Transcript for:
Understanding Virtualization and AWS Compute

okay oh okay I think you were talking of mute unintentionally but let's get started so last week we we started looking at compute right we started looking at compute on a blers and we basically started with virtualization where we talked about what virtualization means I'm pretty sure some of you heard this term for the first time we talked about hypervisor as hypervisor is the software that powers virtualization as a whole we looked at the types of hypervisor we looked at their use cases because this is important as you start understanding compute as and and the level of abstraction that goes across the industry as people as different companies are using using compute to power their applications so we looked at virtual machines and servers ec2 instances on the AWS platform as the computer service that AWS offers we covered what the ec2 architecture looks like and then we talked about dedicated host we talked about shared host and then we started looking at easy to instance types so we said that instance types are very important because it helps you to pick the appropriate amount of resources that makes up your instance and when I say appropriate I mean with respect to your application needs right so it's very important to understand the different types of instances because this will help to guide your decision making or whatever recommendations you're giving your customer or your client for their applications so when you choose an easy to instance type when you choose an easy to instance type you're basically influencing the amount of resources that you want to get for that instance and by by resources I mean things like computer I mean things like storage I need things like I mean things like random access memory and the local storage capacity so the instance type that you choose has a part to play or is proportional to the amount of these resources that you're getting okay and that's why AWS has categorized the instance type into five main categories and these five main categories basically handles a proportion of these resources so we have five categories of instance types that handles a proportion of resources that goes to the compute server and we started looking at what those categories are we started with the general purpose category which is typically the starting point of every organization if you're not familiar with instance Types on AWS you want to start with or if you if you don't understand your application needs you want to always go with the general purpose category of instance so we have the general purpose category we always call it the starting point when you are not certain of your application needs so you use the general purpose category when you're not certain of what your application needs why because it gives an even resource ratios okay it gives an even amount of resource ratios so it gives an even amount of CPU it gives an even amount of ram it gives an even amount of network storage and all logical storage and all of these these things that's what the general purpose does it's it's really ideal when you're hosting application in Dev environments and those applications you're not very conscious or you're not very sure of what the application needs are or when you want to test something when you want to test something you can use the general purpose instance types oh I didn't know we had so many people in the lobby let me go ahead and make someone hor one second then one second so em I've made you co-host I don't know why my participant list wasn't showing okay um let me try let me pull it up as well there we go okay I can see it now okay so I was saying that we use the general purpose category when we're not sure of our specific application needs when you're working in the dev environment when you want to test a specific application that is when you use the general purpose category then we have the compute optimize category we have the compute optimize category and this category is designed for compute optimized instances or application applications that needs a lot of compute okay so high performance Computing application so it is designed for I performance Computing applications for example gaming applications so if if you're working for a gaming compy they will always want to turn to use a computer optimized instance type machine learning applications gaming applications machine learning applications they would want to use high compute okay and then things like media applications batch processing all of those who want to use high compute instance types so most of the time the application owner will tell you the needs of the application and then you'll be able to basically relate those needs to the different instance types that are there then the next thing the next category we have is the memory optimize the memory I optimize again it's giving you a higher ratio sorry mam sorry ma'am please The Bash application what is the spelling of The Bash is it b a t c yes that's right b a t c okay thank you ma'am you're welcome so memory optimized applications are giving you a high proportion of memory as as compared to your other resources so it's giving a high proportion of memory allocation when compared to other resources okay so again it's ideal for applications that need large in memory caching all of those applications that have to cache data so ideal for applications that cash data in memory so if you're hosting an application that cash data for a specific uh uh period of time then you want to go for the memory optimize instance types okay and then we have accelerated Computing instance type we have the accelerated Computing instance type please can you scroll up so memory optimize memory optimized there you go thanks you're welcome okay The Accelerated Computing instance type comes with additional capabilities added to the resources that's why you call it accelerated Computing okay added to your CPUs so added to CPUs The Accelerated [Music] Computing instance types offer gpus so CPU central processing units added to that would come which comes natur naturally accelerated Computing at graphical processing units to your incense types okay so it's ideal for graphic intensive applications applications that has a lot of Graphics like Netflix for example Netflix will go for the gpus or the accelerated Computing instance types so that it can take advantage of these GPU resources okay so ideal for applications ideal for applications that use graphics that are graphically intensive okay for example Netflix then the last one we have is the storage optimize storage optimize storage optimize the storage optimized instance type provides a large amount of super fast local storage okay so it provides a large amount of super fast local storage okay and it's ideal for applications that need super fast storage super fast storage okay all right um give me one second I'll be right back slide yes just take a screen so we started by looking at the general purpose category the general purpose category and then we have the compute optimize category that's designed for compute optimiz applications application that need a lot of compute then we have the memory optimize we designed for applications that require in memory especially caching application because caching applications are caching data in memory right so you need the memory Optimizer basically these are instance types that have more memory than other resources then we have accelerated computing thus brings in more um graphic intensive processing units into the instance type and then we have the storage optimize that's optimize for applications that need local Superfast storage okay when we're talking about the epimeral storage there are some applications that even though this storage is e Epal they would need it because of it high performance because this is the hardware storage right so um we we have these categories of instance types that organizations now just need to decide and choose and it's important for you to know because if you're working with a an application developer that doesn't is not aware of the specific different instance types that are there you can play a better role to uh recommend to them which one meets their specific needs okay another thing that I want you guys to know as far as instance type goes is how to understand the understand the instance type code so remember last last last when we wanted to create an instance we selected a t2. micro instance okay we selected a T2 on micro instance and this T stands for the instance family okay every instance category has a different family a category of family a category a family within that category okay so you need to know those categories and they are well documented on a documentation but I just wanted you guys to know the high high high level so this stands for actually let me bring in this is it's better to see it from here so if you look at this this is an r5d x large okay so R in this case is the instance family just like we see it here just like we had our T2 micro so when we had our T2 micro t2. micro T was the instance family T was the instance family five in this case five in the case of our Hands-On two this is the instance generation so when AWS creates these instance types every year they add more features to it okay so if you look at the RDX r5d extra large it means they in the fifth generation of that instance type with the t2 that's the second generation we have t3s which is a third generation so every year AWS creates different generations of that instance type that adds more features to it so this one is the generation and then this D this additional nature after the generation not all instance types have it for example if you look our t2. micro we don't have the additional capability but the D is stands for for instance types that have additional capabilities again not all instance types have that additional capability and then this is just the size extra large large micro um three extra large five extra large all of those different things it is just the size okay so it gives you um so the size basically means that maybe a t2. micro would have one CPU a t2. medium will have two CPUs a t2. large will have three CPUs again that's that's what it means by by the sign it's also important for you to understand these things so that when you're helping someone a client make these architectural decisions you definitely be able to tell them that okay I don't think you should go for the five extra large it's better for you to go for five instances of large as opposed to going get getting one very large instance so that you can automatically scale as needed okay and things like that Miriam so um understand that CPU stands for the compute processing unit right what does that mean so it's it's basically it's basically a resource within your bare metal that helps to boot or power your your operating system okay okay so it's a resour within your bare metal that helps to build the power your operating system and it's powering it but not but not necessarily accounting for the speed is that what it is no it's not accounting for the speed okay there are other resources that accounts for Speed you have Network bandwidth with your your your bare metal you have Network bandwidth which accounts for the speed between your your Compu instance and your storage service okay so accounts for the speed how fast and the amount of data that flows between those two end points okay okay thank you you're welcome all right moving on the next component of um compute I want to to talk about is instance billing okay instance billing or how your instances how AWS builds you for these instances it's important to know this so that you can know exactly where you can incorporate cost optimization where you can help your client save money okay now as far as instance billing goes there are three three main or or I should say four main different options that AWS offers its customers for them to um choose how they want AWS to build them okay and let's look at those four main options um real quick so the first one is called the OnDemand pricing model okay we're talking about instance billing and we said the first one is the on demand pricing model the on demand pricing model OKAY the on demand pricing model is the model that follows the pay as you go pricing so if you were here in our first last we talked about pay as you go pricing which is basically you consuming you paying for what you consume for the length of time that you consume okay AWS has different ways that it builds for different instances so it could build you on a per second basis it could be you on an hourly basis okay and based on that if you consume your instance for 10 if you run an instance for 20 seconds AWS will build you for that 20 seconds if you run an instance for 20 hours AWS will build you for that 20 hours that's what I mean by on demand and pay as you go pricing model so the on demand pricing model is the default pricing model that AWS offers okay so and and it's mostly used in cases where you don't you B you just you don't want any downtime on your applications so use when you don't want downtime on your applications when you don't want any downtime on your applications because your applications are running yeah cool thank you you want to go on mute denzo and then is used when you are your applications are running shortterm a shortterm task if if if I'm going to run my application just for a couple days and I'm not sure that I will run that application for the entire year you can go with the on demand pricing model so shortterm T it's used for short-term task for example you want to test something you don't want to make any long-term commitments on on on things that you're not sure of that's when you go with the on demand pricing model because long-term commitment means that you have to pay some upfront money and you're committing for a long period of time you don't want to do any of any of those things then you go for the ond demand pricing model it's good for new applications if you're just hosting a new quick question there was no please can you go on mute or you raise your hand when you want to talk I can't I can Okay so when you're holding hosting a new application so you're hosting a new application in your environment and you're not sure exactly whether that application is going to run for a long period of time then you can go for the on demand pricing model okay one thing you should know about the on demand pricing model is that is the most expensive pricing model so it is the most expensive pricing model as compared to the other three okay the the pricee that you see on the AWS documentation if AWS tells you or if you see the price in the in the in the compute Marketplace that's the price that you get charged okay so it's the most expensive now the second the second cost model for E2 instance that we have is called the spot pricing model so we've talked about the on demand pricing model now let's talk about the spot prising model the spot pricing model spot pricing offers the cheapest access to E2 instances okay with the spot pricing model you have the ability to save up to 90% of the on demand pricing so it gives you the ability to save up to 90% of the on demand price so it means that if AWS is charging you $10 an hour for an instance with the on demand pricing you have the ability to get that instance for $1 an hour using the spot pricing model using the spot pricing model however there are some caveats right with this pricing model and that's why it's very cheap so the first thing is that this pricing model is powered by the forces of demand and Supply the the the the the pricing model is powered by the sources of forces of demand and supply and AWS is aws's spare capacity so it's AWS AWS has spare capacity of instances that nobody's using but now they put it in the spot Market a will take it and say okay as opposed to me as opposed to us just leaving it sitting here nobody's using it let's put it in the spot Market but but but the price would fluctuate based on those who need it more okay so when you when many people need it then the price will go up when less people need it then the price will go down okay so it's powered by the forces of demand and Supply so it means that you can see an instance in the morning for $10 an hour and then after that after a couple hours you see that same instance type that you can go down to like a doll an hour because again it's powered it's like the stock market it's powered by the forces of demand and Supply so the reason why you may not want to use this port instance for to host your applications is because when you say when you tell AWS that I'm willing to use a spot instance say I want to use a t2. micro Amazon Linux instance but I'm only willing to pay $2 an hour for that instance meaning that you're telling AWS that this is how much I'm willing to pay okay AWS will take not of that okay when an instance when the instance when the forces of demand and Supply makes the market and the instance is available at your price AWS will give that instance to you however as the market begins to fluctuate and the price changes say to $3 an hour but you're paying just or you're willing to pay just $2 an hour AWS will take back that instance from you okay AWS will take it and give it to the highest bider give it to somebody that's paying more somebody that's willing to pay more that's why I say there's a caveat on the price fluctuation so what a lot of companies do right and when I say AWS take is going to take this instance back from you AWS gives you just two minutes AWS gives you two minutes once the price changes a gives you 2 minutes and then the instance terminates so if you don't have a backup process in place or some some some some you you haven't engineered some process in the background that will be able to take back up of your application and and maybe or transition transition it to another instance type then you may run into lots of issues okay but a tells you that way ahead of time so what a lot of companies do what I've seen a lot of customers do is they would choose a huge number of instances right you not just select one instance type or you not just select a couple they'll select up to like 10 or 15 instance types and say okay across all of these instance types we're willing to pay this amount so that if a T2 on micro becomes available we take it if some if AWS is taking it back then we flip it to a t3micro once it get it becomes unavailable we flip it to the next instance that's available that way you still save money right because anything that you get from the SP Market will always be cheaper than the on demand okay even if it's you're just saving 70% or 50% or 30% of the cost it's still definitely much savings as opposed to the on demand pricing so that is what a lot of customers so again it has these cost models have specific use cases so remember we said the On Demand is good for applications that you don't want any downtime the spot pricing is good for applications that can accommodate downtime okay that's one of the use case so if you if your application for example if you're doing P batch processing in the night when everybody's is sleeping you just processing things in the back end then you can definitely take a spot pricing instance because you're going to save a lot of money and even if you have some downtime because when it's transitioning from one instance type to another there would be some downtime so when that transitioning is happening it doesn't break your processes because of that so as long as your application can accommodate that downtime then the spot pricing model is good for you okay so let's talk about let's take that down so it's it's good so in terms of the use case we're looking at applications that can actually accomodate down time and I'll take questions momentarily okay I'll take questions momentarily applications that can tolerate interruptions that can um accommodate down time so for example I mentioned batch processing so badge processing applications a lot of customers have applications that process things in the night when people are sleeping do analysis do automated analysis in the night okay so these These are good use cases for spot pricing instances sport pricing compute imagine the amount of money you're going to save if you can run an application on instances for 12 hours without with with with like a 70% saving so it's a lot of money okay it's also good for applications that have flexible start and end times okay applications that have flexible start and end times okay the disadvantage here is just that it's not good for critical applications if you have any application that's critical that needs to stay live consistently then you don't want to use the spot pricing but if you have applications that are flexible then you want to use the spot pricing all right let me take questions before we move to the next one a thank you with the spot pricing is it that the company will indicate let's say they want to go for sever options of instances for the sport pricing is it that they indicate the price that want to pay for those instances and and when that price available a will give them that particular instance now what if the instance the instance are shut down or they withdrawn because the customer offer price is lower than the than the for of demand and Supply Price Right does that mean AWS will automatically transfer the workload to the other options that they fire in for the spot prising or do they have the responsibility by themselves to to to tocate the applications in all of those instances I don't yes I I I understand perfectly what you're saying okay so to answer your question straight up AWS is not responsible for transferring your your workload you have to engineer that by yourself okay but the good thing is AWS has given you all of the tools that you need to engineer it okay so say for example you have your applications or you have your E2 instances sitting behind an Autos scaling group we haven't talked about Autos scaling we're going to talk it about tomorrow but Autos scaling is just a service that helps to scale your resources Up and Down based on based on whatever criteria you set in place right so let's say that within the Autos scaling group you have four servers and all these four servers are using the spot pricing model so so say for example this is your Autos scaling group So within your auto scaling group you have some capacity that you need to you need to put in there so let's say that for example within these four servers they're all using spot pricing this is a T2 this is a T3 this is a T4 this is a T3 and then for some reason the T the price of the t2 changes and a BL was offering it for $2 an hour in the market and then he changed to $3 an hour so this one would go to termination okay the church is going the state is going to change from available to terminating mean meaning that AWS would AWS would notify you give you two minutes notification okay so it's left for you now to engineer processes that would leverage that two minutes to quickly take a backup so that before the instance is completely terminated then you have everything backed up or within your auto scaling group you can do that but now once it goes to termination the Autos scaling group will realize that okay this one is unhealthy because it's terminating and then it's just it's going to create another instance that will be maybe like T3 because T3 is available or whatever instance type is available but again like I said it's on you to engineer that a BL is not doing anything for you there okay yes sorry it gives me a p question because though I'm not going to the auto scaling group I think the auto scaling group uses a launch template or configuration I thought on that launch template you can only indicate a particular instance type can you indicate several instance type on the launch template you cannot with when you're using instances directly on the launch template but when you're using things like um micro services eks and ECS you have the ability to indicate multiple instance types and now when I talk about you engineering all of this C using custom engineering you can use something like a Lambda function and add it to that okay with the Lambda function now it would basically update exactly the different instance types that you want to use okay it can be F even though with your launch template you don't have the ability to choose multiple instance times but you can engineer it so that all of that update takes happens automatically without you touching it with your hands thanks for all right you're welcome Victor uh yes um I think youve kind of partially answered that question because I was going to ask um the engineering of this thing seems to be manual and therefore the hus is on the client uh to be able to figure out what they need and how quickly they can respond to these types of changes so it looks like uh you know the cloud architect you know is guarantee not a job clearly if we use this particular type of instance type yeah so it it's just for you to host applications that can accommodate that downtime right the length of the downtime would really depend on the application will depend on all of the things that have to happen in the background for the application to become up and running again so the downtime will vary per application so if you're if for example if you're processing something in the night like you're doing a bch processing in the night if it takes 5 minutes for the application to be down before it gets back up as long as it doesn't break anything then it's fine you would never see you rarely see a company use spot instances to host live application that's receiving live traffic right because trust me your customers are going to get mad at one time because of the downtime and the inconsistencies with your application so it's good for backend processing that's what you should know for fact okay all right my question is um during the moment where AWS is notifying you that uh a better price is in a demand and Supply force can you also double beat and maybe continue owning the instance um I don't know if AWS added a fature to put in a price range but if you put in a price range then AWS would would um I'm I'm not sure there's a feature to to put in that price range but um be as of my last knowledge the last time I use port pricing you don't have the ability to double beat for that particular instance right you can change it after that but to take it and then create another one that's why you have the option to select multiple instances and put in a range of instances right so that when because it's rare that all of those instances would not be available for that price range it's rare so so that if one is unavailable then you have the other one but I don't know if AWS has their future to double beat to double um beat the price that's something that we can check okay make sense yes all right let's talk about the next pricing model which is Reserve pricing model Reserve pricing model the reserve pricing model okay so the reserve pricing offers you or can offer you up to 75% of discount on your on demand price so it can offer you up to 75% of discount Off You On Demand price okay it can offer you up to 75% of discount of your on demand price and this really depends on on on two things it depends on the length of time you're reserving it and the amount of upfront payment your down payment right it's like you buying a house for example if you're in the US when you want to buy a house if you're putting in more down payment then you probably get less interest but if you're putting more a lesser down payment then you may have more interest that's basically how the reserve pricing pricing works so it it depends on the amount so depending on the amount of down payment or we call it upfront cost upfront cost and the reservation time you have the ability to reserve it between one to three years so you can make a reservation for one years you one year you can make a reservation for two years you can make a reservation for three years so you can reserve between 1 to 3 years and then you have the option to say to reserve with no upfront cost meaning that you're just reserving you're just telling making a commitment to AWS that if I'm making a commitment that I will use this particular instance consistently for the next one year AWS has a package of discount that is giving you or you can tell it that I'm making a commitment to use this instance for the next two years with a partial upfront cost AWS that they're giving you or you can tell a z I'm making a reservation to use this instance for the next three years and I want to pay everything right now that's when you get the 75% discount Okay C is giving you a huge discount if you're using it for three years with all the down payment immediately everything immediately so total upfront cost so each of these AWS has a particular amount of discount that's giving you and that's why you see I said at the beginning that it can offer you up to 7 5% okay because in some cases you get just 25% discount in some cases you get 50% discount in some cases you get 75% discount it really depends on the amount of time you're making the reservation and also The Upfront cost or The Upfront payment you're making to AWS but either way a commitment is a commitment if you make a commitment to AWS whether you use the instance or you don't use the instance you'll still pay the money that's the thing that's one thing you should know so when you when you're using the reserve instance pricing it's like you it's like you leing a house so when you Le a house for one year whether you leave in the house or you don't leave in the house the landlord knows that you still pay their money at the end of the month so that's basically the same concept here the last pricing model is called Savings Plan saving plan now saving plan offers it's similar to The Reserve PR pricing it offers something similar to The Reserve pricing but it it it covers it's it's measured more in dollars right so and it covers a wide range of comput so with Savings Plan it's not just an is instances is on compute as a whole so you can use your savings plan for other Compu resources is2 instances Lambda functions eks ECS all of the different computer resources that AWS offers without tying it part to a particular instance type or a particular instance you can you have more flexibility that's why I said it's the same as Reserve pricing but it offers more flexibility it offers more flexibility it offers more flexibility and more flexibility in terms of because it covers a range of instances so because it covers a range of compute sources covers the range of Computer Resources all right so that's that for instance billing whenever it comes to cost optimization on compute you definitely want to look if your cost or your client is going to save money by using by leveraging any of these saving options they can go for savings plan they can go for um Reserve pricing they can go for spot pricing on demand pricing should mostly be used in non life environments for say for example test environment or Dev environment where they're just spinning instances here and there and test and and terminating it but if it's a production environment and they're sure of what they will use in that environment definitely you want to take advantage of these saving options that AWS offers okay yep you want to take advantage of these saving off options at a office all right um Theo I see two hands up let's take Theo and then we take it all right EV I go ahead okay Pro please I want I want to ask um so for the instance we created which of the ricing models are we using is it the first one the on demand yes on demand is the default if you don't choose anything the on demand would be what you you'll take K okay thank you you're welcome that's a good question all right the next thing I want us to talk about the storage let's talk about compute storage compute compute storage okay so ec2 provides flexible cost effective easy to use data storage options in different categories so when you're using an E2 instance you're not tied to one particular storage options the different flexible cost effective storage options that you can use and the important thing here is to know those storage options because when you don't know you just stick to one thinking that you don't have options but AWS has different options right and and with these storage options based on your particular use case and your application you need you can use them in combination mean that you can use multiple storage options or you can just use one of the storage options based on your your needs okay there are three main categories of storage on AWS the three main categories of storage on AWS we have the block storage we have the block storage and the block storage is meant to meet specific needs and for specific types of data that you want to store we have the file storage which again is meant to meet specific needs and for specific types of data and then we have the object storage so we have the block storage we have the file storage and then we have the object storage and I want us to talk about those the block storage the file storage and the object storage so let's start with the block storage okay blog storage let's start with the blog storage the block storage chops data into blocks so to speak it chops data into blocks that's why it's called the block storage okay the way that the blog storage works is when you to explain it in in a the the simplest way for you to understand it architecturally let's say that we have a piece of document like this and you want to store this document in a blog storage the block storage when you want to use a block storage is going to take the data and chop it into small blocks so it's going to chop the data into small blocks and then it will store the data as small small blocks the reason why it chops data the reason why this kind of data chops data into small blogs or this kind of storage chops data into small blogs it's because it makes it easy for it to be compatible with different operating system okay so it chops data into small blocks because it makes it compatible with different operating systems and it speeds the process of booting the operating system this is just how the storage works okay this just is just how the storage works whenever you're using the blog storage to store your data that's how it works so it's going to take your data and chop it into small blocks and each of these block would have a unique identifier it gives an an ID to every block of data that it splits from that the whole data so in order for it to house the small blocks it creates something that's called a volume so in order to house the small blocks of data it creates a volume okay so if you want to visually look at it we're saying that we have something like a volume like this and then inside this data this volume we have blocks of data this is how the block storage looks like okay you have a volume again this volume this is what we call the volume this is what we call the volume and the purpose of the volume is just to house the the blocks of data that has been compatibilized with um with the by the block storage okay now when an application want to retrieve this data when an application want to retriev the data it makes a call and that call would map all of the different blocks categorically based on that ID remember when he's chopping the data it gives each block when he's chopping it into block it gives each block an ID when an application wants to consume that data then it pulls it back together so that the application can consume it as a single unit okay can you go back to the next page please page sure yes M okay so what get attached what gets attached to the instance is the volume okay you attach an a volume to the instance this is just how block storage Works whether you're using block storage in an instance or you're using block storage on a database server all of it works basically from an architectural standpoint this is how it works okay so if you ask me why does it why does it chops the data or users this approach to chop the data is basically because this block store most blocks most block storages are used to boot operating system when you're starting your computer when you're booting your system the block storage is used to boot the system and this when it chops the data this approach helps to make it compatible with these different operating system you can never use an object storage to build an operating system you can never use a fire storage to boot an operating system you use a block storage and it's just because of how the way that it works okay so if we want to look at an is instance say you have an E2 instance this is an E2 instance then you would always have a volume tied to that instance and that volume is going to be your block storage okay Miriam so does the volume have a unit of measurement and can then be many volumes attached to an instance good question so it really depend it really depends as far as an instance go that two types of volume okay the two types of volumes we have the instant store volume and we have the elastic Block store volume also known as EBS okay so these are the two types of volumes that we have on AWS now the block storage is only used to boot the operating system no so no it's not only used to use boot the operating system you can actually start data in the blog storage right but most block storage always have the boot storage within it okay thank you you're welcome lesle from what I'm so from what I'm getting the um what volume means is pretty much like an ID that it gives like each of the blocks that uh it breaks it into so that's is that correct no that's incorrect a volume is basically that physical infrastructure that's used to house the block storage all of the different blocks look at this architecture this is a volume okay this is the volume and then this is the blocks and each block has an ID so this one has an ID this one has an ID this one has an ID this one has an ID all of it has an ID and then when an application wants to use it then all of those blocks will come together to form a single unit so that the application can consume it when it but when it's just sitting and no application is using it it's sitting as blocks of data that is sitting inside a volume make sense yeah thank you you're welcome any other question so so how does this differ sorry I didn't raise my hand how does this since we're talking storage now how does this really differ from S3 so so so I'm just trying to put my head around it so the bucket you know S3 is like the volume right and the objects are like the blocks but just in a very general sense yes yes yes in a general sense yes you can look at it like that but now the difference between these instant store or elastic Block store and S3 is that this ones are block storage S3 is an object storage okay remember I said there were three types of storage okay S3 is an object storage you can store all kinds of things in S3 but the block storage is suitable for particular kinds of data especially esally when it comes to Performance okay all right let's talk about instance store and then we talk about EVS so the instant store I think I mentioned this last week is basically that physical physical disc that's directly attached to your bare metal okay this is the physical disc that is is directly attached to your bare medal it's directly attached to your bare medal okay and we talked about this the good thing about an instant store is that is a it is super fast it is high in performance there's so applications out there that would need a a they call it a hardware storage would need the hardware Hardware storage to perform the way that uh um the application needs to perform so it has the highest performance and now when we talk about performance of storage right what do we mean the two ways that you basically evaluate performance of storage you evaluate it using what we call iops iops and we evaluated it using throughput think about a car for example when you want to buy a car how do you measure the performance of a car horsepower right for those who are familiar with cars am I right yes you are okay so when you want to purchase a car you look at the horsepower of the car to to basically evaluate which car is more performant than which and it's the same thing with storage options so when you want to look at storage option the reason why we say that instant store is more performant is based on the iio and when we say when we say iops what do we mean iops is basically a metric that measures input output of data read write okay input out output the measures the read the number of reads and the number of write operations that happen per second okay so I I and O means input output per second iops means input output per second but basic Al that measures the read and write per seconds so this means input output per second and it measures the read and write operations per second it measures the read and right operations per second okay so this is basically the amount of data that is being read can be read or written per second if you can read and write more more amount of if you can have high read right per second then you per performing more than the other one so that's how we measure performance of storage and then when we talk about throughput when we talk about throw put throw put has to do with data transfer speed data transfer speed basically it measures the amount of data that is being transferred from one end point to another per second okay so write this down if you're not taking note please write write this down throughput measures data transfer speed which is the amount of data that is transferred from one end point to another per second the amount of data that's transferred from one end point to another per second so when we talk about instant store being highly performant as o as opposed to EBS volumes it's because it has high or high UPS and it has High throughput make sense or any questions so I I I just have one quick one the the um the very first option the one that you said is uh is um attached to the bare metal and is used instant store yeah instant store that's the one yeah so the instant store I remember you said something about instant store being um highly volatile if I'm not mistaken that youal it's also ephemeral storage yes ephemeral that's right temporary because it goes away as soon as the device is Switched Off that is correct okay y so it's it's known as an Epal storage because it's very temporary it doesn't persist Hardware Hardware failures okay so it's ephemeral because it doesn't persist Hardware failures this means that if the bare medal has issues and AWS has to fix those issues there's a high probability that your storage the data that you're storing in an instance store will go away if you start and stop an instance then you there's a high probability that you your data would go away if AWS has to shift your instance to another host there's a high probability that your data that you stored there is will go away and it is free okays doesn't charge you for the instance store but still there are some instance times that gives you good amount of instance store because it helps to boost your application performance it help helps to boost your application performance okay one thing about the instance store is that you can only attach it to the instance when you're creating the instance you can only attach it to that instance when you're creating an instance you cannot create an instance and then after that you came and put a instance store volume to it no you can do that with EBS you can attach an EBS volume to an instance after the fact but you cannot do that with an instant store volume okay you cannot do that with an instant store volume now talking about EBS volume ex can you repeat that again I I miss that pleas so I'm saying that with instant store I didn't I didn't write it down just because I'm running out of room but I was saying that instant store volumes can only to oh what am I saying can only be attached to an instance at lunch time as opposed to CBS volumes that gets attached after or that can be attached I shouldn't say gets attached it's like that's the only scenario that can be also attached after the instance is created and also be attached after the instance is created all right let's talk about um EBS volumes and then I'll take questions okay you back up one one page for the sure sure there you go thank you so ABS volume is also a block a block storage which means that it still has the same architecture where we have the where we basically have the block that sits inside volume okay so it's still the volume but in this case it's more of a network storage it's not like a hardware storage that gets that comes with the hardware it's something that is on its own and you get to attach it to the instance which means that it is not ephemeral it it can persist Hardware failure your instance can be down but your EBS volume Still Remains intact okay so ABS volumes on the other hand can persist are more persistent because you can persist hard where failures or not just halfway let me just say it can persist beyond the life of the instance I'm more persistent persistant because they live beyond the life of the instance okay and then you have the ability to encrypt or unencrypt your EPS volume so it can be encrypted it can be encrypted you can use EBS volumes along along with instant store you can choose that okay you can choose that and it comes in different sizes you have the ability to choose the size of the EBS volumes that you want you can detach it so let me so you can detach and attach to another instance you can detach it and attach it to another instance one important thing about EBS volumes I'm going to write it in the next page you can attach multiple volumes to a single instance but you cannot take a single volume and attach it to multiple instance okay let me write that down you can attach multiple volumes to a single instance but you cannot attach a single volume to multiple instances at the same time a WS however so let me put it however to know that there's an exception however AWS has something that's called multi attached AWS has something that's called multi attach which gives you the ability to get a single volume and attach it to multiple instances but it is on it's not it's not for all all instances it has a particular instance type and a particular volume that it works for okay that it works with however you can use the multi attach option which which is compatible with particular instance types so not all instance types have the multi- attach options okay go all right while you guys are getting up to speed let me check those three hands that are up Les go ahead not here yeah oh sorry my hand was up from last time sorry um Miriam I'm sorry this might be a stupid question I'm new to the it so um I'm just trying to gain Clarity by relating this to maybe re word scenarios so um when we talk about ec2 and all these that that we learning right now is it is it pretty much building a virtual computer for a company that's exactly what it is okay okay it's exactly what it is it's like your laptop your laptop comes with an internal drive right that you use to store all of the stuff that you that you use it's exactly what it is but now we're just saying that a is offering your laptop to you as a service okay so we're looking at all of the different features of of your laptop so that when you want to build your own you know exactly what feature to leverage and incorporate in that laptop okay okay and then the followup question so um H I don't know how to phrase it say just say what comes to your mind okay um so in this stage right now we are pretty much uh what is it building application that will enable the computer to run right are there going to be other stages where we what am I trying to say where this application will actually be doing the running if that makes sense so right now we are building application that will enable functions right and then down the road function will be doing the running does make sense we're not we're not building applications okay okay because in in real life the applications will be the the work of developers we're not training you to be developers developers will write those applications and give you to deploy it on the compute server okay for example when we talk about applications what do we what are we talking about we're talking about something like slack you have slack on your laptop right that is an application you have zoom on your laptop right that is appli that Zoom application was developed by a developer developer wrote a code and that code made an application so that you can download it into on you can download it on your computer but that's not that's not what we're dealing we doing with we're dealing with right now we're just trying to get the computer up and running okay then we're looking at all of the different aspects of that computer okay about the instance Style that covers the CPU that covers the storage that covers the memory and then now sorry now we're talking about the storage that gets attached the kind of storage that gets attached to the computer okay so remember with with with uh computers that had the exteral hard drive that you attach it to it right technically the same kind of thing we're talking about now so because you your Hardware comes with some storage but sometimes that storage will not be enough you need more and that's why you can take an EBS volume and attach it to it so that you can give more you you can have more room to store data okay we we will not necessarily talk about how you develop we will not talk about how you develop applications in our last class we simulated creating an acc2 instance anding into it getting into that instance right and the goal of that simulation honestly was not about the Apache application that we we used on the server to to be able to see the application running it was more about how you get into that server the process of sshing into that server because that's what you'll be dealing with if you try to install that application and it's not working it's not your business you just reach out to the application owner and say oh it's not working and then they'll fix it but now if their application is fine and they send it to you and you put it on your server and they cannot access their data that the application is hosting then they'll reach out to you that and say that okay why am I running out of space why is it I don't have enough space to store my data then that's when you now would need to make sure that you increase the size of the EBS volume so that they can have more space to store their data I don't know if it's making sense perfect yeah it's clear thank you okay Prince and so if you if you're using spot price Model and you're also using instance store volume and like when this when it switches over will you lose that data as well yep you lose everything okay V you're welcome Victor oh yes um I almost forgot the question but I think what I was trying to ask is you said something about a unique identifier now there are two pieces of of Hardware there's the volume and there's the blocks so which one do they both have their unique identifiers because which one do you reference the blocks it's just it's just the block the block okay the block okay all right so there are different storage types sorry ma'am can you go to the to the first page this one where we have the how can this one the first P this one no no the first page where you stopped where you stopped oh the last page this one the last yes exactly this page let me just screenshot it sure sure absolutely thank you ma'am you're welcome okay so remember how we talked about the performance of the block storage with EBS volumes you have EBS storage types that are high in throughput and you have EBS storage types that are high in iops okay the two types of EBS storage types so we have EBS storage types high in throughput we have EBS storage types that are high in throughput and these are called the HDD okay these are typically called the HDD and high storage types oh high iops storage types are called the S solid State Drive SSD SSD SSD okay so if your customer says that okay we need High UPS is more important to us than Thro then you want to go and look at the options that AWS has for SSD drives if your customer tells you that frut is more important to us than iops then you want to go and look for the EBS options that offers more throughput okay more throw and remember we say Throw what do we mean data transfer speed if your customer's business use case is streaming data okay if they are streaming data which means that data is being transferred from one end point to another they need that data to be streamed faster then they'll be looking for more thrp put than iots okay so again it's all it all depends on how you understand their use case but typically they'll always tell you now if they are using that EBS volume to boot their operating system or they are using it for databases SQL or no SQL databases then you would want to go for the more I helps okay because with database you have read WR operations people are reading from the database people are writing to the database so you need more iops than throughput okay let me put that in writing okay so what do I say throo optimize or HDD is suitable for streaming data because it offers High data transfer speed and then iops optimize as solid state drive SSD it's suitable for databases and boot operating systems because it offers more read and write operations off us more read and WR operations okay now it's always good to understand it so that you know exactly what to recommend to your customers okay if you most of the time customers don't know exactly what they need you don't want them to Fumble around when they have a solution and a subject matter expert Like You on the team you want to be there to speak up and make them give them high good recommendations so that they will not they will not make poor decisions all right now EBS has a fun a feature that's called snapshots EBS has a feature that's called snapshots ABS has a feature that's called snapshots and these snapshots act as a disaster recovery measure for your data that's stored in the EBS volume so EBS snapshot act as a disaster recovery measure for the data stored on your volume because even though even though the EBS volumes are persistent Beyond the life of the instance we all know that an instance is resilient only within an availability zone so it still means that even though the it persists if the in if the availability goes down then everything that you have on that volume will go down and you don't want that because it contains your customer's data so with EBS snapshots you have the ability to create a snapshot you can create a snap shot it will take a snapshot of everything that's on your server and then you transfer the snapshot to another region so with this you can create a snapshot and transfer it to [Music] another region as part of your Disaster Recovery strategy you can transfer you can take a snapshot and transfer it to another region as part of your Disaster Recovery strategy and this is what we're going to do for our hands on tonight so basically we're going to create an instance and then we on that instance we're going to do certain things on the instance so it's going to be a Windows instance right so with the with that Windows instance we're going to install notepad++ if you if you have a Windows computer you have no not pad on it but you don't have notepad++ notepad++ is something that you need to download from the internet so when we're creating that Windows instance we'll put in user data that would install notepad++ for us we also put in information to create a folder okay when we create that instance we're going to go into the instance Center we will see that notepad++ has been installed we're also going to see that the folder is there and then we'll take a snapshot of the EBS volume and then when we transfer the snapshot to another region we'll use that snapshot to create an instance and then you see that that snapshot has all of the stuff that you installed on the original instance that's how you test your Disaster Recovery strategy to see if it works in your environment okay now one thing you should know about snapshot is that snapshots are incremental snapshots are incremental which means that if you set a policy to take snapshots in your environment from say every hour for example at the end of every day it's not like AWS will be taking full snapshots all the time it will take just incremental snapshots so if you say I want to be taking snapshots in my environment every hour then let's say at 1:00 you wrote my name is Miriam for example and then you take a snapshot at 100: p.m. then between 1 p.m. and 2 p.m. you add more data and you say I am a girl the 2 pm snapshot will only take a snapshot of I am a girl it's not going to take a snapshot of the whole thing because when it takes a snapshot of the whole thing then it's it's consuming more room and is going to charge you more because it's duplicating the snapshot so a lot of customers are skeptical about snapshots because they think that it would be costing them more money because it's duplicating snap shots that have already been taken however AWS says that the snapshots are incremental in nature okay it only takes a the difference of the data between the previous snapshot and the current snapshot I hope that makes sense okay a awesome so that's really important for you to to understand it and then you have the ability to use data life cycle manager to schedule your snapshots okay you want to schedule it to happen hourly daily weekly all of those things you don't have to manually do it you can schedule it with that feature within um within um E2 or compute okay one questions ma'am the the recovery since you're doing incremental is it just um so if you wanted to recover my name is Miriam I am a girl um do you comb has a way of combining the 2:00 and the 1:00 to give you the full data that's a good question so a combines that on uh in the back end okay when you want to recover that data you just need to take the most recent and then you see that when you restor that most recent it's going to have everything that you had in all of the previous snapshots okay so it happens automatically and so this is one way that you you handle Disaster Recovery in your environment for your computer servers another way that you handle Disaster Recovery is using Amis okay we'll talk about Amis tomorrow but we're going to simulate Amis today by basically creating you guys saw Amis last time right when we're doing a hands on you saw that Amis is basically an image of your operating system you can have a Windows Ami you can have a Linux Ami you can have a Mac OS Ami by all of those different things it's just an image of your operating system so you can also use the two ways that you can back up your server you can either back it up by taking a snapshot of the EBS volume or you can create an image using Amis okay so as far as Disaster Recovery goes you can back up your ec2 instances by either creating a snapshot of the EBS volume or creating an imaage which is an Ami now when you create those and restore it then you give you your server you can create an EBS snapshot you share with another region or move it to another region you can even move it to another account and then you can use it now to create a server and that's what I want you guys to do I want us to do um after the break for our hands on where we'll see how Disaster Recovery works with this okay because when you're designing you definitely have to make your you make everything resilient and fall tolerant and make sure that okay if an a is down the whole environment is not going to be down we should have some Disaster Recovery strategy in place okay all right I see two hands up I'll take those two hands and then we can go and break prins Prince is not there let's take Emma Emma um sorry EV okay Prof um please I just want to be sure that I understand this well so when we save data on our instance we can can we edit the data you you can you can right you can ED the DAT because again this volumes when they get attached to the instance they empty okay you're putting data in there okay storage volume but you can you can go there you can write on it you can edit the data but there there are steps to it okay okay so when we take the snapshots can we also edit those like after taking them and moving them to another region is it possible when you take a snapshot of it in order for you to edit it you have to restore it right you restore it back to a volume and then because you can only have access that volume through an instance it's like your hard drive right or let let's let's just say your USB when you have a USB and you have dat data stored in your USB you cannot access that USB on its own you have to take that USB pluck it to a computer and then through your computer you have access to the data that you have on the USB am I right yes bro so it's the same thing with the volume even though you take a snapshot you restore it to a volume you still have to connect it to an instance before you can access the data okay thank you you're welcome Victor yes I just had one I don't know why that was a bright idea so so when you said you could actually do your disas Disaster Recovery using EVS volume snapshot or create an image it just occurred to me that you could also do the creation of an image if you were trying to capture um a golden template yep we're going to talk about custom Ami we call it custom Amis right because we're going to talk about that tomorrow because in most cases companies will not just take the Amis that you have on the marketplace and use it in their environment they would want to do some pre-installation on it they want to customize it okay so we call it custom Amis or golden images okay and AWS has a a service that's called the ec2 image Builder wow so it's a whole service that helps you to build it has a pipeline that helps you to push all the things that you want and build a golden image okay we'll talk about that tomorrow cool thanks okay you're welcome but that was a good question okay let's take we're going to take a regular 15 minutes break we'll be here at back here at 920 and then we're going to go into our handson and for our handson tonight we'll be simulating Disaster Recovery using images and EBS snapshots okay I want you guys to understand that concept and understand how it's implemented because it's really critical when it comes to um compute and saving your data against disaster okay so we're going last CL we we simulated our handson using a Linux instance this time I want us to do it using a Windows instance so now you guys get familiar with um how you get into a Linux instance and how you get into a Windows instance with our Linux instance which part did we open to connect to the instance who can remember 22 yes part 22 we open part22 for lindu and we call it s s but for our Windows we're going to open port 3389 for Windows and we call it RDP okay so these are the things that I want you guys to get used to so like somebody ask you this in an interview you can even close your eyes and you respond to it okay and then you understand the steps on how you get into those instances all right um let's take 15 and then we'll come back can continue all right see you guys later so um arrang but today we would be dealing with EBS snapshots I think Prof already talked about EBS um elastic block storage it's part of it's a type of storage that you can use for your virtual servers we will talk about um we would um create Amis from EBS from U EBS snapshots and also see how we can automate the process of having snapshots so because in most organizations example M again we have about um snapshots every day so there is it's always good to have um an automated process to um create the snapshots from your volumes before data life cycle manager came into place um Engineers on the job had to write custom scripts buy scripts and stuff like that to be able to back this data so AWS saw the need that oh people need to back up data they need to create snat UPS on these volumes and um they made it easier by adding this feature so it's not um very old I think it's like three two three years I don't remember but it's relatively recent so we would be doing this any question so far I'm assuming um The Silence means we all together great do we have we have yeah it's in the chat yeah Prof can you share the wrong book on slack to somebody beat me to it give me a minute oh somebody T already put it okay um hi Prof um Prof Susan had mentioned about something about downloading um notepad++ are we going to be doing that first so um give me a minute the notepad++ is going to be uh installed into the server by user script so I think you guys already discussed with process about user data which is basically a way you can customize um your server during the boot process so rather than waiting and ssing into your server at the end of when it's uh it's running and installing uh applications or whatever um software you need there before you can use the server you can bundle all that as a script and pass it as a user data during the launch at launch time and all that is um made available when the server Comes live so basically what we want to simulate here is we want to ensure that we just give me a minute okay great we want to ensure that um we customize the initial server so that's just like a company having a server in into that server they have installed the applications they have installed uh uh uh uh accounts for different use uh uh workers and stuff like that whatever information you can put in the server whatever customization you can do so you've basically customized your server and in this case we will simulate the customization by installing via user script node plus node Plus+ and also creating a folder in the machine and uh and a file inside the folder with some customized info I would share um the user data for you and we would do that in uh I would say instance one or the primary instance then simulate um the backup strategy Dr strategy by taking a snapshot of this instance moving that snapshot to a new re new um um region moving the snapshot to a new region and create an Ami creating an Ami from and launching another another E2 instance with that we can move into the new instance then ensure that the information that was in our primary instance is is also present in this in the new instance make sense chud doesn't make sense yes good yes it does yeah great so for this case for our handsome note note Plus+ is our uh sample application for the case yesterday it was whatever the Min man is using to host their website okay good so um somebody will have to share and let's start the user data already please I will supply the user data to you once we get them so let we start with scenario one um to simulate backups with evbs snapshot so the goal as we already mentioned we need to create an instance so you you can see that the things are building we've already we are done with um um vpcs and stuff so now we are putting is two instances into all those vpcs um they have their subnet they should have their right route uh write rout tables and stuff like that so just we we keep adding piling the information and at the end of it everything should uh fall into place and make sense who sharing please leard leard so we need to create an E2 instance so for our hands on today we would be using a Windows instance I'm sure we all know how to create these two instances already right Miriam yes OB yes sir we do good so that's this should be fast so I still have screenshots uh to guide you on creating the um instance so you we would um simulate this for our primary instance in us one which is North Virginia so we go to North Virginia region and we create an instance so you can give the instance a name I called it am Ami with with Windows you can give your a descriptive name as you team fit or you call it instance one whatever works and we need to select an Emi and we said today we using uh Windows Amazon machine image so you select um windows and that would give you the different um Windows available um Amis in your region so by default it picks uh Windows Server 22 2022 base which is free to available or eligible so you configure the uh different information concerning the instance like the instance type T2 micro so it's already then um selected by default T2 micro are we together I guess we can be really fast here right so after uh taking instance T2 micro let's go to Advanced details you need to scroll down to the end of the page towards the end Advanced details open the drop down menu and you'll see more information about it so um there are different uh features about your instance you can configure under the advanced details but what we are interested in is the user data section so I would share right now the user data script that we would put in there we don't need a key pair we will create a key you can create a keyp here you can create it at the end when you're launching instance some of the things I assume we already know how that's done right can we use the keypad we created for the last instance um I think so you can create another one Keepers don't cost money REM mind you you need a keypad if you select a keypad on the console and you attach it to this instance and you do not have it on your local you will not be able to associate into that server okay so ensure you have it so I shared a user data in the chat please Leon can you open it can everybody open it and we copy the user data into the user data uh um space it's giving it's giving me a security issue that it might be handful to my it's because of how the extension of it that's fine okay nothing inside that can have your system and uh also like so what should I um what should I just uh double click all right now you're already doing it with Visual Studio most most um scripts it's always good to use um okay Visual Studio Visual Studio ID that's with uh what is that notepad yeah can you close that and open it with Visual Studio hopefully your visual studi is already um you already installed all your plugins for let's see or I could also just take the take over the screen for a few minutes and I discuss with what what this does I think that's better so everybody can see it give me a minute you stop sharing or I think because of s i can basically override you okay so I want to share which screen is this screen can you also see my screen no NOP I'm already sharing oh are you see my screen now yeah yes sir huh yes good so um basically the DET where do I click next what after clicking on the advanced DET the drop down where do I click next to input the just give me a minute um I'll get back to you let's discuss the um user data then I'll get back to you okay make sense sure good so [Music] um this is the user data that you're going to put into the user data space so basically it's a simple powershare script I'm not a powershare expert but basically what this does is we are your screen we not seeing anything you can see your desktop not not what you're showing them can you see my CLE yes yes that's not what I'm supposed to be sharing okay give me a minute please I want to share screen can you see my screen now yes what do you see code so basically this is the um powershare script which you have there and it's a very simple powershare script like I said I'm not a powershare expert so just some tweaking was done on this and this basically creates a lock file called user data datal locks. txt which inputs every um uh that locks every information as the script runs and um there is a part of the script that downloads the not notepad++ installer the binary into your into your machine and and installs that and we also have um the code creating a folder creates a folder called gch Model B 2024 in your um Drive C in in the in the server which you're going to uh spin up and in that folder we also saying that it should create another file called info.txt and in the info.txt file write some information so basically what will happen is this is the customization we are doing into the server using user data which I am sure you guys talked about already hopefully and um we are installing not note uh part Plus+ we are creating a folder in uh Drive C creating a file inside the drive called info.txt and putting in some information there so you can customize whatever information you want want to see the other parts are just doing try and except ex try and accept if it um doesn't do what we asking it to do we asking the script to also lock some error messages into the um user data lock. txt so that we can know whatever the problem is make sense I'm sorry you're going a little bit too fast for some of us who don't know how to read this how do you know what is being done like what indicates um the action that is being um taken what do remember what indicates we are passing in um we are passing in a user data to customize our server as we spin it up as we no I'm talking about reading this like reading this looking at it how do I tell this is what I'm looking at this is what you know do you have um a little bit of of of of uh programming background but no if no we will get to it I think there's a section of our training that would introduce you guys into some programming like I said this is a power powershare script there are different types of script you will come across python you come across different programming uh stuff so this is a pow share script and the power script is doing a couple of things for us it's creating a file inside the server which we are going to launch in that file we basically a lock file we're creating a lock file and you're calling the lock file user data lock. txt and in this lock file we want to put in every information every error information that we uh um face while we launching the the instance okay but basically or what we interested in is we want to install um a notepad++ into the server okay and we're going to the script is going to to to the internet HTTP GitHub where the not notepad++ binary is in downloads the binary into the server and installs the binary okay so okay so the dollar sign is what tells us what we doing apparently the dollar sign is like um a variable so it's a placeholder basically you you keeping in some information there then you can always call that information and the subsequent lines of your script okay okay perfect yeah that's what I was looking for thank you so we installing that and we are also creating a folder okay we're creating a folder called J model BGE in Drive C in the server one and in that folder we are creating a file called info.txt okay so we are going the file is the dollar sign folder part as you can see dollar sign folder part is called jch mobile b.txt so we want to create a f info. TSC in J Tech Mobile JJ Tech model B 2024 and in that file we would script also write some information in the file basically a simple text file and input some information so at the end of it what we want to see we want to see all this information in our server okay this is the customization that happens so that was just a brief so you understand what the user data is doing sorry was that the file that you just sent into the chat can you resend it please because I don't know what happened to my laptop sure I can do that thank you so um Le can you reare so we continue yeah I suceeded in opening it good yeah that's the f so I reared it for the person that needs it so please just copy the user data and put it inside the user data section through in the launch interface Miriam are we together Terry yes sir we are great Franchesca yep great is it possible to do the uh test key uh key pair one more times paron Advanced details was a question to me use a and copy I want to use a you like take all my screen it's it's always good to use control a please just click in there and do control a and let me see what happens click just click somewhere yeah contr a good that's it the atts I do you just take over my screen contr a contrl c contrl c also copies then contrl V paste it's a very handy shortcut you you have to master it as you go forward Okay so once we have the user data in there we also need to configure our security sir please I have a quick question yes so for the uh key pair when we sssh we use the that perm are we still using the same for the window I don't yes perm should be fine all right so if you if you're um um connecting into your windows in since using a client called puty then you would use a pkk PPK I think but if you're using um the remote desktop I think in your for those that having Windows machines if you're using remote desktop I think it's called remote desktop uh DP what RDP desktop yes so you that P should work so for those for Mac users please you would have to download remote desktop from your app store and we would have to use that to SSH into the windows this Windows instance so let's continue please so we don't this should not take us up to an hour to be done inst so we can launch the in would I would like to verify your um inst the V VPC here yeah just give me a minute good so VPC is default we want to enable public IPS we want to create um a security group allows RDP traffic from anywhere this looks good are we all together um Can can you show the key pair part can I can I use any of this so this is the thing I'm I'm saying that yes we need a keeper and if you've launch an instance as I suppose we all have you can create the Keeper at this stage or when you launch the instance without selecting a keeper it's going to tell you that oh do you want to create one and you create one at that at that phase if you're using a key pair it it's it's in your console that you have this key pair ensure that you have the public key like we said key when you create a key pair there's a public and there's a private okay so the public is what AWS keeps and the private is what you keep on your local so if you use that public key to launch that instance and you do not have that private key in your local machine you will not be able to connect it to that instance so for those that are using an key pair that's already that already exist ensure you have the private Keys okay makes sense mean you should just create a new one it's better to create a new one like I said it doesn't cost money okay so I can launch right now great are we together are we all here yeah so what we've done is basically launch an instance I'm sure you guys have done that before we did this when we were doing what was it did you did you Creed a key paay yes you can create a key PIR so you can create a key PIR once you click on launch it's going to give you the option to either proceed without a ke if you don't want one or create an ke at that phase yeah did weate go ahead I was just going to say did we create one I missed it yes create one just create one please okay thanks quick question and also like group are we um are we in um put in the details now or are we going to do it that do that later it's already there we already checked if you go up to security groups we ensure that it allows RDP from anywhere and that's fine okay and we are assuming that you're putting your um instance in a VPC with um um everything configured and routing and also Auto assigned public IPS please we are building on to what we already learned concerning networking okay so I'm assuming that at this stage if you face challenges you can always talk and we would try to bring you up to speed but um do not forget what we already did because we will just keep pil information and adding okay make sense yeah can you drop I didn't know we had to we didn't have to delete so after you see where it says your lunch loog log drop it down the arrow let the arrow drop down um not there not there calm down okay okay the arrow yeah drop I just wanted to see if I had the same thing okay that's it thank you so some other person was facing a challenge can you speak up pleas yeah I was saying that uh you know I usually just delete every single thing after our sessions so I didn't know that I had to create a VPC but no if you you cannot delete the default VPC that's AWS default so if you delete whatever you delete there's a VPC there so if you do not select a VPC when you're launching AWS puts that thing in their default VPC okay that works to create the instance were you able to create the instance yeah yeah well the fun thing is uh I I get I got a error message saying that um you know I had no VPC available but then there was like the uh hyperlink where I could create a default VPC which I clicked and that automatically just like created one for me so okay good so um there is no way you can create an in instance without a VPC right because these are things that are VPC bound you must keep it in some in some sort of network that's why we started with VPC before we got to the stage Mak sense good so are we all together Chantel so I have a question yes um I was build for $8 just for VPC of which I stop the Gateway and every other stuff and that bill was for VPC I was like VPC go go to cost Explorer Go to e and really drill down into where the cost is coming from at a high level they will say VPC but you need to go see the networking company that is costing you that money VPC just creating a VPC doesn't cost it's free I'm what you put in that VPC that cost where are you cost management please can we do this at the end of the session this would drag us okay we can look at your your your your build at the end let's let's finish up with our hands on so those who want to leave can leave are we all together do we all have an instance yes great if we do have an instance the next phase is to create um snapshot of our EVS volume so you select the instance which you just created yeah I cannot see my my pen right so basically you select the instance are we together yes once you select the instance then you it will give you um detail information for details status and alarms monitoring security networking storage select the networking tab the storage tab that will display information concerning the storage this instance e EBS hello there somebody with music in the background please Opie I think you're the one can you mute yeah apiz I justed sorry so you select on storage then this would give us the storage that is attached to this instance which you selected and this is the volume this is what we want so can you click on the volume that will open a new console for you are we together yeah um Lut um snapshot says um Zer stroke to mine says sorry zer stroke five mine say zero stroke two yeah if you go to volumes how many volumes do you have here uh I'm sure I'm sure it's du to the instance it's us to the instance I have yeah how many instances do you have I have about five instances yeah so I think it's that's how where the information is coming from basically is saying that recently backed up volumes no volume of no volume has been in backed up in basically about five that you've had or something like that so that's that's not should should not be a cost of concern for you okay okay if we are all here then we need to select the volume once you select the volume it will display more information concerning that volume you see you can see this volume yes you can name the volume you could also tag the volume during the instance creation process that this will that will give it a name here okay okay um Leonard is naming his volume Windows volume once you select the volume you can click on actions top right and create snapshot are we together just hold here no one second please um I'm sorry what did we do so you select the instance the inst okay you just Vol yes once you select the volume give me a minute I take it all over from the beginning select the instance once you select the instance it displays more information for you you select the storage tab storage tab will display storage do attached to that specific instance and you can then select the volume you click on the volume once you click on the volume it opens the volume console for you there you will need to select the volume that we just created you're not can you be going doing that as I'm speaking storage go to storage click on storage good you click on the volume not yes click on the volume that opens the volume console MH great then you select the volume M once you select the volume then you click on actions create snapshot okay I think this steps are in the Run book yeah they are so once you click on create snapshot you need to provide details for the snapshot that you want to create a description for the snapshot and if you want to give your snapshot a tag then you can create a tag are we together Jude Jude a are you cameronian I give your give a name for your description um add a description for for your screenshot so I was just trying to ask like what example in in the he wrote Windows EBS snapshot for AMI equation Windows EBS so description is op right it's basically describe what what you will see and you would you would um understand what is this is for so so if I say to replicate to replicate in another region does it make sense because the description also no but it also depends on if you understand what you're doing right yeah I'm just saying generally whereever you see in an ads resources description is for you to give an information about it something that maybe later you would just look at it and say okay I know what this is about or you would um your colleagues will also look at it and say okay this we know what this is all about so it's just describing it all right so mine is Windows EBS now snapshot for AMI creation so that I know that this was the snapshot for AMI creation you I'm sorry Windows EBS snapshot for AMI creation that was my description am amand J jenta Emma are we all together the matter yes thank you say I'm good great once you give a description then we can create a snapshat please scroll to the end of the page and you create a snapshat so just FYI maybe it's something some of you have noticed when you're creating a resource for example a volume an instance during the um details you they say give it a name but once you create that resource you still don't see the name as have you guys noticed yes yes because if you want to see the resources for example this volumes where you have Windows volume if you want to see the name here in that resource almost at the end of every resource there's something they call Tag this is what it is that tag that is displayed here make sense M so once you're creating a resource and they say give that resource a tag it is a tag that will be displayed here not the name you you gave to this resource when you creating it can you come again please I was just an outside anyway what I'm saying is once you're creating a resource with AWS there is always the option for example create a VPC you to give the VPC a name right there is a difference once you give that VPC a name after you create it or for example a volume you give this volume a name when you creating it but once it's created in the name um uh field we still it's still empty because the information that is displayed here is the tag of that resource during creation make sense so during during the cre creation of a a resource there's always always a field at the end which is optional it's always optional T yeah right so if you want to if you want to display that information here or like the name here you need to fill that back F okay yeah right that's basically what I'm saying and now we create a screenshot are we together yeah yes sir so you've created a snapshot so we need to copy the snapshot to another region so we were working in North Virginia now we want to copy the snapshot to another region this is Dr so for example our servers are running in North Virginia what happens if the whole region of North Virginia breaks down okay happens it's can be an earthquake it could be flooding and all the data centers in North Virginia are gone so we can do a Dr scenario by um copying your snapshots into another region and keeping them there so that if North Virginia is down you can spin up instances in Frankfurt you can spin up instances in Ohio Oregon and stuff like that okay are we together yeah M good so in order to copy the snapshot to another region we need to go to snapshots so on the EBS volumes go down to snapshots so you scroll down to EBS volumes you know the other side good and snapshots good so you can see that you would have one snapshot which is the snapshot we just created if you remember what I was talking about because we did not give it a tack that's why this feels empty you gave this volume a name or the snapshot name when you were creating it or something like that yeah so we want to create a snapshot we want to move our snapshot to um another region so for demo purposes we're using what region is that Us West two so you select the snapshot and you go to actions mat tax go to options and it say says copy snapshot copy snapshot so you you select a snapshot go to actions and copy snapshot this will give you a menu or a display box for you to um um key in more information about your snapshot where do you want to copy it to Miriam are we together belus yes sir we are together ITC what is that it checkout that somebody's name yes yes sir we are together okay abdalah yes sir Kinsley and Flor hope we all doing fine and together so you put in the information here so we want to copy our snapshot to another region you give the snapshot description you select the region where we want to copy our snapshot to so AWS will display all the regions for you so you look for the region where you can send your snapshot to so we want to do send our snapshop to us West 2 Us West 2 you having EU to you could also use that but for demo purposes just use us West to so somebody don't go to us or you don't go to us and say oh I don't find my Snapshot so she use what EU Us West Us West us you're all in the US somebody's in London I heard so you can if you want to use Ireland it's an Ireland region and London region you can do that but for demo purposes let's all use us West 2 once we have we selected Us West 2 we can encrypted click on copy snapshot so snapshots can be encrypted and decrypted okay so you can encryp once you click on encryp it start asking more information about KMS those are things that you understand once we do KMS so let's just leave it blank for now and we uh click on copy snapshot so if you want to give the snapshot a name like we said you add add a tag so tag the key would always be named then the value is whatever name you want that um uh snapshot to have so key is always name key name name name name name name so value so you give the value how you want the snapshot to be called then you copy snapshot are we all together once we copy the snapshot the snapshot will be copied from the US from North Virginia to Oregon in us right Us West 2 is Oregon yes it will be copied from us West one Us East one to us West 2 copy copy snapshot then you would need to switch your regions in order to see this the snapshot so we go to the new region us W two oh so the the process is still running in the background okay yeah so this volume say is it 30 gab yeah that's the default that was given based on the instant size so if you during the launch process if you want it to be 500 or 5 terabyte you can change that at that at that phase I think I'm almost you think you're lost yes so after we um copied Snapchat what what did he do next snapshot you just copy it to um you did you select snapshot you selected the region you sending the snapshot to yes that was us um West two yes two then you clicked on copy snapshot yes then you switch to the region where you're sending the snapshot to you can see it there um toon us okay yeah so in new region we're waiting for the snapshot so somebody right now is traveling from from North Virginia to Oregon with a snapshot so to speak but it takes time though I just did this it took about like a good three minutes yeah yeah you're moving data from between regions 3s is is is wonderful right imagine you had to do this with a physical data center you need to get some trucks and how many hours would it take you to move from Virginia to Oregon okay completed um I think you should refresh your page so once you have the snapshot in know region refesh but down there you can refresh your browser does the speed also depend on the fact that we using the free tier or has to it okay your computer what what does the speed depend on the fact that we are using the free tier or not necessarily no I do not think I'm surprised mine is completed so it's maybe your latency too yeah I have a lot latency your latency so let's give him another minute good now his snapshot is in Oregon so in Oregon so please visualize this we have our insute instance that's running in one region we for Dr proposes we've created a snapshot of that is2 instance and we are sending it to another region so your applications keep running in North Virginia but if something happens then you can spin up the same server using this snapshot in another region or you could still send back the snapshot to to to North Virginia and create it once North Virginia is back online okay if there are any questions please you just you ask or keep your raise your hand and I'll will get back to you so now that our snapshot is ready and available we can create a new Ami from it so in order to create an Ami from our snapshot you just select the snapshot go to actions come quick question in uh in that transfer is it generally advisable to encrypt it or just leave it unencrypted actually it's advisable to encrypt it so I do not we did not get into encryption now because you guys have not done something we call KMS okay so once you say I want to encrypt this it will start asking you for the key to encrypt that information and then you will start you would need to select the KMS key to do the encryption so once we you do KMS most of these things now would make sense you would understand it because if we start talking about KMS keys I think it's a lot of information at once so some people might get a little bit confused or overwhelmed okay but normally in your in your working environment these things would always be encrypted you would hardly see it not encrypted it's possible for for for known prod environments but for production environments this things are always encrypted are we all together thank you Mao I'm with you great so we want to create an Emi from our our snapshot you remember that each time you're creating an instute instance you're always selecting an Emi an Ami is just a pre preed image that AWS prepares for you some other big customers have also prepared their Amis and they in the AWS Marketplace which you can use companies prepare their own Amis so for example you might be in a company and they take the uh the base image from AWS like we just had and they bake it bing here just means customizing so the the the word is b they bake it with their information customize the the Ami to what they want and create a new Ami from that so that just means now we have taken our base image from aw has customized it and we've created a new Ami so we in internally if you want to launch a new instance or a new server we can then use our customized Ami so this is um in summary what we doing here so from the snapshot which you created we can then create an Ami so you select a snapshot and you click on actions and you have a drop- down menu from which you can select create image from snapshot um prop the create volume on that uh actions drop down will that mean something like making a copy of a snapshot is that what what it means yes okay so we are creating an image for now okay okay just wanted to know yeah another advice I would have for you guys is play with this console it's your free free account play with it understand most of the things that are there all right don't be Sate on clicking just click go around and see what's there at times just click and see what happens all right and make sure you clean up after that because you can click somewhere and it starts AWS up meod starts running I so that's how you get used to it okay so I have a question please yes so I wish to know what's the the function of the am Ami that we are using here what is the rule like in this process generally an Ami is pre big image for your your server all right if you have a your laptop it comes with an operating system there right it comes with some um how do you say for example if you buy a new Windows machine it comes with uh the the Windows operating system it comes with some windows applications already inside right so that's basically the same thing in in the cloud so we want launch a new server what do we want to launch is it um an Ubuntu server is it a Windows server is it a mark server is it a red heart server so all those uh preinstalled or pre uh a uh um what's the best word configure default all those defaults are already in the image so the in the image so the image in this case is just a container for all the defaults the OS you want the default applications that come with that OS and stuff like that okay so a I'm sorry I'm still waiting for my um copy to appear is that normal I think I am lost where are you waiting for your copy to appear keep refreshing yes I've been doing that but it's still not um showing you maybe have low latency to come okay so we said create image from snapshot so it pops up with um a menu for you to ke in information about your new Ami your new image so you give the image of description and you scroll then I want you to look at the Ami give me a minute can you scroll up you're creating an image from the Ami so once you give the image description you give your image and name then you scroll down and you launch create an image you create you create image are we all together leet yeah I'm event are we all together so I'm with you you can add your tax along the way yes bro good so um please launch create your image we should be done in under 30 minutes okay today should be long uh short tomorrow I think we'll have a little bit longer session Miriam did you get it no ma'am not yet I don't know if I did something wrong or it's just that low latency refresh your browser please been [Music] doing once you create give me a minute we come back to M miror once for those are here once you create this the IM then you need to to go to the Ami section of EAS to console do you need to select okay oh sorry so you go to Amis and you should see your your your Ami which you just created are we together yeah if you look at the here it's written here owned by me so this is yours can you click on this drop down menu yeah so if you want to see other images you would see you you you select it okay if you want the images from AWS which you would normally see when you're launching an in instance from scratch it's public images so if you click on public images then you will see all the different emis from awsn okay so if you want to see your EMA your Ami you need to click you need to select owned by me so you own that Ami so nobody is saying it so Victor's Ami is not seen by Mao and stuff and and and and VI ver I'm good Miriam are you still facing issues yes sir I'm still waiting what's the issue I don't know did did you change your region to make sure that um you change it to Oregon where you lost it's the Oregon region that I'm waiting um for it to upload it's still pending you you you do not see your image your snapshot there no I don't is it in the pending state or you don't see any information I don't see any information it says you currently have no snapshot in this region yeah did you select oron yes I did I'm curious I'm not so sure you did maybe no I was very careful about that yeah we are always careful but at times we still make mistakes okay maybe yeah because if you selected Oregon then it should be there I think you selected a different region and because you're not there you can't see anything okay so is there any way you can go back and check what region you selected this is troubleshooting no oh wow so just go back and do the same thing and ensure you select Us East US West too check on us and the E2 yes go to us east2 what's that and check and see maybe that's what you what you selected EU I'm 99.9999% sure that you selected a different region yeah so just do it all over again okay just select the what was it again the snapshot and copy snapshot to a different region okay all right quick question Pro so in say in production uh you happen to have sent it to the wrong region how do you walk that back because like she's just uh sent it to maybe sent it to the wrong region how can you now retrieve that and um or stop it or there's nothing you can do you've sent it to the wrong region you go to that region if you want it to another region you can still send it back to where you want it to go to or you send s you send another copy from the base image from the base um yeah from the base snapshot makes sense yeah but no one will have access to that snapshot right nobody can see it except you just in if your if your colleagues have permissions the same permissions you have in the account they will see it okay so this all now boys back down to I am is the person is your is your colleague having I am permissions that Grant grants him here uh uh um access to e to if yes then the person can see it you if your permissions are are really fine greened to such an extent that because I can give you permissions for ec2 in Us in in specific region if you go to another region nothing happens so it's based on how uh the um Authority that the person has based on the IR am policies you attach to the I am identity okay I think I have a question that lady sent that thing to London how do I know because I sent mine to London I see in lond all right let me check all right let's continue like can um you know if we created the first instance without a Keir because I think that's that's the issue that I'm having so you know I went out there and now I need uh I need to find the key of the first instance so I can put it into the second instance but I think that so you're you're already way ahead of us right yeah yep y we'll get to you all right got you so let's continue please um from this Ami you select the Ami and we can launch an instance from this Ami so just go to the [Music] Ami select the Ami which you just created and then your this becomes highlighted or this becomes active you can launch an instance from Ami so now we are launching a new instance in Oregon using the Ami we just created so you give the instance necessary information so you give it a name um a little bit faster so if you see here Ami from catalog it already selects the Ami which you created if if you give your Emi a name you would see you should see that here oh Leonard give me a minute please are we adding another tag no he just he's just trying to customize his stuff so give your instance a name and let's go so do you do you create another key pair so you can proceed without a key pair okay because um or create one another keypad it doesn't matter but you can proceed here without a keypad what this means is if you remember you created the primary image with the keypad so this as we go further you you'll see what happen so Pro you can proceed with our kead to launch this instance she doesn't on proceed without ke wait wait wait wait wait wait come on I should proceed without I said click on I wanted to you to read something he say not that you will not be able to connect to that to this instance unless you already know the past where into into Ami yes so because you're launching this instance from an Ami you created from another machine if you do not give it the keeper you you need the password from the primary instant and we will get that password as we go ahead okay so should I just or what do I do proceed without keeper and and create instance okay yes guys I actually found it I sent it to Ohio sorry congratulations told you I told you I was very sure that you sent it to the wrong region that's very common okay um it's very very common to be honest I was caught into an incident about two years ago or something and the customer was freaking out that all their in their infrastructure is gone somebody deleted everything meanwhile they were just looking it in the wrong region so they looking for their servers no server and there was an escalation and somebody have deleted our whole environment but they looking for the information in the wrong region so it happens it's very common so now we want to click uh Connect into the instance we just created so just like we connect to U uh instances which I'm sure you've done before you click select the instance you click on connect at this juncture Mark users you need um s into Windows instances with directly from your terminal so you need RDP so remote desktop client so you need to download remote desktop what um os are you using this is a Windows user okay are there mark users here who us so you need to download remote desktop Clan I want looks like you just go to your App Store and type Remote Desktop Client I can send a screenshot can I get a screenshot promot desktop CLI please let the mark user share so we can just do that for three four minutes Prof we connecting to the second to the instant we just created right not the first one right not yes connecting to the we connecting created with the image with the snapshot yes please scroll down let me see remove your search and just type remote desktop remote space desktop what happens remove your search yes this is it right here remote Microsoft Remote Desktop um which one is it yes or the first one yes first one get that's what I'm looking for so there there there are so many uh client applications which you can use all right for s for these things there are so many there's one they call um I forgotten the name but I saw it on your terminal there on your interface so this is the client which we need remote Microsoft Remote Desktop so all Mac users you need this to be able to connect to your to this Windows instance do I need to do anything here is it already did you already download it I believe it's downloaded yes so basically you can if you come to add PC that you would give you a window to to to add PC and give give it information so PC name you look in there they say host name or IP address so you need to copy the public IP of that PC and put it here how however there's another shortcut so let's go back to who was sharing or if you want to continue sharing that's also fine for me are all Mark users in sing are we all together yes sir great okay um Windows users I'm not very ver with Windows I'm sure we'll get it through so Connect using RDP so click on download remote desktop file so if you read this it says that you can give me a minute it says that you can connect to this instance using the remote desktop client or by downloading the running and running the RDP shortcut file below so AWS is trying to make your life very easy are we together yeah so if you download this clan click on download download remote text file download remote Tex file okay then you run the file so my says so this is the interface for Windows guys so you see it has already given you the information username and password you need username and password to connect to this client so you need the password for the server where do we get the password of the server if you remember the message you had when you proceeded without keeper it told you that you need to know this password from the machine where you created the Ami from or the snapshop from franches are we together no I think I'm lost when when you click on the download which one am I to select I have the Ami and then I have the new key pair I'm not so sure you where we are and when you click on the download remote um file can you do that again uh Leonard just yes download remote file this would download something into your local machine into your downloads folder just like keeper so you go to your downloads folder then you will see that file what's the name of the file please it's going to give it the name of the instance which you had so okay so it says um I am I Ami instance so I so that's the instance you gave the name you gave yeah so you just run it run it by double clicking oh okay the mark user are we together Mar I think you it's um can I please share my screen for a minute and you I think it's not giving me permission to R it sure okay thanks M are we together so you're right here click on connect oh okay um oh so rather opened here and I was looking on the okay so this is where we are are we all together Terry that's where yes sir yes sir we're together how Terry Mara belus Emma yeah I'm good yep we're all good so let's continue the ma are we together okay let me stop Shing so where do I get the password from or that's where he that's what we want that's what we about to get so we getting mine is not con who who who Who's that you can share mine is not cor he said remote access to the server is not any um rofi yeah hope I pronounced it right is that is that are you the one sharing no I'm not Shar let's let's see your screen for a minute please you can share please Snappy let's be fast I would not like us to always um go beyond our PL time can sleepy paron you look tired no I I am tired but I'm not sleeping I'm fine I can still go for the next 3 4 hours actually I see my screen on is this your screen connect connect yeah download download remote desktop F click on download remote desktop file go to the download the F yeah double click on that click on connect okay in the in the Run book for Windows users I apologize because I don't think I have fa what you ped because I'm not using um windows so this is what is really not okay go back to this instance let me see go back to the running instance click on that instance 999 the ID there is Okay click on the security security here Security Group it let me see inbound rules go down let me see edit inbound rules is the problem yeah no that no but all traffic is fine the problem is the source source source so click on Source cck I any anywhere ipv4 please delete the roomle delete the rule add rule all traffic just go to all traffic custom TCP goes just all traffic so we don't stay drop shooting your stuff all traffic Source anywhere save go back and run that again go back to your RDP client and run that thing again it's open you can just go back to your downloads for that and you see it there okay run it again connect yeah it's working now this um wherever sharing can you share Leonard whose screen is this oh should we continue here no should that we those use oh that's a window okay is there somebody with Mark facing issues too Mar is Martha here Mara yes I'm here so um we created an Emi from an primary instance I will call it primary so we we we get it from a primary instance that's in us one in that primary instance we launch it with a kead and we created an Ami from a we created an Ami from a snapshot from that primary instance so if you were proceeding without keeper it told you that if you you would not be able to connect to that instance unless you know the password of that primary instance so we need the password to to be able to connect here so we let's go back to the primary instance so you're in Oregon you need to go back to how is it called Virginia nor Virginia click on nor Virginia select the instance which you created here it's okay because it's not looking is looking for that select the instance which you created you remember it um something Windows scroll down yeah yeah so you click on connect RDP C scroll down thec password so it says password and get password get the password so when you were creating this you you uploaded a keypad right you remember so we need that key private key file which I said you need to have it on your local if you don't have it then you can't go beyond this so upload that file into this into this so just click on upload private key file it should take you to your local machine to look for that file so what was your F name go your download it's right there yeah that's it ww I remember yeah open so now scroll to the end because the private key corresponds to what it has as this public key you can download this if you put in the wrong key penile it won't work so you can decrypt the password now we can copy that password into our clipboard we need to go back to our instance in US West 2 which is your I think you have it in the next tab just give me a minute give me a minute give me a minute yes go there oh all right just just go back to each two instance and do it all over again you can leave the North Virginia open and you just use another tab for for Oregon that should also work back are we together has somebody suceeded already Miriam no yes I'm connected you see like me I I'm wondering if uh I won't be able to do it because I did not uh put a keeper in my first instance you need a keeper I need a keeper so basically like that first instance is and that second instance they both because you're creating thei from it you need the kead or you created a new instance with a Keir and you try and see what happens but I think you will have a problem there is a way to to to to bypass this and and change this password but it's it's beyond the scope this 15 minutes session which we have that works that the L can you put put your password here yeah I did hold up did you just copy it what did you do how copy it and bring it here okay copy and bring okay refresh this please do this all over keep the the password in your clipboard and just rerun the remote deom client oh that's it yes click on yes you can always just read everything that comes on your screen read that's a good thing about this things at times they telling you what to do yeah I was said is pela on on the call read no she traveled oh dear sorry vacation read read read so now we are in our Windows ec2 instance U Prof yep um the Microsoft Remote Desktop that I downloaded what do I use it for because um it worked for me just downloading the file the was it yeah um are you the the person with the mark yeah with the mark yes okay if you do not have that CLI and you use use the shortcut fire it would work but what do I what do I do with it because just um it's just on my desktop I haven't configured so you you you needed to have the application for you to have for the shortcut to work because in the background it makes an API call to that client before it can open that dialog box for you oh okay okay so you can close it you can close it that's something that happens at the API level so okay you can close it but if you do not have the the application and you just go with the shortcut with the shortcut RDP F I don't think it will work all right thank you good so now we are in the Windows server in or Cloud Windows Server okay so you can you can just verify we remember the user script which we use on the primary server remember this is the server from our uh uh um our own a Ami right yeah do you follow Franchesca um say hold on I'm still I think I'm yet to get my I don't know where I placed my your F your your p f yes I can't I don't know where I placed my um so you need to look for it most likely it goes to download by default a to download that thing into your download folder so go there and look for it but what I'm saying is password can guys give me a minute let me explain something so this is what we are seeing on L screen is the windows instance he has a Windows machine so he might he might think that it's still the same but this is the windows Cloud Server okay and if you remember this server is Created from an Ami which depends on a snapshot from a primary instance in North Virginia so in that primary instance we installed no plus not uh p++ and we also created a folder and some information in there so to demonstrate that what we did in that primary instant was backed up in our snapshot and also now available in our new instance if you just go to search not note um notepad you should see our new application so cancel that you should see the application we we installed note part plus put plus let me see enter nine expect to see notepad Plus+ did somebody get the right result um what did he do yes BR I think I did I missed that what did he do what what did you do he just looking for not not plus Plus+ so if somebody get got it why are you not getting it Leonard can you close that screen and try again refresh not part note note note note not plus plus enter just click on enter okay this is weird did somebody get the the notepad++ yeah okay for you leard we come back to this go to your drive C and let's see if you have the folders difficult so you go to my computer I think you go to the folder I think that should take you to my computer U to the folder here yes I followed all the steps but I don't know why I cannot outp into the server we we'll get to you just give me a minute so for for troubleshooting let me let's just ens sure we see some stuff here then we can start troubleshooting so go to this PC we want to look for Drive C you remember the script was to create a folder in Drive C go to drive C click on Drive C cck on Drive C yes where is my pen double yeah local dexc local dexc now y right there double click on it so you remember this the folder we said it should create you remember J take model bash and in that folder there is a file with some information which we told the script to put there so you just click on it so we can see here at least we you have your folder you have your info.txt file and in that file you should have some information there that's all coming from the user script so you could also do this test your um your private time just do simulate the same thing without even using the user script you s you op you RP into your server create a folder upload whatever information music on your laptop into that um Server create an snapshot create an Ami from it in a different region and you should still see that backup thing in there same thing all right together I I got everything you described except the the note notepad++ so I'm curious why I got not Plus+ I think you have to type the notepad and Plus+ then it comes up well I mean I see everything else that um you're seeing now but the application isn't where it's supposed to be like uh if I do a search for it it doesn't show I'm sure it's there click on if you click on the Windows uh logo over there and type notepad you should see notepad++ okay let me try that whoever got it can you share and and um and we'll see what you get I got it can I share please if no one is sharing you done yeah you guys can check okay so should we stop the recording at this point or yes please I we need we need to troubleshoot and see why we're not getting there not I don't know if you guys copy your scripts all uh right and uh hopefully you did who's sharing and what's the issue it's me so let me go back give me a minute um um give me a minute do you guys prepare want the debuging sessions in the recording I think it makes a a little bit too long oh leave it for me to go all right go ahead so I think I copied so here I went to this is it [Music] what's going so let me just see how to the process see let me try to over the process back again in different pH then see what it gives me okay so I selected my stance and then connect oh sorry download no not this one not this okay so yeah get password get password get a password from here get the password right yes right back and this is it and then copy and then I need to go back to very go president Trump performance excuse me as having been very confident and sort of cerence meaning speaking sentences and it's talking s there I think so yeah by the moderators people debating always do but it became embarrassing Victor we can hear your TV thank [Music] you so this is it I can't paste it I don't know why uh can you do contr v v v there you go okay oh oh just try do how M yeah I got the same Works yes yes yes yes and then you should be good oh okay you may have to do it a couple of times before it takes before just wait wait please and proceed because this is where I'm stuck okay well are you still recording no he stopped recording it's still recording I thinking so it's it's going to take time to um to load but does anyone has their um not plus notepad++ I have it too not that's work he's asking me if I want to download it no it should be it should be downloaded that's what the script was meant to do where we yeah had so whoever whoever is sharing go to the search bar right under and type right under under search bar search bar down you you can cancel that cancel that screen yeah right click on that and then go down there yes notepad let's see what comes up notepad yeah notepad notepad++ give it some time it's a new machine so yeah she doesn't have it no just Ty just wait some time he has to download you will come wait some time don't rush don't rush right again type notepad Plus+ make sure you put Plus+ how many plus two two plus type it no that's a minus it looks like it didn't install for some people if you type just notepad it's supposed to show all your notepad I I did that it didn't show when I Plus+ it came up enter enter l no it's not there this is regular not I did my not and I have Plus+ I choose a Plus+ one right I wait for some time for it to download yeah that could be I did I had the note the normal notepad I wait for some time then I got the second notepad two plus because it's an Tolo remember that the the bootstrap says uh install notepad++ when Windows boots so which means that for some people it it may take a little bit longer to right down still booting still boting yeah okay I'm in the meantime franches can we check your C drive to see if the folder was created I don't think try not again let's see close close I still don't have it no wor someone said something that I think I think might be the reason because after after your computer starts that application needs to download so it might need some time I think that could be reason yeah I use some time and I got it did some people get it if you got it please just show your screen so others can see it then you can um um the process or rep the you wanted me to do something sorry yes I wanted you to check your C drive please if that folder is created but mine mine doesn't show the folder so you want me to share again yes please can I share and check your C drive okay someone is sharing let the person finish I think we are done we can I can stop the recording and we have I can stay here for the next 10 minutes to um okay help one or two people are blocked this is somebody with a not PL Plus+ this is what I expect to see good cck on can I share so where's the recording I stop it