7.1 AWS Cloud Modeling

00:09

Hello again, you cloud Model builders and welcome to module 7.18 Ws Cloud modeling. Remember building when I was a kid? These great model airplanes in Carson model kits I would get from the hobby store. It's been weeks building them, and I have to put on the decals and paint them. And most of the models ended up finishing were complete disasters, though, because as a kid

00:28

I was always in such a hurry to get to the end result. I would just use gobs of glue to connect the parts together.

00:33

Thes things looked like the parts were stuck together by chunks of chewing gum and you so much glue. And I was just horrible that the details So we won't even talk about my painting and D count skills. Suffice it to say that we had a lot of models of cars, planes and trains to the dumps landfill.

00:49

The good news for us is we're gonna build some great cloud instances of Amazon's elastic compute cloudy. See two instances,

00:55

types and features. We coul cloud modeling, and we are gonna fill up our trash heap with these cloud models, that's for sure.

01:03

So if there's one slide from this course that I would want you to take a screenshot of, it would be this one. There are way too many AWS products and service offerings to cover in this course, which is a foundational course covering everything cloud in not just a W. S. But for this discussion. We're gonna learn more about this little corner of the AWS portfolio. But in actuality,

01:22

it's this giant, epic, huge, transformative and disruptive service that has changed the way the enterprise thinks about delivering a gym. Users.

01:29

It's critical computing hardware in application infrastructure, this thing called easy to or elastic compute cloud.

01:36

So Amazon elastic, compute, cloud or easy to provides almost infinite compute capabilities in the cloud, and it's both very reliable and very secure. You can run any kind of workload in what's called an E C two instance without having to invest a lot of capital expenditures to get your organization's computing resource is

01:52

the model for cloud computing is pays you go P A. Y g, which means you pay just for the resource is you're going to use

01:57

on an hourly basis. You don't have to go through this huge procurement process and then deployment process to stand up. A new computing resource is like you would on premise in your own data center, instead of a few clicks from your, um, Amazon management console and your new instances up and running

02:14

the page ago. Model works great for customers because if your organization has a spike in the trafficker workload, you can quickly spinoff additional servers

02:22

almost instantly. And when the traffic reduces, you can get rid of those servers. You need a bunch of new workload for the weekend, and then you want to turn it off come Monday to business as usual. Well, with the elastic compute cloudy, see two and pays you go. No problem.

02:35

The first part of modeling our cloud computing E c T resources to determine what kind of easy to instance our organization requires.

02:43

Do we just need a general computing resource for Windows Server? Or do we have a CPU intensive application like a city government geographic information system? G. I s where we're gonna need to spin up.

02:53

Service is that will help manage traffic and population challenges using special on three D visualizations, or do we need a bunch of storage memory or both for a big environment resource planning AARP database.

03:05

Well, depending on what we need, E C two has compute instance options that meet our needs and our company's pocketbook

03:13

When you spin off servers in Amazon, easy to you have complete control over the type of storage you want to use. The network configurations, the security configurations and so on. The easy to Web interface allows you to configure a server and minutes compared to what could be days in the traditional on premise physical server model with easy to

03:30

the time required to obtain and boot the new servers a matter of minutes,

03:35

since now it only takes minutes to deploy a server. You could play hundreds or even thousands of servers almost instantly, and you can quickly scale up and scale down. Depending on workloads and traffic volumes. You have the choice of multiple instance types, operating systems and software packages. US and easy to administrator can select and less control.

03:53

You're configured memory CPU resource allocation, easy to instant storage,

03:58

including the boot partition size that is optimal for your choice of operating system and application.

04:03

The benefits of easy to are many, including the biggest advantage, which is time to market. You can deploy any server almost instantly, and as a result, you don't have to wait for weeks or months to get a new server. This also will help foster innovation because you've been quickly get Resource is for that new project, that new lab. That new pilot testing environment or prototype environment

04:21

scalability scale up and scaled down a pure elastic business. Imperatives based on the peaks and valleys of your workplace. In enterprise,

04:29

you don't have to worry about the high cost of over provisioning just to support a few peaks in the volume of your workload. Traffic E C two is reliable and secure across a huge variety of instance types, various operating systems and software packages that you get to choose from

04:45

the E. C to instant types are broadly divided into the following categories. General purpose, computer optimized memory optimized storage optimized and advanced computing. The tea instances or to use the military alphabet tango instances and the general purpose category like t one, Andy too

05:00

are the smaller general purpose instances. Compared to the general purpose in one through M fives,

05:05

general purpose instances provide a balance of computer memory and network resource is and are good choices for many of the applications. You will spin up in turn on some of the general purpose instances, like the tea to provide burst of performance, which means these instances provide a baseline of CPU performance but have the ability to burst above the baseline.

05:25

This happens because a T two instance will accrue CPU credits when performing below the baseline.

05:30

That's providing you burst herbal performance above the baseline. Depending on how many CPU credits have accrued over time, they m through through M five. General purpose instances are great overall choices for things like building websites, development environments and testing and staging environments, but don't provide the burst of all performance like the tea, too.

05:49

The compute optimize See three, C four and see five's or for work loads that are heavy and compute and are the optimized instances ideal for compute intensive requirements. High performance processes, air used, delivering a lot of processing power for compute intensive workloads like media trans coding, large concurrent users supporting applications, high performance computing,

06:10

gaming servers and so on.

06:11

Memory optimized x one x one e R three. And our four instances are ideal for workloads that you're planning to run to have a lot of memory requirements. Any application that processes large data sets and memory will benefit from these memory optimized instances to do things like running Oracle and S A P databases,

06:29

big data process engines like Presto, an Apache spark High Performance Computing, H P C. And so on.

06:34

Storage optimized instances can be used for workloads that require high sequential, read and write access to very large data sets on local storage, delivering thousands of low latent see input output. IO operations per second or I Ops and Advanced Computing P to P three, G three and F one

06:50

are for high processing computing requirements such as running artificial intelligence.

06:56

Aye aye machine learning algorithms, molecular modeling, computational dynamics and finance and so on. These advanced computing instances provide access to the hardware based accelerators, such as graphic processing units, GP use enabling high performance, parallel computing and high throughput.

07:15

So after successful modeling, there's just a few short steps and you're going to be off to the races. There's only a few steps to provision or easy to instance, and Then you can start your new models engine. First, you're gonna select your Amazon machine image. Then you're gonna configure network. You're going to choose your instance. Type

07:30

your availability zone, which will control the late and see to your new platform your elastic block storage, which is a fancy way of saying you need how much disk space or solid state flash drive space for your new platform.

07:42

And then you're gonna take your foot off your models break, put your foot on the gas, start your engine and your new instance you've just modeled is off to the races.

07:49

So I'm in my AWS management consul home page. I'm gonna go up here. The top hit service is,

07:56

and you'll see the very top option here under compute is my easy to

08:01

now. This front end allows me to control anything in everything from my easy to platform service's from setting up a new machine, setting up a new instance to managing Compute deprecating a new instance to turning it down. Whatever I want to do for that instance, I can do it from this page, but I'm gonna go ahead and click a launch. In my instance

08:20

once I have successfully modeled. I know how much optimized computer and memory storage I require for my Amazon machine image. I can come down here and start selecting the platform that I want to turn up and my optimized model.

08:35

Well, I've got all these different various Windows operating systems, Lennox operating systems all the different. Various Sues Lennox and Red Hat Lennix all the different Lennox options available. I'm gonna go ahead and turn on the souse Lennox machine

08:50

and here's all my instance types. You remember we got our micro's or small and mediums or burst a ble. Aah! Possible T ones and T twos and t three's. We've got our M three and M five images here and across our

09:05

spreadsheet here with Guide. How many CP use how much memory, whether or not a blast elastic block store is the only service that we can turn on from a storage. Or maybe we can even turn on SS D's and how many SS D's and help build out a raid array because we need high availability on fault tolerance for our storage

09:24

one. We're ready, for instance, and we choose it will go ahead and

09:28

it next in. This is where l configure my virtual private cloud configurations of might know where configurations my identity management. If my service is a T two or T three and it's burst of all eligible,

09:41

I'll go ahead and click enables. And now, if I've got credits because I've been running below my baseline for awhile now I can perform above my baseline

09:48

and do it at my ST V because I can use that burst of service against my credits that I've developed over time, I come here to add storage. Well, maybe it starts up with general purpose, solid state drives, but I need more ops I ops than that. So I'm gonna click this button and come in here and say, Well, I need Ah 1000 I ops. So I need

10:09

a little bit higher performing storage and I need

10:11

32 gig

10:13

32 gigabytes of memory. So go ahead and select that and we'll hit review in lunch. And here we can see our micro image. How many see pews are instance.

10:26

What's our storage? Well, you can see we've got 1000 I ops and 32 gigabytes of storage is what it's gonna turn up with. And then when we hit, when once we hit launch,

10:35

I'm gonna download my one time password, which is my crypto keys. So I've got a one time password to access my virtual machine. Meanwhile, in the background, my virtual machine is standing up in the cloud, so it's spending its provisioning itself. It's doing all of this stuff. Yeah,

10:50

auto magic that happens in the cloud through software defined capabilities of the

10:56

drives, the cloud, and it could take 2 to 3 minutes firm and machine to turn up. It could take a long as an hour when I'm ready. When it's ready, I'm notified, and then I use my one time password to log into it and start spinning up. My service is

11:09

pretty nifty. Really cool stuff.

11:13

So that's it for module 7.1, and yea, the teacher says just this once the students get to skip the learning check and skipped the exam. Why? Because the elastic compute cloud the EEC to module this module 7.1 is so big and so important that

11:28

it's time for you to start doing some homework. So I'm gonna give you some homework on this module and I expected to go into your cyber supplementals and start taking advantage of some of those labs

11:37

are amazing Cyber A teaching assistants have taken the time to build out for you. So get some of that hands on learning. And as part of that hands on learning, I expect you to go into a W s and sign up for a 12 month free tier account. And with that, you're going to start doing some of this on your own. And as part of this part of the homework, I want you to

11:56

breakdown all of the differences

11:58

on the different compute optimization memory opposite optimization storage optimization, advanced computing optimization instances. So when we're talking about cloud modeling, what's the difference? A T one anti two and the C one to a C three in an M one to an M five. So, no, you're instance types. And when we want them,

12:18

what do we mean by modeling? And how does cloud modeling help us? Well, we built our first model today, and now that you're going to go into the labs and do some of this yourself, well, you won't be putting your cloud models in the trash heap like I did with all of my model planes and model trains.

12:31

Now, the five. No. The five steps necessary to provision in Turn up a new easy to instance. We talked a little bit about it We showed you.

12:39

And I want you to do some of this on your own.

12:41

So that's it. So next time we're gonna talk about a W s storage elastic block store cloud backup in cloud retention.

12:48

So thank you all, you cloud champions. On behalf of all of us at the Cyber Security and I t learning team, we want to say thank you so much for joining us. We appreciate it. I do want to let you know that we had a little internal sit down and little internal meeting and all the teaching assistants have been told and instructed that if any of you out there say your dog ate your homework

13:07

and not gonna cut it, you still got to get your homework done.