[MUSIC] Welcome to Cloud Essentials Module. Cloud compute, storage, database, and networking. This is the essentials of the foundation of the infrastructure of the cloud. We're going to speed through this, because through the journey of this training you're going to learn and depth about each of these. But I want to spend some time on it here. Now as we get started in the essentials. So, cloud compute instances, cloud storage, cloud data base and cloud networking. Just a cursory view of that in this course. First, cloud computing is instances. Talk about that in depth. Cloud storage, object file, and block, what does that mean? Cloud database, relational, and non-relational, SQL, and NoSQL we'll talk about that. And then cloud networking speed and ethernet, routers, switches, and gateways as the essentials of the network structure. We'll talk about that. All right, now, cloud compute and the definition of cloud instance. A cloud instance is a computing event typically running on a virtual machine, a server resource in a public or private cloud, both. It is also provisioned compute hardware, with storage and networking. Part of that event. A couple examples, Google GCP. An instance is a virtual machine hosted on the google infrastructure. Talk more about that. And next is Amazon EC2. Elastic Cloud Computing (EC2). AWS provides a wide selection of instant types optimized to fit different use cases. Their words, instant types comprise a varying combination of the CPU memory storage and network capability and capacity. All those included. Now, let's take a look at an application from a data center to a virtual private cloud instance offering. Here we got to see is a VPC is the fundamental elements of the cloud industry. It is a logical set of infrastructure resources in the public cloud environment. VPC allows an enterprise application like application. A here is an example on the left, for the full benefits of taking advantage of the private cloud elements. All the security behind the firewall and that in a public cloud environment. And the VPC allows that to happen. So, if you have an application A on the left, you're thinking about moving it to the right, which is in the public cloud. And here's an example of a Google cloud platform, GCP example. Here's what you got to think about. Your administration staff for IT and your DevOps team get involved in this to start architecting it. And the architect in the middle there, the inputs, the outputs, the interconnects, what's needed for access, and what do you need to tie into. On the right, you start to create your subnets with instances and this is the notion of instances. Everything built on it's compute, it has memory storage and networking. Here's how you logically build it out. The picture in the middle on the right is actually Google's data center. And you get a virtual slice of that, you get a virtual part of that in the VPC setup as you put an application in place. So, that's a quick run through of an instance used as you're looking at the in application. Now, let's look at amazon's cloud offerings. General purpose, compute optimized, memory optimized, acceleration, and compute going together, and then storage optimized for your application. You note here that they're well categorized, there are 275 instances that are offered are quickly point and click part of your well architected domain or VPC. And you can see here a simple example of that in the bottom right. You have your gateway, you have your auto scaling and scaling zone, you have your instances to be used. It's simple example, but how instances are used are showcased in the bottom right on the bottom right of the chart, the bottom left of that chart. Now, let's pull out an 8100 series Xeon. Here's how it gets offered up. You have virtual CPUs of choice, you have your size of memory for the choice and the platform. You have your storage local attached and you have your bandwidth and then transfer speed to the storage. That all comes together with choice. And you can see there's different choices across the 8100 series Xeon here for you if you go on to the amazon catalog. Also, quickly moving forward here. Here's an example of packet, typically offers up bare metal systems. And they're a company that actually provides a number of intel bare metal offerings in the platforms. You can see a E1 small X86 on the left, which is an atom based processor. And you see a G2 large X86 on the right, which is a Xeon 6126 processor. And you can see the platform down the list there. Let's spend a minute on the right. So, you have 28 cores in that platform. A 2-processor system, you have 192 gig, you have almost a terabyte of storage and SSD, you have a couple of video cards for processing, and you have 210 gig ports cards for network connection. A pretty robust system and you're paying $5 an hour. So, let's talk about costs for a minute. You're paying $5 an hour in that, which is typical for a large system. On the left, a smaller system, you're paying 7 cents an hour for that. Quite a difference. And so, will you architect and well architect your application domain in the cloud. You have choice, and you have the ability to choose the lower side performance and capability and the higher side performance and capability. Let's talk about storage. There's three types of storage, there's object storage, file storage, and block storage. And you can see typical objects are things that you would typically have in your everyday use, whether that's a sound file or a video file or a picture or whatnot, objects really. And objects are best stored in what's called buckets. And you can see here, AWS S3 is an offering for this. It's a bucket object type storage offering. You can see in the file domain typically stored in hierarchical folders. The AWS EFS; elastic file storage system is an example of a product there. And then blocks, this is typically enterprise databases and we'll talk about that in a minute. And you can see AWS elastic block storage offering for the storage. Here is memory storage hierarchy that you typically see, particularly in and around an intel view. You have memory on top, which typically is DRAM, DDR3, DDR4. Typically hundreds of gigabytes of storage, memory, and nanosecond speeds, very fast based on the CPU and the platform and the application. Now, the storage on the bottom you have a couple of things that are important here. One is SSDs have continued to ramp in popularity. You see tens of terabytes local to the platform at microsecond speeds, fast and capable but memories faster still. Now, on the bottom you see archival type or longer lead time to get information type storage, typically a hard drive or a tape in the millisecond speeds. And this is terabytes and terabytes of information. Now, back up to the top, is there's an opportunity that intel is crafted in the industry is crafting on larger memory close to the CPU. And that's obtains persistent memory as our brand and product line going into that. It's been well placed. It's for application that needs a larger memory footprint but speed near DRAM speeds that is showcased here. And then, if you want a fast boot, you want to think about and obtain SSD. Which has performance and capability near the platform in the platform for that speed of use or in memory database as well. A couple of places that intel shines. Now, here's a quickness glance at amazon S3 storage classes. You see on the bottom S3 standard that typically is a scent or a few cents of gigabyte per month. And then if you move to Glacier or glacier deep archive, it's a fraction of a cent per gigabyte per month. Just an example of how the storage gets presented in the cloud environment, it's typically size cost per month. And one thing to note here, and this is interesting, typically in the hyper-scalers in the cloud environment, if you transfer data in its typically free. If you get the data out, it's typically paid based on the amount of data that you get out. Just something to keep in mind. Now, databases. Quickly on this two types relational and non-relational. And let's go down the list, relational is typically structured data, it's user data or inventory data. Typically an Excel file as an example of that, you can think about the data in the different cells within Excel. That's typically well structured and its relation relational because it can tie into cells next to it and other Excel files aside to it. Again, typically tables and data records. And the query tool here is SQL. We'll talk about that. A couple examples of the product from MySQL and open source database and query language and Microsoft SQL server, a product that's licensed. Now, non-relational. Typically unstructured information, photos, text, video. Typically you think of in a word file, the tax within the word file unstructured. The schema is typically free formed objects and texts as we just mentioned. And then, it's good for really searching and pattern matching. The database itself is can put a lot of pieces together and kind of a point to point fashion and and bring it out. And the NoSQL is the query language for that database. So, you can think of MongoDB and Oracle, NoSQL database as examples of this. There's many others, there's dozens and dozens of databases in the industry today. Here's a couple examples of that. Now, SQL and NoSQL, what does that mean? You can see that some of the structure and the SQL on the left relational database structure. It's structured query language, SQL is structured query language, its ability to communicate to a database and back out of the database. NoSQL, again unstructured data. NoSQL really means non-SQL or non-relational or non-structured database query. So, the query provides a mechanism for storage and retrieval of what is modeled as free formed fashion of data. And examples are key value oriented, or data document oriented, or column oriented, or graph oriented type databases. That's a quick run through. Cloud networking essentials. This is important, we're just going to spend one page, one minute on this. On the left is an architected virtual private cloud. You can see the three domains in the middle, there's three elements to the VPC in the middle. You have a development environment, your test environment, your production environment. And then you have your networking lines tying it together. Here is a gateway, a transit gateway in the middle networking coming in and out and provisioned. And then it ties out to an on premise customer gateway. So, gateway to gateway is talking. So, what does that all mean? How do you work around the definitions of networking when you're thinking about this? Okay, first thing is speeds. You have typically 10 to 400 gigabits per second speeds today in equipment. Typically it's 10 to 25 gigabits per second in a modern system to system, rack to rack infrastructure. In a hyper scale modern data center, it's typically 25 gigabits server to server, server rack to server rack, pod, pod. Then as you go up, as you architect the network, you have faster speeds tying it all together. Because you can imagine all the aggregation of the individual server speeds go up and get more demanding and volume. A moment on ethernet. Ethernet is a standardized networking language protocol, allows communication between devices. You can think of this server to server. And it's also known as the IEEE 802.3 standard that was established in the early 90s. Still very robust today, it's the protocol that brings it all together. Now, last networking equipment. So, the networking interface, typically the NIC, the network interface card can be on a motherboard is the input and output to an individual server. A network switch allows IP packets to be routed between the servers in the rack or across rack, multiple racks, pods. So, it's a switch is more local to the devices. Routers ties local networks together from the switches. And then, gateways, as we saw an example on the left, tie into and out of your data center to the wider area networks or call it way and the wide area networks. Disparate networks tying it together through the gateways. Those are just a quick flip, there's other things like hubs and different elements and devices in the network. But those are the four things you need to know. NIC, switch, router, and Gateway. And the last thing I'm going to say on this, we're going to tie it all together is VPN. Intel uses VPN, most everybody uses VPN. It's private tunnel to encrypted tunnel across the network. And it allows some condition of anomie, some condition of security within the network tie point to point from my device to the intel IT networks. It allows that virtual protection of data as it transits the internet. Now, a summary. Hey, there is a lot of things that were talked about, we just skimmed it. But these cloud offerings which are compute instances, storage options, database choices, networking provisioning. It's a robust set of solutions that are at your disposal. These technologies are available on premise as well as the hyper scale data centers, and is available today. The cloud industry and the supply line continues to rapidly innovate on these offerings. Higher speed and networking, better database choices and performance and more compatibility. Better storage choices and of course better compute choices that we bring to market and many others. And all of this again, I'm going to say this again, I said it in one on one. The rapid growth of the cloud industry requires knowledge and skills to influence and maximize impact for our customers and for intel corporation. We're on a journey here. We truly are on a journey, this isn't one of another classes that you're going to take through the fundamental series here. Understanding this, understanding of cloud offerings, make you a better world-class influencer. Thank you for your time. [MUSIC]
