Ubiquitous, on-demand network access to shared pools of configurable computing resources ideally requires minimal management effort or service provider interaction. This course covers the essential characteristics of data processing in the cloud, service and deployment models, and key components of implementing Amazon Web Services, as well as constructing Hadoop clusters and performing MapReduce operations.
Introduction to Cloud Computing
Loading...
Del curso dictado por Arizona State University
Data Management in the Cloud
16 calificaciones
Explore Related Videos
At Coursera, you will find the best lectures in the world. Here are some of our personalized recommendations for you
Conoce a los instructores
Mohamed SarwatAssistant Professor
Computer Science and Engineering
Ming ZhaoAssociate Professor
ASU School of Computing, Informatics, and Decision Systems Engineering.
Before we learn how to do scalable data processing in the cloud,
let's first try to understand what cloud computing really is.
So, the objective of this unit is to understand the basic concept of cloud computing.
So, what exactly is cloud?
If you asked this from different person, you may get different answers.
Do we have a official definition of cloud computing?
And the definition here is probably as official as you can get.
This definition is provided by the National Institute of Standards and Technology.
And this definition is actually quite long.
So, let's just try to read it and I'll help you understand it,
I'll help you dissected later.
So, what this definition says is,
cloud computing is a model for enabling ubiquitous, convenient,
on-demand network access to a shared pool of
configurable computing resources that can be rapidly
provisioned and released with minimal management effort or service provider interaction.
So, from this really long definition,
we can already see a few things.
First of all, cloud computing is a computing model and
this model will allow the users of cloud computing
to have access to computing resources.
And these resources, they are from a shared pool of
resources and this kind of access is down via network,
via ubiquitous, convenient, on-demand network access to this shared pool of resources.
And moreover, these resources can be rapidly provisioned and released
with minimal management effort and the service provider interaction.
So, at this stage just try to understand that cloud computing is a computing model
that will allow users to access a shared pool of resources across the network,
pretty much from everywhere and the resource can be rapidly provisioned
and released without involving a lot of human effort.
So, now let's just try to dissect this definition.
To understand this definition,
we can start with the key essential characteristics of cloud computing.
And the first one is On-Demand Self-Service.
So, what On-Demand Self-Service means is that a consumer
can unilaterally provision computing capabilities.
And these capabilities include server-time,
network storage and any other kind of resource that
a consumer needs to run his or her jobs in the cloud.
And these capabilities can be provisioned as needed
automatically without having to require human interaction with the service provider.
So, that's what the unilaterally really means here is that the user
can request resources and release resources,
by himself or herself without having to involve the service provider really.
And the capabilities again can
include any kind of resources that the user may need to run his or her job.
So, that's basically what On-Demand Self-Service mean.
That's the first essential characteristic of cloud computing.
Now, the next one is Broad network access.
And this is a fundamental characteristic,
one of the neighboring technology of cloud computing that the resources,
the capabilities from the cloud computing are
provided through the network and this network is ubiquitous.
So, as long as you have network access,
then you should be able to use it,
to be able to access the capabilities that are available in the cloud.
And in order to enable the ubiquitous access to their B2B some kind of
standard mechanism so that no matter what kind of environment the users have,
the user can use the same mechanisms to access the capabilities available in the cloud.
And later on, when we study some concrete examples of cloud computing,
you can see that one of these standard mechanisms is web services.
I just give this as one example.
And these standard mechanisms will allow users to access
cloud resources from whatever kind of environment that the user has.
And this environment, the user environment can be either a thick client platform.
That's the kind of laptops,
workstations that we are familiar with or it can
be a thin platform such as the smartphones and
tablets which are becoming more and more widely used than personal computers.
So, the phones and tablets they're called thin clients
because we have a much thinner software stack in those kind of
devices and the traditional laptops and
workstations they're considered as thick platforms
because their software stacks are much thicker.
So, to summarize, broad network access means that
the users will be able to use some standard mechanisms
to access the cloud resources across the network
no matter what kind of thin client or thick client they are using.
This is second essential characteristic of cloud computing.
Now, the next one is Resource Pooling.
This is a very important one.
This is in mainly from the user's provider perspective.
It says that the providers computing resources are pulled together to
service the consumer, the cloud users.
And another way to- a more formal term for this resource pooling,
we use resource consolidation.
This is the same thing is that we consolidate resources together and we form
a resource pool and use this pool to service the resource request from the cloud users.
And on one hand, we cross provide these consolidate resources, put them together.
On the other hand, this pool of resources will be used to
service multiple consumers and this is
the second aspect of this characteristic is that
the resources are provided to the users following a multi-tenant model.
A multi-tenant model just like when landlord rent out an apartment to multiple tenants.
And a cloud provider will rent out his resources,
the pool of resources to multiple users,multiple consumers.
And in these resources,
they can be rented in the form of either physical resources,
meaning physical CPUs, memories,
and storage capacity or can be in the form of virtual resources.
So, what exactly does virtual resources mean?
We'll talk about this a little bit later.
And these resources no matter physical or virtual,
they are dynamically assigned and are released to the users according to their demand.
This come back to the On-Demand characteristic again.
So, these resources they are dynamically assigned and reassigned.
So, resources are not- typically,
they are not dedicated to any particular user and in any particular quantity.
They are dynamically assigned and reassigned based
on what the consumer really needs at that moment.
Now, let's express that there's a sense of location independence.
Meaning that, from the consumer's point of view, the consumer,
mostly the consumer doesn't know where the resources
that he or she is getting is really located.
So, that's what the location transparency or location independence mean.
We can also call this location transparency.
So, when a user runs some resources from the Cloud and runs his or her application,
the user generally does not have
a location where his or her applications is really running.
But there are some exceptions that the user might have some knowledge,
a higher level of obstruction of location.
The user does not know,
never know which is their server,
which is the exact like the server is located to run the user's application,
but the user may have some knowledge or even some kind
of control in terms of, in which county,
which state, or maybe even which data center his or her applications should be run.
But that's about it,
that's the the obstruction location
that the user might have knowledge of I mean the user have control of it,
but not more than that.
So in general, the user have low control or knowledge of
the exact location of the resources that he or she is getting.
Then again, these resources can be anything, okay.
Just imagine what kind of resources you can get from a cloud provider.
Just imaging what kind of resource you might need and there
are also some resources that you can run from a cloud provider,
and that includes storage, processing, memory,
and network bandwidth and maybe many others that you might need for your applications.
Next up, the next essential characteristic is rapid elasticity.
So, elasticity means that resources can be provisioned and they are released,
the resource can scale out and it can scale in as needed.
And not only the resource can be a provisioned on demand,
but also resources is kind of on demand is done elastically.
Elasticity means that the resources are provisioned and released really quickly, rapidly.
So, that's what rapid elasticity really mean here.
So to reiterate, the resources can be
scaled rapidly upward and inward according to the demand.
So that's something the cloud provider needs to be able to support,
is to be able to supply resources and
recycle resources according to the user's demand
rapidly and from the consumer's point of view and not only that,
it seems that the consumer will be able to get
infinite amount of resources from the Cloud.
Of course, nothing is really unlimited.
If you look at the largest cloud provider like AWS,
Amazon Web Services, even they are resources that are really unlimited.
But we say it's unlimited,
we are saying that's from the perspective of the user.
So from the perspective of the user's application, what the user's job,
the user will always be able to satisfy his or her demand.
Just from the user's perspective,
the resources available at a cloud provider is unlimited,
and it can be can be provided in any quantity at anytime.
Now, finally the last essential characteristic that describes,
defines cloud computing is as major service.
So cloud computing is not only a computing model,
it is also a business model.
It is just like, again,
just like when we rent things instead of buying things.
When users use cloud computing they rent resources
instead of buying resources and own resources.
So the cloud systems,
so there are little bit of way to measure how
the services are provided in how much quantity.
So the cloud systems should be able to,
so there has to be a measuring capability at some level of
obstruction to measure how much of resources are provided to a user,
how much resources are consumed by a customer.
And the units of this measurement depend on the resources that are provisioned,
it also depends on what is the cloud provider's capability.
So, for example, in terms of storage,
in terms of a capacity in units of bytes,
in processing we can measure it in terms of few cycles or number of CPUs,
it depends, in bandwidth,
in terms of bytes per second,
and number of active user accounts,
these are just some simple examples.
Only when there is a way to manage the resources,
they're provisioned to each individual user,
then the cloud provider is possible to optimize the control
and optimize the resource or provision accordingly.
And the fact, from the consumer's perspective,
it also makes the resource provision sort of transparent to the consumer.
So to summarize, these are
the five key essential characteristics that define cloud computing,
if we can go back to last one is managed services,
second last one is rapid elasticity,
and before that resource pooling,
before that is product access,
and first one is on-demand self-service.
And then after we understand
these five essential characteristics then we can go back to this.
We need long definition, you see,
actually the definition covers all these five essential characteristics, right?
This resource pooling here,
shared pool resource, there is product access here,
and there is, what is next one,
rapid elasticity is here.
The major service is not explicitly matching the definition,
but is implied and we are missing one,
what is that, resource pooling,
I think, we covered all of them.
It is on demand self-service,
yeah on demand is mentioned here with a minimum/maximum effort.
Okay, that describes the self-service part of on-demand services.
So if you really understand these five essential characteristics
then you understand what is cloud computing.
And then when you see
any particular cloud computing example and believe me there are so many,
so many cloud computing services today,
that you'll be able to understand why they are
really cloud computing and why they are considered as cloud computing.
Explora nuestro catálogo
Inscríbete de manera gratuita y obtén recomendaciones personalizadas, actualizaciones y ofertas.
Coursera brinda acceso universal a la mejor educación del mundo, al asociarse con las mejores universidades y organizaciones, para ofrecer cursos en línea.
© 2019 Coursera Inc. Todos los derechos reservados.
