1 Hour Introduction to Azure - For Absolute Beginners

hi there this video was taken from my course aids at 304 Azure architects if you are interested in becoming an azure architect Microsoft does have two exams on it the ACE at 308 at 301 this video is from the 300 course there's a link to that course in the description of the video if you want to be notified anytime there's a new video that goes live on this channel go ahead and hit that like button and hit the subscribe button and you'll be notified anytime there's a new video for you to watch anyways so I wanted to start off this course with a bit of a baseline set the problem with teaching through video or teaching even live online to a group is that I have no way of really knowing your individual level of experience and adapting the content to you specifically and so whether you're at the beginning the left-most of this chart where you're a beginner you don't really have much experience with Azure or even with cloud computing or you're a little bit farther along the way maybe you have experienced the AWS but not with a sure you understand the cloud concepts and what the benefits are but you don't really have the specific skills working with PowerShell working with the portal etc or maybe you're far you know midway down this line here and you maybe you've even passed one of the existing associate level exams be at the administrator or the developer you work in some specific areas of azure but you don't have exposure to the entire broad set of services that Azure has or maybe even you're above that and you've been working with Asha for a couple of years and you've got a broad range of skills you've just not taken the architect test yet and you want to take a course like this to catch yourself up on all of the areas of azure that you're not exposed to on your day to date well I don't really have that way of adapting this course to you specifically and that's just the reality of it and so as we're going through this course I want to start off sort of here I sort of picked an area on this line about 25% down where I want everyone to get to at least this level of knowledge and so this video is sort of an attempt to get you if you watch this video and follow it you'll at least get to where that spot is on the line now if you're very far to the right if you're already an intermediate you've already passed an associate level exam or you've been working with Asher for years maybe this stuff is not great for you and you want to sort of put this on a faster speed I know with video technology we can set videos to a faster speed and so you might be what go through this at 1.5 X or faster or maybe even skip it entirely that's totally up to you but basically I want to get everybody from that leftmost spot on the beginner at least into a novice level in a single video or set of videos here and so it's a bit of a challenge but we're going to we're gonna push everyone at least a little bit more to the right now I will say that this does not necessarily replace the azure fundamentals level exam okay you're not gonna watch the single video and then go off and get certified as a fundamentals level on Azure it doesn't cover all of the topics on that exam but this is basically I don't know if it's equivalent or this is basically where we're aiming at we're gonna get people into this general range here so we're gonna start off talking about cloud computing in general not necessarily specifically Microsoft Azure but what's the big deal why you know we maybe you've worked in IT for a little bit of a time you've worked with servers client-server setups or hosting what's the big deal with cloud computing what's although the fuss about so this is in my own words but basically the big deal of cloud computing is that it gives you this ability to rent computing resources and not just you know you can rent a resource at Bluehost or any of the hosting providers but it's available for your use within minutes and you only pay for what you use whether it's in time or in gigabytes or other measurement and that costs can literally be pennies you could literally get a computer use it for a little bit and end up paying 30 cents and that would not came out about 10 years ago that was a revolutionary idea okay so before cloud computing was available or if you're using services other than cloud computing services either you own the equipment and so you go off and you get the server and it's a massive upfront cost and time it takes weeks to install the software and then get that into your own data center or you have to do a to release type of thing which may be less upfront cost but it's still you end up signing a lease for three years or four years and and you still need to take take a week to lease some equipment or if you go you know let's talk about Bluehost and those shared hosting models that's not necessarily reliable performance you lose something in that shared environment where you lose control other customers have the ability to negatively impact you and and things like that and it's not necessarily you still have the commitment issues you still have the amount of time it takes and the loss of control so I can remember a story not too long ago maybe five or six years ago working in an environment where we knew we needed to grow the environment we had a project coming up we needed a server a dev server and testing server production server and so you go IBM is the host and you go and you say well we need a server they go off and take some 48 hours to come back with a quote they go off and find the latest hardware prices they put their own estimates around it how long it's gonna take them to get the machine how long it's gonna take them to set it up how long it's gonna take them to get it available within the environment and it comes back with a fifty thousand dollar price tag for a single server and I have to go to my boss and say we need $50,000 for this project he doesn't have the authorization to spend fifty thousand dollars just like that so he has to go to the finance team and say we need fifty thousand fifty the finest team goes well you got to talk to the client blah blah blah it takes weeks to make this decision finally we get the approval blah blah and so that's the environment many corporations and enterprises are coming from ok so to go into the cloud environment where you can literally make this happen this idea and then within 20 minutes have a server at your disposal try something out and then say oh that actually we'll do something differently and then stopping that it's it's revolutionary cloud computing if you go in there it's not just us Windows servers Linux servers there are hundreds of unique services machine learning and databases and storage and we'll get into that in a few seconds here so back then you know 10 12 years ago this infrastructure as a service this is what they call when you go and get a server in the cloud or networking in the cloud that's infrastructure provided as a service that was sort of like the idea that that Amazon had that was sort of revolutionary now let's talk about the different companies that are available in this space now hopefully you've heard of these people before Amazon Web Services AWS is was the first mover I think they're about twelve years old at the time of recording and as of today there's still the largest cloud provider many people think they have a huge lead in the space and I'm going to show you in a second that their huge lead isn't actually that huge Microsoft Azure started about ten years ago and is the current number two provider of public cloud services the gap has really shrunk Microsoft was growing outgrowing Amazon your for years and so basically they've caught up in a lot of ways Google which is you know you're surprised because Google was sort of into search and YouTube and ads and they've had this they were famous for having millions of servers long before you know Microsoft decided to get into cloud and there are actually legging behind when it comes to providing public cloud services not only legging behind in in revenue in that particular part but in the their capabilities and then their their offerings and the their vision etc I always turned to this company called RightScale who produces what's called the state of the cloud report every year and it's always interesting when they come out with this on the last state of the cloud report for 2019 actually showed Amazon AWS losing 1% of the what was called adoption in the enterprise so if you look only at these fortune 500 or these big enterprise size companies they've actually lost a point of market share in their Asscher has continued to grow and they grew 2% over 2018 Google again 1% that's I mean to fall up at 1% or increase one percent that's pretty flat really we've got the other players there's many different cloud providers some of them not even listed but we're talking about teens or talking about Alibaba at 4% etc and so in the enterprise will see that these 1 2 and 3 Amazon AWS Microsoft Azure and Google cloud platform are really the three top cloud providers and the other ones even if their adoption rate or their market share isn't so far they're really far behind in capability according to many many things Gartner group now you'll see also from those percentages that you know Amazon is at 67% in Asia at 60% well the whole thing adds up to over 200 percent so multi-cloud is pretty common you'll have companies that use Amazon for this particular project and use Azure for something else they'll put their stuff in an s3 bucket and they'll run their virtual machines in Azure you know different tools for different purposes there are you know it's also a protection strategy in case the prices on AWS get out of hand you can have them compete against each other etc many different reasons for using multiple cloud options now I'm trying to express my excitement about the cloud I personally think that this has been an awesome revolution I've been so thrilled and excited to be part of it in the last few years in terms of teaching a sure but I also want to sort of put a stop and say it let's just you know really the cloud is is great and I love it but it's really just a computer and you knows it's there is no cloud that's just someone else's computers so I'm going to continue to be excited about it I'm continue to try to get you to get excited about Asscher particularly but you know it's just we're talking about computing here so that's sort of an overview of sort of the really high-level overview of the of the players and of what the cloud is let's get a little bit deeper you know bring the plane down a little bit lower and talk about the overall benefits of cloud so I started off this video talking about renting computing resources available in minutes charged by the minute cost you a few cents is sort of the thing that grabbed people's attention and you really can't create a server in just a few minutes I'm gonna pause the slides here and we're gonna switch over to the azure portal so what I'm gonna do here is I'm going to pretend like we just had this idea that we need a virtual machine we need a Windows Server for some project and so let's see how long it actually takes from the idea to get anyone so I hit the stopwatch start now I'm not going to basically teach you or talk about exactly what I'm doing here we're just gonna pretend like we just had this idea I'm gonna call this new idea in this course we're obviously gonna get into how to create a virtual machine and all the wonderful stuff around now I'm gonna pick a bit of a bigger machine so we're gonna get the t4 machine then we'll use we'll get a Windows 2016 to give it the administrator user ID and password I'm going to want to RDP into that box I'm gonna skip all the other options it's offering me this machine at 16 cents Canadian per hour and if I just click create so just to get to the point where I've ordered the Machine a minute and 15 seconds here minute in 20 seconds from the idea to go and make me a machine let's see how long it takes for a sure to actually make it and I should say one of the cool features here is we can actually watch it in progress so creating a virtual machine is not just oh we just need a machine we also need a network we need an IP address security group and right now we're getting storage storage created so you know that's that's all the things that are involved in creating a VM all right so right around a five-minute mark it says my deployment is complete I can now sort of open it up and see that this has in fact done its work going to the resource within the portal gonna minimize the menu here we can see that we have a virtual machine running it's given me a public IP address and so theoretically I can use RDP the windows remote desktop protocol to get access to this machine using this IP address and the ID and password that I gave it so let's go back to the stopwatch sure so five minutes into it and we have a machine for use now as you know the first time that you run a machine it takes a few seconds to start so let's go into the connect button click the RDP filed RDP file comes to local I click connects remember I created a specific account for this like I set this when I created the VM I'm gonna say ok I have to accept this certificate and now RDP has started and so now it's gonna take a minute or two for Windows to configure itself and I'm gonna go create a profile for myself and we have to configure this for a particular role etc the server is just going to be empty it's just as if you installed brand-new version of Windows on a server that you have direct access to so at this point there's not really useful until I make it do something but the point is within six minutes of deciding that I wanted a virtual machine I am here we're looking at the home screen of this brand new machine and it's got a second gonna pop up with the do you want your PC to be discoverable see no it's gonna pop up with the role manager here and we can decide this is a nice machine or file server or some other purpose for it but that basically this is a demo of how quickly you can decide that you decide you want resources now I briefly touched on this earlier but basically I'm only gonna be charged for usage here so this machine is currently running and it started off less than 10 minutes ago at this point I could hit the stop button at this point and I sure would shut down the instance and make it you know release the IP address and make it no longer accessible to me but I then I could start it up again and it will restart so asher is basically going to only charge me for the seconds or in this case maybe 2 or 3 minutes that the instance was actually created and running right so it's actually fairly easy to clean this whole thing up let's go into the resource group and we can see there are seven resources that were created as a result of my request including the virtual hard disk the you know C the C Drive essentially network interface card storage for the Diagnostics IP address and stuff well let's say this ID I did spent a couple hours on it I was like okay well we proved that's a viable idea but we're done with this machine for now I I can quickly you know delete the resource group the entire set of it and just as soon as we had the idea within two or three minutes all charges will stop all the resources will get cleaned up and we will have no lingering costs associated with this so deleting and stopping the resources just as easy as it was to create it so in fact you know the power here is there's no contract I didn't sign a three-year deal to get that machine and no contact I didn't have to call anybody I didn't have to fill out a form and wait for someone to call me back I didn't even have to file a ticket I just went and created a resource used it and deleted it and not another human other than the bunch of you that are watching this video nobody else in the world knows that I did this let's talk about costs for a second so not only is that the flexibility of creating these machines we're literally talking about pennies per hour at this level so I chose this from the pricing table but the D level machine d2 with two CPUs and eight gigabytes of RAM is 3.7 cents per hour and that includes this is a Linux machine so there's no operating system cost and in this particular case we didn't need to we'll talk about reserved instances later in this course for cost saving purposes but this is you know sort of an option of what's available out there I would pay 10 cents an hour so if I if I just created a machine like this it's 10 cents an hour charged by the minute and so 10 cents an hour charge by the minute really is about you know one third of a cent per minute something like that and we use that machine just now for about four minutes and so maybe I just had about a one point two cents of cost for what I just demonstrated to you now I've been talking about virtual machines and that's certainly how you know AWS and then as you're started in that infrastructure as a service space but the innovation is not stopped and basically there are so many new ways of doing executing your code in the cloud you don't need to have a Windows or a Linux server and then do all the installation and do them deployments manually and all of this the old days stuff so now we have all these different computing services so we talked about virtual machines we have operating system choices asher has something called web apps which you just zip up your code package it configurations and send it up to Azure and they will they will run it you don't have control over the hardware you can choose performance tiers based on certain values but you don't get to choose two CPUs a gigabits of ram all this stuff it's not a server model and then to go what's what they've been calling server lists or functions as a service you can actually just create your code right in the browser you go into the portal we'll demonstrate that at some point in this course and write some lines of code hit save and from that point forward azure will execute your code whenever it's triggered so functions are something that's good for not even worrying about development on your machine it's just a small piece of code logic apps these are workflow apps it very much reminds me of windows workflow services or sequel server integration services and those types of no code models boxes and lines drag-and-drop connect things together events that fire trigger something else etc there's also service fabric this is not a super popular way of doing coding but if you want to break your big monolithic applications into dozens of small little services asher has some abilities to automatically manage he'll monitor and reporting all that wonderful stuff onto these micro services and scaling containers are taking over the world their Microsoft has really embraced containers in a big way used to be a sure container services and they moved to the kubernetes standard so as azure kubernetes services IKS but it's not just having a nodes and clusters and orchestrators you can also have containers in web apps you can have containers in service fabric you can have container instances which just take we'll do a demo in a second but we can get code running in a container in literally seconds you know it demonstrated a virtual machine that took about eight minutes to get started if you want to beat that time deploy code into a brand new container and it can be done very quickly we even have services where you've got jobs that have got millions of parallel tasks and we can scale up some machines in that kind of metaphor and that's just the compute right that's just the executing code in the cloud elements the stuff that the cloud can do goes way beyond just compute we're talking about databases many many different types of databases for all the different needs out there we're talking but not even data but storage right files and backup files and and websites and things like that networking services and includes some security in there like firewalls and things like that load balancers the machine learning is a big growth path and Microsoft has these cognitive services that they offer if you want you know text-to-speech and speech-to-text and image recognition and all of these wonderful things chat BOTS that's all in the machine learning category event hubs and queues depending on your application communication style then on the right we've got everything from identity and security elements analytics blockchain you know if that's something that excites you so all of these services are things that you can go into the cloud today and rent by the hour by the event by the execution etc so really cheap relatively cheap services that you can even don't duplicate this in your own data center without a lot of custom code a lot of installations is stuff like that so talking about containers let's quickly create a container and show you how quickly that can go all right we're gonna start up our stopwatch again and this time we're going to use the command line instead of the user interface that we did for the virtual machine once again I'm going to create a resource group again the purpose is not to teach you how to do this but to show you how easy it is to do so this is an empty resource group what we're gonna create our resources and I'm going to run a command here that creates a container a Windows operating system container and grabs an image from docker hub so this is a docker hub URL for Microsoft Windows iis and we'll let that run is going to take a or so but it's going to basically create a container running is and we're going to see very quickly hopefully that you know familiar screen that's the first screen of is when I excessively installs and there's no applications installed we're gonna see that sample app so we'll give that a second alright so we see that within about four minutes time here we have succeeded and so I should be able to go to this URL which I labeled it a Zed SGD demo and we have the sort of default home screen for I is when it's just first installed and that's running in a container and that took about four minutes as we saw so certainly Linux containers the great thing about containers is that you can package up your code like say you've got your own application is with your own websites your own folders everything all is installed and it takes the same amount of time to deploy so because it's created as an image sitting in docker repository docker hub and you're deployed to container and you've got it running now I wouldn't necessarily recommend the ACI that we just demonstrated as a production instance because it's a solo instance here there's no multiple nodes running there's no fault tolerance etc so we get into Azure Carinae service and stuff in future modules of this course but to get code running you can do that in four minutes five minutes got a public IP address even this one has a fully qualified domain name and so if you've got to demonstrate some code deploy some code super quick super reliable stuff like that deleting it so let's say we we didn't want it anymore a Zed container delete give it its name I called it demo container I see yes and look at that it returns in less than a second here so deleting things creating things can be very quick in the club so let's get back to this sort of high-level overview of microsoft azure of cloud computing we've seen how powerful it can be microsoft azure first came out in february 2010 to the public as windows azure and in 2014 it was renamed to microsoft azure 54 regions around world and that is the most of any cloud provider so comparing even AWS might have bigger revenue and a little bit bigger market share roger has sort of established the investments in the buildings and the wires and the computers and the locations exceeding even I sure in that respect they've got data centers and regions and India the Arab Emirates and even in South Africa we can grab this map Microsoft has the Azure regions page on his website and it's always updated there's always being regions being announced so those 54 regions means that it is sort of looks pretty well spread over the globe of course there's concentrations in North America concentrations in Western Europe Asia but it's starting to establish itself more in places like Australia we saw the Indian and UAE there's a single region in Brazil and I'm sure that's gonna grow and South Africa as a recent announcement too so these are all sort of growing over time and that's a lot of investment there now you might see that number 54 and you think wow I'm gonna go into Azure and I'm gonna create a server in imagine a region not all regions are available to all customers so there is the public cloud so we've been so we'll be talking about the public cloud mostly in this course but let's not forget there are three individual government clouds for the u.s. the German government has its own there at the German government the German population has its own cloud for German citizens and other restricted regions will talk about these terms you might hear public and private and hybrid when it comes to cloud public cloud essentially means it's available to anyone anyone with a credit card can just sign up maybe even takes a few minutes to create an account there out of those 54 regions that I talked about 28 of them are available for anyone to create there are some restrictions and additional six regions that are technically public but you can't just go and create a resource you need to qualify there are some minor differences between the regions and so you might really want this really powerful GPU server n-series CPU that has you know GPU capabilities and that might only be available in some regions so depending on the type of resource you want you may be restricted to some regions and not others generally most contents available in all regions one thing that happens with data is it becomes very sensitive as to where data is stored and I noticed that she does a pretty good job of respecting national boundaries and so if you're getting your data into Canada the Canadian regions while the data stays in Canada data in France stays in France data in UK is basically stored and stored in state as in UK etc so they do a pretty good job of respecting those national boundaries because the way that data travels over the world has become sensitive for certain industries now we're talking about the other cloud is the government cloud they have their own URLs even right it's not the same as the Azure public cloud eight regions currently and even within that it's fractured between the US government's like these federal state and local which goes into one and it even has its own like defense network in the secret network I guess this book for the spies and so you can't just get a virtual machine in the public cloud connecting privately over over Asher into a virtual machine in the government cloud there is restrictions there's completely separate networks not connected in any way Germany has one of the toughest data protection policies there's even a data trustee and there are when you're creating a machine or creating data you're going to have certain subtle differences between that in regular Asher and so it's considered its own cloud as well there are four regions in Germany now you may have heard that was the public cloud right anyone can sign up for those things I mean even the government if you qualify your state government or you work with them you can sign up private cloud those are basically those are clouds that do have those restrictions so the Department of Defense like you can't just get your credit card and sign up for a DoD cloud right so those things are private you can even download what's called a j''r stack install that on your own hardware within your own data centers and you would have a sure effectively running within your own data center and you can take advantage of those commands I just showed you how to create a container well you could do that on your own hardware in the same sort of manner with Azure stock and I know that it's been introduced a few years ago and it's slowly gaining some speed and they're talking about re are connecting it now so I wouldn't put a lot of investment into Azure stock right now it's serve definitely Microsoft's trying to find its place for it you can also have sometimes called the internal cloud or the corporate cloud like you if it's network it's just for the employees of your own company that would be a private club the other term you might hear is hybrid and so that is the combination between some sort of public cloud and your own network so it's used in a lot of places basically it's a combination of your own private software hardware applications and the public cloud so you can have your primary data center in your own premises and then if there's a disaster you have the backup ready to go in the cloud in case of a disaster so it's not used except in the case of emergency you're a sequel server has the extension option where you can push data from a local sequel server into the cloud take advantage of the cloud storage there are the opposite where you can have Azure public services connecting to private services in your network through a gateway relay any kind of mix-and-match between public and private now I do want to just spend another moment talking about a Microsoft global network here so you know we just talked about the 54 regions and how all of these things are available to you in the public cloud to be able to create computing resources anywhere you want in the world effectively for pennies but how is Microsoft able to do this well they've over the years they've invested billions and billions of dollars in these buildings and all this computers and the wiring and all that stuff here's a picture a bit of a slideshow moment here but here's a picture of a data center and Amsterdam this is a Microsoft data center in Western Europe here and you can see all the cooling units on top of it so inside there are a couple of floors of servers you go into it this is in Wyoming here but you go into it it's a warehouse just full of servers and Microsoft has standardized all this hardware and so whenever they're installing servers in brains building brand new data centers they just have a they know what they're doing there they're building it based on this plan I saw this picture in my my searches here of Microsoft Azure servers that are designed to run in the ocean floor so this is off the coast of Scotland they put some servers into this tube and they dip it into the water plug in electricity and in networking now the advantage of this type of thing is that there is just it's a sealed container there's no dust getting in there there's no contaminants of any kind there's no light so look you know inside and so those machines can actually just run should be no real hardware failures there's nothing that can get in there that can damage the hardware they're supposed to be more durable than a regular data center where people are coming and going the other thing too is it can take advantage of the water the cold water for cooling purposes so we don't need to run to all those air conditioners on top of a building and that costs money it has a cold water all around it and so Microsoft's is sort of innovating on these types of data centers now the other thing is not just the buildings and the servers but they've got their own fibre and so this is not running on the public internet actually Microsoft said I think it's the largest or the second-largest private network in the world so these lines all represent fiber capacity that as you can see on the west or through North America under the Pacific Ocean under the Atlantic Ocean all around Africa all around Europe and through Asia and so these are high-speed multi terabyte multi petabyte traffic's that can travel over the world and so when Microsoft is dealing between their North American data centers and their European data centers they're running on their own network not only of course is it causing them nothing to do that but you know it's protected from the travels of you know AT&T and other and other providers who are running the public Internet so they're investing in this in many different ways that we don't get to see effectively one of the things that I saw was they connect like Spain and Virginia for their data centers most trans-atlantic connections are between the UK and New York right so they're even protecting regional things in a slightly different way that you would think so now let's switch it up a little bit and let's talk about the buzzwords effectively you'll hear these things as being advantages or features of the cloud and we're gonna just sort of dive into some of these now first concept is the economies of scale now it's just the in fact I guess that Microsoft can purchase a server install it and run it for cheaper than almost anyone else in the entire world right there at that scale where they can run a server for a lot less than you can run a server ok this includes not only the purchase of the hardware because they can now go and negotiate if they're gonna go buy a hundred thousand servers from somebody they're gonna get a really great price that you're not gonna be able to get there was a slide that said they have electricity is about 20% of the cost of running a server on your own and Microsoft's building data centers next to rivers and they're being able to draw on the river for to power other data center etc so all of these ways that Microsoft's getting electricity cheaply there's certain countries if you're a familiar with Bitcoin countries that are known as having cheaper electricity costs and so Microsoft can install servers and get the absolute slowest rates only even the labor cost now you might think Oh Microsoft's employees are not any cheaper than my employees maybe they're more expensive but you're only running a hundred servers Microsoft's employees are running hundreds of thousands of servers each and so the cost per server an employee can just can just cost can be distributed over way too many servers and the other thing you don't really think about on the demand side is the utilization of the servers and so while you're running servers and you may have daytime and nighttime and certain capacities that go up and down seasonal Christmas time versus the summer seasons Microsoft's able to because of their global reach and millions of customers they're able to manage the demand nighttime daytime Christmas winter summer because of the way that the world works so they're able to get higher utilization rates out of their their hardware and magic then you would out of your hardware so the first concept we'll talk about is the concept of high availability now you may be familiar with this because it's not just a cloud concept but it's a concept in general the availability of a system is basically the percentage of time that whole application like the thing isn't at whole is able to successfully respond to user requests and so you can use the word uptime as a very very proxy or very similar word but it's not just the uptime of that server it's the connection to the internet it's the backend services the database that serves the server the load balancers and all of the services around it so it is uptime but it's not just uptime of a single machine now you might think ninety percent is a high score and if you pass your exams your parents might be pretty proud of you with a 90 percent grade but 90 percent availability in server's not that great it's literally two and a half hours of downtime per day that means you were you are aiming to you know be down from midnight to 3:00 a.m. every night and that's not a really available server right if you you know radically scale that up and say okay we're going to go for 99 percent availability that's still about 15 minutes a day of downtime I mean that's basically the time it takes to reboot a server and so if you said the nature of our server the name sure of our application is that we need to reboot it every day we cannot function more than a few days without a reboot so every morning at 3:00 a.m. we're gonna reboot the server that's act is technically 99 percent availability because you you can't keep that server running more than you know with 15 minutes a day it needs to be rebooted so it's pretty good veil ability maybe your customers wouldn't necessarily notice except if it's like everyday in the middle of the day unexpected that would be a problem now we get to the concept of high availability so that was just availability in general but high availability is intentional right it's intentionally creating a system that seeks to avoid downtime and that requires thought right that requires planning in advance and it's not necessarily the cheapest solution you're gonna have to go for higher-level more expensive services from the cloud providers including Asher you're gonna have to go for duplication replication backups all of this stuff it adds cost so to get from we were just on the last slide of 90 and 99 percent now we're talking ninety-nine point nine and when we get into these nines you might hear them referred to as three nines four nines five nines so three nines availability is a minute and a half a day and that's you know that's actually pretty good if your application isn't available from you know we're just randomly from 303 o'clock to 301 p.m. I mean you're not really gonna people will deal with that but that's not necessarily super ideal we get to four nines availability that's only seconds right that is nine seconds per date that's very difficult to achieve for four nines on a single server because servers need stuff installed servers need rebooting from time to time when you get into three 9s and four nines you're talking about having multiple copies of your application in different regions load-balanced etc now you might think okay my goal is to be 100% available 100 is the goal well that's sort of a mythical creature it's like the Loch Ness monster or a Yeti or something where you know you might you could set that as a goal I guess but it's generally not entirely possible and I say this very carefully because there are companies out there who are bigger than you and bigger than your client like Facebook like Google like amazon.com retail site all of these companies who do not have 100% availability I can remember just over a month ago a month and a half ago Google search being down I went to google.com and I got an error message and you stop for a second and go what what should what am I seeing here google.com is unavailable and then you refresh the browser and it comes back but that's not a hundred percent if I get an error from Google calm even if it's just for a few seconds that's down and there are times to when we hear about these cloud providers having regional outages and so your Microsoft Azure will occasionally have a bad deployment AWS will occasionally kill their a whole region and then you hear about these companies like stripe and and all of these ecommerce companies and eBay's and things like that having outages because the cloud provider has an outage and so it's a hundred percent is generally thought to be impossible and prove me wrong there I guess next concept we just thought so that's a high availability it's intentional right but you can't get to 100 but you get 99.99 four 9s five 9s those are attainable scalability scalability is the availability ability of your system to handle the growth of the users or the growth of the work for those users and we take a graphic like this it's really crude graphic that I drew but you can see that every system has a max capacity every system ok even amazon.com has a max capacity but mam azan dot-com sets their limits so high that it's higher than their usual spike in traffic ok I worked in an environment once where we did load testing and we you know ran a load test against our main website then you come back and it says it could handle this many thousand concurrent users and you find out where that bottleneck is is it you know in the traffic to the database is it in the i/o system writing files to the server is it CPU is it memory what is the bottleneck and then you you work on that you run the test again it's a recurring process to say what is your bottleneck but every app has a bombing every app so scaling is not the act of being able to scale but necessarily but be is it possible is it easy can you just add a server and boom you've increased your capacity so can you increase their capacity that means no code changes no refactoring of your code no reaaargh ejecting of your solution if you can grow your solution then you then you can say your solution is scalable and so a related concept to scaling is this concept elasticity right you'll hear especially with AWS they tend to have ile everything's elastic there you have elastic databases and elastic computes and elastic storage what is elastic rate that the elastic element is that the system can automatically grow and shrink based on the applications demand so the rubberband aspect so we had that graph and we said about scaling where we have a maximum capacity where you have an increasing user demand over time and that failure point okay but if your application had it was elastic then the capacity can grow and so as your demand grows capacity raises to meet it right and then when the demand eventually shrinks as it normally does then the capacity shrinks back down that is an elastic system auto scaling and you'll see the capacity is sort of staggered it's got steps not smooth because typically when you have four servers and then you add a fifth server the growth in capacity is not smooth it's you just increase your capacity 20% and then you add a couple more servers you've increased your paths and you know they're 20% so capacity tends to you know comes have come in bursts essentially and so this elastic thing the requirement for that is this the system can detect when it's getting busy and add more resources when it is and because we're in the cloud it's cheaper you'll end up saving a lot of money because your system is elastic now it's important to understand where these failure points can happen right so when you know when we're talking about availability and we're saying why isn't it a hundred why can't you just install an application into a server and it runs why well the thing is we're living in the world of computers and hardware fails right you have a home computer sometimes your power supply dies sometimes your harddrive dies sometimes the memory goes wonky that just happens hardware erodes over time those connections get a thing there's a little bit of corrosion going on dust gets in there hardware fails we live also in this world where you know your regional elements are out of your control right there could be a storm a flood power outage I mean I'm old enough to remember when the whole east coast of the US and Canada were down no electricity for a day and you know that that stuff happens and it's hard to keep your data center running when everything else around you is is out of power right so your batteries only last for so long and the gas and stuff lasts for so long but if the internet is down the internet is down we also suffer internet outages where traffic starts to fail packets are being dropped routers have gone bad it happens now we're also at the mercy of these cloud providers for their own deployments and their own software we're wearing our software but they're also running their software so it does happen that microsoft azure will deploy something to a region and that was just some unexpected thing happened a bad deployment they have to roll it back we don't forget we live in a world of security and we have to patch windows sometimes we have to patch our software Microsoft has to patch their own as your operating system things need updates and so that requires a reboot sometimes and that requires a downtime so downtime happens in a lot of unplanned ways and also even in planned ways which are the updates to get that high availability if you want that 99.999% what about your own code what about your own app you know how are you going to a deploy version 2 of your app without taking down version 1 well there are technologies for that like service fabric you can deploy version 2 of something and it'll just switch over with no downtime for users but that's something that's intentional thoughtful planned in advance is not by default but how can you update the database add to call and remove a column change the structure of something while your application is still running all those sorts of things don't forget we live in a dangerous world where denial of service attacks are a thing people can go to application and send a million visits at the same time to it and all of your infrastructure even if you're elastic and scalable gets overwhelmed your load balancer gets overwhelmed the incoming stuff gets overwhelmed an existing customers suffer also just mistakes happen right you can accidentally check in some code that's not complete missing a file and hit that deployment switch and suddenly you're on production it's happened to all of us it'll happen continue to happen I remember even with AWS happened once where somebody was entering a command line they were trying to remove an item from a routing table but they didn't specify the IP address range properly they made a typo if you will hit enter and the AWS engineer wiped out the routing table and AWS eastern region was down for six hours because of a human error it happens so which I'm trying to bring this home here yowzer goes down AWS goes down Google goes down every major cloud provider has their moments where they have some sort of operation hopefully not frequent but you need to design your applications to keep running during these things you don't know I mean I guess if it's a priority to you you need to how do you design your app if as your eastern region goes down how do you your app continue to run well there are ways to do that that's I'm not going to cover that during this intro here we're still in the intro sections of this course but getting a teeny high availability it's an architectural problem there are solutions for it all the time these companies including Asia coming out with improvements so as your front door service was just released last couple of months that's an improvement to the way that these things happen and can be managed on a regional level related to these things of scalability and keeping availability keeping it running what happens if it does go down right to keep you got keeping it running on one side the flip side of the same coin is being able to recover from failure and the two act two questions that are the key questions for recovering is how that's an acceptable amount of downtime okay so let's say you're down for an hour and it was a disaster and is that acceptable and let's see it's not let's say you're like no we need to be up and running within 5 minutes well how much data would you be willing to lose to get that up and running in 5 minutes you need to have your data backed up and stored in alternative locations and accessible from those locations and so how often are you doing backups you're doing backups every 5 minutes are you doing backups constantly is it a constant synchronization and so these things are part of disaster recovery so I wanted to wrap up this introductory section here to talk about where we're at ok so I started off this section talking about the continuum of knowledge where you're starting from the beginning you've got no knowledge very little knowledge of what computing is what asher is and you can go all the way through to the expert level where you're ready to attack taking passed the exam and you pass it with a great score and even beyond right there's more to Azure than even beyond what is tested at the expert certification level but I wanted to get people to sort of like a little bit farther down the continuum from the beginner level at least into this novice level like I said it wasn't my plan to get you to pass the fundamentals exam but if you take what I just did and go and take an AZ 900 course and go to take an eighth at 900 exam you might be pretty far ready for that but so hopefully we got a little bit farther down that continuum now I need to warn you right Asher is huge there is a lot to learn you're going to feel overwhelmed by this you know there are hundreds of videos to go through and not evenly that but hundreds of topics and hundreds of services if you look at the compute options that we reviewed there are eight or ten different computing Coda's and Azure database options there's eight or ten different database options storage options networking options load balancing options how to achieve high availability etc etc etc etc so a jury is actually huge and it's very difficult in any kind of course to be honest to convey expert level knowledge in hundreds of areas so there is a lot now think about what your job is if your job is to become an Architects are designing solutions for a sure well then your job revolves around the question of why okay we do need to understand all of the azure options all of the offerings so there is this broad level of knowledge in terms of what is available but why why would you choose that if out of those 10 compute options between web apps and VMs you know server list model with functions and in service fabric why would you choose one over the other all of them have pros and cons there are cost implications there's performance implications control implications and the expert level knowledge required etc so all of these things have pros and cons and our jobs as architects ultimately is to understand what's available and why you would choose one over the other and so a lot of what we're gonna be talking about is how does this work and why would you choose it now you'll find if you get if you're interested in getting into the exams in terms of our architecture exams that there are how elements to it so it's not just why right so Microsoft is going to ask you to create a VM not just understand why you would want to create a medium but they're gonna say go ahead and do it so we do need to go deeper than the just the why also I believe in terms of learning this is going to help you so doing is actually better than hearing it's actually better than seeing someone do it doing it yourself and so my intention in my hope is that you're going to not in this section but we get into various talking about databases and virtual machines and web apps that you're going to go and do it you're gonna you're going to say oh well I'm gonna go and try to create a web app and see how easy or how hard it was okay so doing is better than seeing and literally listening when it comes to learning I want to end with sort of a with my approach okay I'm notch I'm not here to sort of preach at you but I do want to encourage you I want to encourage you to not be afraid of it don't be afraid of Asher don't be afraid to ask questions don't be afraid to look something up go do a search in your favorite search engine that Microsoft has a great Doc's website documents and so their help website is van or try it yourself right you don't need to ask me to you know what happens when you try to create two web apps with the same name well go go and create a web app and then create a second web app with the same name and see what happens so that this is the great thing about a sure it's cheap or free to do certain things and you'll see for yourself and that's I think what you'll learn better than if I told you and I looked it up and I answered the question so don't be afraid of azor and don't be afraid to go and explore with it and play with it and you know expand your own understanding my final philosophy I put resourceful Ness wins and I'm gonna put that on its own slide here this is not just for Asher does not just for this course but for every course you'll ever take I think this is true for your career I think this is true for anything if you are presented with a problem there are two types of people if you're presented with a problem and you stop and you wait until someone comes along and helps you that is less than ideal okay you are not going to go as far as the person who's presented with the problem and solves it goes around it goes under it goes through it destroys it okay so there's no brick wall it's only in your mind so resourcefulness wins not only with this course not only with Asher but with everything no that's the only hopefully the only preaching is part of this whole thing but I want to encourage you to be resourceful and to not let anything stop you if I have a video coming up and it doesn't quite get into enough detail from something that you're really interested in don't let that stop you go find it and go go deeper okay resourcefulness wins Thanks thank you so much for checking out this video that was taken from my course for a Zed 300 Azure architecture technologies again to get as her architects certified there are two exams the A's at 308 and 301 if you are interested in training on Azure architecture there's a link to my course that's in the description for this video if you want to be notified when there's a new video that goes live for free go ahead and hit the like button hit the subscribe button make sure you are in that to get noticed Haitians I want to thank you so much for checking out this video I hope it was helpful to you and I hope to see

Channel: Scott Duffy @ GetCloudSkills

Views: 15,690

Rating: 4.8709679 out of 5

Keywords: Udemy Instructor, Scott Duffy, azure, azure architect, azure architecture, microsoft azure, microsoft azure introduction, microsoft azure overview, microsoft azure tutorial, microsoft azure training, azure tutorial for beginners, azure tutorial, microsoft azure tutorial for beginners, what is azure, microsoft azure fundamentals, azure microsoft, azure training, what is microsoft azure, azure training for beginners, azure beginners, microsoft azure for beginners

Id: _TUaTNc8czU

Length: 54min 24sec (3264 seconds)

Published: Tue Feb 18 2020