi123
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Traditional Datacenters have servers, cables, cooling systems, network setup

etc placed in a secured area. Cloud computing have redefined the way in which
infrastructure can be redesigned. There are some drawbacks with traditional
infrastructure, like-
1. Applications are hosted on on-premise
2. Huge capital expenditure is there to set up the infrastructure.
3. time and efforts are required
4. Infrastructure needs to be updated whole time
5. Resources could be either underutilized or overutilized. If infrastructure is
surplus, resources could be under-utilized
These all factors affect the total cost of ownership which is the sum of purchase
price of an asset and operating costs for its lifetime. A simple example would be
the cost of owning a car. You can buy a car, but you will still need to pay
license fees and insurance premiums, and it must regularly be serviced.
Now, suppose I have an online retail business where in normal days, network
spike is not much and I have enough resources to handle the data generated in
normal days. But when comes sale day, this network spike reached to a point
which my available resources cannot handle so this and generated data is too
much that a lot of data loss could happen, though my resources are overutilized
at this time.
So, scalability here is not much because it will require more infrastructure but
cloud can provide it.
Business drivers of cloud:
1. Cloud computing is scalable and elastic computing model. Elastic
computing means, cloud has the ability to expand storage, memory etc. or
either decrease it on demand. Basically, cloud services are able to add and
remove resources on demand. Elasticity is important because you want to
ensure that your clients and employees have access to the right number of
resources as needed.

2. Capacity planning to plan system resource configuration as per demand.


You can do it via three modes
a. Lead mode to add resources in anticipation before actual demand
increases.
b. Lag mode to add resources when demand increases to full capacity.
c. Match mode to add in small increments to match demand
continuously
3. Cost reduction to reduce OpEx and CapEx → OpEx stands for
Operational Expenditure which is the day-to-day running cost of IT
systems. CapEx is Capital Expenditure which is upfront cost of acquiring
IT systems

Some concepts I have come across:

On-premise → localized IT setup which has no physical or digital


component sourced from anywhere outside the perimeter

Scaling → increasing resources to meet demand. It can be horizontal


scaling by increasing instances of the virtual machine, or vertical scaling by
increasing capacity of components.

SLA → Service Level Agreement → measures availability, reliability,


performance

Goals of cloud computing are:

a) Reduce cost of ownership of IT assets


b) Scalability- We need cloud scalability to meet customer demand.
c) Availability- We need cloud availability to ensure that customers can access
cloud services whenever they need to and from anywhere in the world.
d) Resilience- fault-tolerant
e) Reliability - We need cloud reliability to ensure that our products and
services work as expected.

Cloud deployment models are modes in which consumers can access cloud
services:

a. public cloud → completely owned by provider, takes care of maintenance


and scaling. In this,consumer has no control over resources but is most
cost-effective as only OpEx is charged, no CapEx

b. Private cloud → owned by consumer; most expensive as CapEx+ OpEx


is charged, maintenance and scaling is duty of consumer. It is the most
secure model.

c. Hybrid cloud → two or more models (public, private) deployed and


interconnected. Here, sensitive data and devices are connected to private
cloud and less sensitive ones to public. It makes the most sense security
wise + cost wise → most practical solution.
Cloud bursting can happen here in which an application is running in a private
cloud and it bursts into public cloud whenever the demand for computing
capacity rises.

Cloud Service Models:

They are models in which cloud services can be made available to consumers:

a) IaaS → package of “raw” IT resources like hardware, OS, network setup,


etc., with dynamic runtime scaling. This is completely unconfigured and open
for total customization and configuration by the consumer.
App, Data, Middleware, OS, runtime are managed by consumer while
virtualization, servers, storage and networking are managed by cloud.

b) PaaS → preconfigured IT ecosystem which is ready-to-use for consumer. It


has lower level of control but great for migrating on-premise to cloud; example
– Webserver.
Only App and Data are controlled by consumer and rest everything is controlled
by cloud.

c) SaaS → software, standalone or a suite, provisioned on-lease to consumer. Its


backend IT setup is hidden and has lowest level of control.
example – G-Suite
Everything here is controlled by cloud.

Cloud Service Providers:

AWS, Azure and GCP

Billing pattern for all three: pay per use, per second, per GB hours

AWS Infrastructure:

 26 geographic regions and each region has 2 or more availability zones.


Av zones is a cluster of data centers that are fault tolerant, scalable etc.
AWS has 84 AZs
 offering over 200 fully featured services
 Local Zones:
 Wavelength Zones:
Points of presence:
AWS

You might also like