Developing with . NET and Azure Storage Services

1. Introduction: In this module, we're going to be looking at as your storage accounts. For this Monday, we're going to explore the Azure storage account offerings. We're going to look at Blob Storage Table Storage, Queue storage. And ultimately we're going to develop a solution that will interact with each of these components of Azure storage. So as usual, we're going to go ahead and create our storage or cones. Look at the different options and explore all the nuances, and explore how you can integrate these offerings into our own solutions. So stay tuned. This is going to be a fun module. 2. Understanding Azure Blob Storage: To kick things off, we're going to start by looking at the Azure Blob storage offerings. Now, Azure Blob Storage is a cloud-based service for storing unstructured data like text files, images, videos, etc. It is a scalable and secure and very cost-effective solution for storing large amounts of data. And it supports security features such as encryption, authentication, and authorization. There's built-in integration for Azure AD and your data will be encrypted both in motion and at rest. Data can be accessed via HTTP or HTTPS protocols anywhere in the world, and that makes it a suitable candidate for CDN things. So CBN, those letters mean Content Delivery Network. So because of this flexibility, Blob Storage is perfect for that. Integrating or building apps to support the Blob storage offering is very easy if you're using dotnet, Ruby, Java, Python, and several other frameworks, you can always go and look if the appropriate framework is there for your reference. But of course it's dotnet developers scores and it has us covered. Now it's commonly used for scenarios where we want to serve images or documents directly to the browser. We need to facilitate distributed access to a set of documents. If you want to stream content like video and audio, if you want to store log files and continuously right, to those log files. Or if we want to store data for backup, disaster recovery, and archiving. Know if you're into data science and big data, you'll realize that there is an offering called the Azure Data Lake Storage Gen2, which is Microsoft's enterprise big data analytics solution for the Cloud. This offers a hierarchical file system, as well as the advantages of Blob storage. And those advantages include the fact that it's low cost, it has tiered storage availability, strong consistency, and disaster recovery capabilities. Now we just mentioned storage tiers. So there are three kinds of storage tiers. The first one that we look at, this hot, hot storage tier is used for frequently accessed data analyst. You'll see that it's the most costly options. So if you know that you're always going to the Blob storage to access this file, maybe videos that are being streamed, audio that is being streamed. Then you would want it on the hot storage tier. Then we have the cool storage tier, which is used for less frequently accessed data. And of course it is a little less costly than the hot because the traffic is not there against the file or files. So whenever you try to access a file, it may take a little longer to load the file. But of course this is better for files that are to be stored long-term and not interacted with as much. Then we have the archive tier, and this is for a rarely accessed data. So this is for files that generally speaking, organizations need to keep files or keep records for up to seven years. Alright, so you'd want to archive files maybe after the third year you can put them into archives storage because you won't be accessing them as much. And if you do go to access, it is going to take much longer. It's going to have to rehydrate that file or that blob, then it would bring it back. So of course, that is used for extended for backups and it is the least costly option. Now a storage up colon has at least two main components that you see across all the other storage types. But since we're talking about blobs, I'll kind of tailor the inflammation relative to blobs. The first thing you have is the actual accounts. Now this provides a namespace in Azure for your data and every object I use stored in Azure Storage would have an address that includes this unique account name. Then we have the container and our container organizes all the blobs similar to how you would expect a file system like Windows or Linux to organize your files. You get that kind of folder structure hierarchy available to you in our container. Then the blob represents a physical file which is stored inside of the container. There's a diagram showing that hierarchy. Several types of blobs are. Blob is a blob, I mean, it's a file that is stored in this storage or cones, but then different blobs needs to be treated differently for different reasons. So the first one we have is a block blob, which is probably the most common one. We use that to store text or binary data. And it's made up of data blocks that can store up to 190.7 trilobites, not terror, but Tilly bytes. Then we have append blobs. Now, this does have blocks similar to the block blobs. Well, they're optimized for append operation. So this would be ideal for a scenario where we need to be logging data from a virtual machine or in your App Service. If you remember, when we're sitting on App Service logging, we could have chosen to log to a blob, right? So this is what we would use. It would provision an Append Blob for that situation. The next type would be page blobs. Page blobs store random access files up to 8 kb in size, and these are generally used as virtual hard drive files for our virtual machines in Azure. Next up, let's look at as your Table Storage. 3. Understanding Table and Queue Storage: Now as your table storage is a service that stores and non-relational or non-structured data, also referred to as SQL. And it is a Cloud service once again, that gives us key attributes storage in a schema-less design. Now you would've seen referenced as your table storage when we're looking at other Cosmos DB. And it is one of the APIs that Azure Cosmos DB support. So if you start off with Azure storage tables or Table Storage rather, and then you need to migrate for more storage and more redundancy, then you can move over to Cosmos DB. But this is an excellent place to start. If you're not sure of your data needs and you need to be able to adapt to new requirements regarding or detail, then this is an excellent solution to start off with because it's schema-less, so it will always be able to evolve with your needs. Now access is fast and cost-effective and it is generally lower than other SQL offerings for similar volumes of data. You can also consider that if you wanted to start off with our relational or Azure SQL Instance, and you know that you're not going to be storing that much data upfront that you might start with tables and then go up. But of course, all of those things will affect your decisions though. The next storage type that we have is as Queue storage. Now this is a service for storing large numbers of messages. These messages can be accessed anywhere through authenticated calls via HTTP or HTTPS. Essentially, when you set up your Queue storage, you're going to get a link that allows you to subscribe to the storage and read and write messages off the queue. Queue message can be up to 64 kb in size. And you can have as many messages as possible based on the capacity limit of your storage or cones. Know generally as your cues, storage or queues in general, are used to create a backlog of work items. The process asynchronously. This means that, let's say if you're building an application that needs to send an e-mail, do you want the application to pause while you connect to the e-mail server, figure out the contents of the e-mail dispatch, the email, wait to see if it was successful, the dispatched, and then the user can continue, maybe not. So IQ would allow you to take that information handed off and then move on with your application. And then you'd have another service that's probably going to take that message off the queue and send the email eventually so that we, the user experience is not affected. And that's just one example. Queues can be used in a number of ways, especially if you're talking about microservices. But we're going to go through some examples with this stuff. So I'll see you in the next lesson where we get started. 4. Create an Azure Storage Account: Now from our portal we're going to create a new storage account. So let us jump right into it. We're going to go over to our pills and choose Create a resource. And from here we're going to look for storage. So I can just type in storage, press Enter and okay, well, let's just use the autocomplete storage account. We want a storage of cones, so let's select that and create from here, let us fill out or form as usual. So as this storage is a Microsoft managed service that includes Blob Storage data Lake Gen2, azure Files. We didn't really talk about Azure Files, but I will explain that a bit. Queues and tables. No, it's important that you get these settings right because some things will be available based on your tier and based on the type of storage account that you provision. So the first thing as usual is to select the cart resource group. So let's go ahead and do that. And then for the storage name, I'm going to say as you are horse, I'm of course going to choose the best zone based on my solutions needs. And then we get to choose a performance. So do we want standard which is the recommended one, or do we want premium, which is four scenarios that need very low latency. So for this demo, we will keep standard. Of course, if you have other needs, you choose the one that is appropriate. Also for the redundancy. I'm just going to choose locally redundant. And redundancy, something that you should also always consider with every service once again, based on your needs. So I'll choose locally redundant storage. Go over to the next option. I'll leave everything that was ticked, I'll leave it ticked. And then here if we were doing some big data operations that we could choose a Data Lake Gen2 for hierarchical namespace is we're not doing any big data operations, so I'll leave that alone. And then when we scroll down, we can see here do we want the hot or the cool? So remember hotkeys for frequently accessed day-to-day usage scenarios. Maybe like CDN. Maybe that is where the entire organization is storing files and reading and writing. Versus the cool storage, which is probably where you would put files that are accessed once in awhile. Alright. Then we have Azure files. So as your files is a provision of the storage account that gives us a fully managed file share in the Cloud. And they're accessible through industry standards, Server Message Block or SMB protocols, network file system or NFS protocols, or even just a RESTful API. In more plain English. As your files would be like an in Cloud file share that you could set up for your organization to replace the local shear. So because we're not facilitating file shares, I will not allow or enable large file shares. And then we can go over to networking and everything there that is, the fault is fine. Data protection. So here we can set up options for how we deal with the data. I will leave the defaults ticked, but I encourage you to read through each one just to see what it's saying and change those values based on urine needs. Of course. We can also enable Version Tracking for our blobs. If you have several files, are several versions of a file, it will keep that log. And you can enable a change feed just in case you have some service that needs to subscribe to the blob state, okay, well, this change, I need to trigger an action. So those are things that you can consider if you want to automate some other workflows based on the blobs and activities done against your blobs. Now, when we move over to encryption, I did mention before that data is encrypted at rest and in transit. So they are just showing you the type of encryption. I'll leave the defaults there and then we can choose which services support customer managed keys, and I'll leave those as default. So once again, next up would be our tags. What? I'll just jump over to review. And once that review has been completed successfully, we can go ahead and create our storage account. Once the deployment is complete, we can click Go to resource. And once we're there, we will see our blob storage or storage account is created. And from here we can do several things. So when we come back, we'll explore what the different options are and hold can interact with the different components of our storage account. 5. Azure Storage Emulator: Now it is worth mentioning that we have access to a local emulator for Azure storage, and it's called azurite. And this is open source and cross-platform. So you have several installation options and you'll see here it provides cross-platform support on Windows, Linux, and macOS. And it allows us free local access to an environment for testing Blob Storage, Queue storage, and table storage. So all the activities that we're about to do involving all of these services we could actually do if we implement our provision, this local emulator for those services. Now setting up the services is relatively easy. If you're using Visual Studio 2022, Then you get it automatically. If you're using Visual Studio Code, you can get it for extensions. If you're using npm, you can run a command if you're using darker than you can pull that image and install. And you can always go directly to the azurite Git repo, repo and clone it down. And to run it, you have several ways to interact with azurite if you're using Visual Studio, visual Studio Code, etc. So you can go ahead and peruse this documentation if you just want a local environment to set up. And then you can go ahead and access them. Know, we're going to explore the Storage Explorer in a bit more detail as this module goes on. But I just wanted to point out that in order to access the local environment and simulate interactions with each of the different components of a storage account. 6. Storage Account Management: So from this dashboard you can see that several options that we may not have enabled or enabled and probably wanted disabled later on are available for change here. So if I wanted to change the default access tier from hot to cool, I could easily click it and then go ahead and change that. Certain things can't be changed or you can just jump up and change performance from standard or premium. But you could change some other options. So that's later on he decided or the organization decided that it needed to allow large files. And you'll see here that it supports fast storage of up to 100 kb. You could go ahead and do that. Could also change this stock quotes here, but we'll leave it hot for now. Later on, we can either remove the resource, our drop it down to a cool tier. Now back on our overview pane, we see that we have the option to open in an explorer. If we click Open Explorer, you'll see that it is something that you can download that will allow you to manage your files and your containers and everything from your computer. We can also use a storage browser. So we can click on storage browser and that will bring us to similar enough interface that shows us the number of containers, file shares, tables, and queues in the system. We can also go ahead and interrupt with those components. So I can go ahead and create a new container here. And I'll call this container. Is of course container. I can set the access level whether I want it to be private so nobody else should get to it or blob for read access for blobs only or container with anonymous read write access for continuous and blobs. Now leave this as private. And in advanced, I can set the encryption scope and immutability support. I'll just go ahead and click Create. Now once that blob has been created or that container other, I can go to the container and then I can add a directory or I can upload a file. Now remember that the container is where the files really go and it tries to give you that hierarchical view of files and quote unquote folders or directories view if you must. So I can add a directory and I call it new directory. You, is it coarse tree? There we go. Then inside of this makeshift directory, I can actually add up all the directories if I want or I can just upload. So when uploading I can choose to override if as already exist, I can drag and drop or browse for files. And then when I look at advanced, I can choose the type of blob that I'm uploading. So there's a block blob, Page Blob or Append Blob. I generally don't change this because I really only upload block blobs. Or you can choose the block size if you wish to access stereotypes. So if it's hot, cool or archive, and we can upload a folder, we can do several things. So I'm just going to do a simple upload. And I'm just uploading a random PDF that I found on my computer. Nothing too fancy. And there we have it. I just uploaded my first Blob file and it is hot access. Of course, I can go ahead and edit it. I can get a URL, I can generate a SAS token. And we're going to look at all of those and what they mean. And I can even delete. Now remember that that enabled the soft delete. If I delete this, I can actually go back and show active and deleted blobs and that will bring back anything that I have deleted. It will show me and it was showing me the amount of time that it's being retained for. And then guess what? I can actually ON delete it. If I right-click, I can click on Delete and it will be restored. So that's a nice little fail-safe for if we accidentally remove any files. No, recall that fast shares will require a bit more finished in setting up, but it is an excellent way to move your File Share from your local computer or your organizations fascia from the local servers and put them in the Cloud. And this is especially useful for countries that require distributed access to the same set of files. Alright, so if you want a private file share for our distributed setup with redundancies in place, this is an excellent option for you. Let's move on to cues. Though. I went to add a new queue, and I'm going to call this is a horse. Q. Let's go ahead and click Okay. Then you'll see off the body of the URL. So we are going to get around, when we're doing development, are going to get around to looking at the URL for our Blob container and the blobs and everything. But right now we're just doing an overview of the storage account. So from this queue URL, I can actually subscribe to this skew and I can put messages on the queue. I can read messages off the queue now and we say message is what exactly do we mean? When I select the queue, I can click Add message. So if I add a message and say this is a random test message. And I can set amount of time, it should expire. And so by seven days, if this message is not processed, then remove it. Of course, once again, your needs will differ, so you choose the one that is appropriate for what you need. You could also see that the message whenever expired, so once a message is equal to under Q, never delete it until it's processed. Alright? And then you would say encode, we would want to do that. I'll click. Okay. So when I click Okay, we now have a message sitting down on this queue. And once again, when we say messages, we're talking about data that needs to be processed. So this message, as you can see, is really just text. And I just accidentally added the same message twice. It's really just takes, alright, so that means that this could be a sentence. It could be some hoard. It could even be a JSON object, which is a very popular way of transporting data like bodies of data onto cues. So what could happen here is that you could send over a block of JSON, JSON object as sticks. And then you can use a third-party service or another service to actually process this JSON message, which will have all the information needed for an operation at hand. Several things and several ways you could use cues. Once again, the major use for queues is to offload processes that might cause delays in the actual applications runtime. So it's quicker to using the email example again, it's quicker to hand off that e-mails body, which could have been this JSON object with the sender to receiver email addresses and then the body of the message and everything else had it off to the queue and then continue with the application. And then in the background, eventually this message will be picked off the queue and send. When I say ventrally, I mean, it could happen. See this same time it could happen 20 emails later if you have hundreds of users sending e-mails in the system, you don't want hundreds of users and the one on waiting for the system. There are several ways what we're going to look at some examples in the rest of this course. So that's all we add messages to the queue. And of course we can clear the queue and we can also dq message. So I could click this message and I can say dq. Now, the next one would be tables. Table Storage, once again is a schema-less design. So I can click Add Table, give it a name. And wherever you see me not using dashes, because certain services have stricter naming conventions and others, so some of them will allow the hyphens, some of them will not. So you will figure it out as you go along. It will always let you know when you are violating the naming convention. But when I click Okay, that creates a new table. And then once again we get a URL for this table where we can subscribe to putting data in and reading from the table. And then if I click on the resource itself, we have a very similar layout to what we saw when we're using Azure Cosmos DB. We have partition keys, we have row keys, and it will have a timestamp. So basically when I click Add entity, I have the option to give a partition key and row key, and I can add other properties accordingly. Alright, so let's go ahead and add a new entity. Now let us say that I wanted to store some grocery items. And what I'm going to do here is I went to see that the partition key is the category of the product. So if I'm storing something that's dairy, and then the, let's say property number one is the price. And then I can choose the datatype. So price would be double. Let's say it costs $2.50. And then the next one would be the name of the item. I don't know why I put the press before the name, but let's just say that this is milk. And then insert. Now, I could have generated a new row key. I could have put one arc would have used a GUID. I intentionally left it blank, but I'm just showing you the points that we can enter our partition key and row key, but then we can enter as many other columns and column value peers are key value pair key attribute peers as we need and like we said, as your needs grow. So again, the data so I can add another entity that maybe later on I realized I need to store more properties than just price and name, right? So if this time it was say staple, and then let's give it a row key this time one. And then the price is ten. Let's just say the name is bread. And what else would we need to storable this? Maybe I want it to actually put in a category. Once again, that's a staple. Maybe expiration date. I'm just I'm just making up stuff right. Because once again, your needs will differ. Expiration date has the time, right? So you choose the appropriate data value and then I put in an expiration date and I'll just delete the last one. So let's say that this was a new look. As our needs increase, the number of columns we can put in will increase, r can increase with very little consequence. And that is where that no sequel aspect comes from. So it still looks relational, but we still have that advantage of that schema lists design. Now that completes our tour of our storage account. When we come back, we'll get into looking at how we can go about creating solutions that interact with these different components. 7. Storage Explorer: Now if you opt to use the Storage Explorer, you're going to get that Visual Studio code slash visuals, sorry, as your Data Studio kind of feel. So first of all, you may want to manage our cones and authenticate accordingly. And once you're authenticated, you can drill down into your storage or cones. So once you have sorted out your authentication, you'll be able to go ahead and access the different parts. So here's that storage account that we created earlier. And once again, we can drill down into it and we can look at different file shares, the cues, the tables, all of those resources that we had looked at earlier. You may even see some other ones that you didn't see in the browser Explorer. So you can go ahead and do that. And all of the options that you had in-browser are also available here. So you can actually manage your files, your blobs, your folders, everything right here. And this is even better for machine-to-machine management of the Blob Storage. So I think that getting the Storage Explorer is a very good idea. And it does make management of your storage of cones much easier than having to go to the browser every time. 8. Web Application Overview: Now we're going to develop a web app that's when you use the different components given to us by the storage or cones. So our web project idea has been to bear registration form for conference attendees. We're going to use asp.net, MVC. This time we have used the Razor pages, you've used blazer, we're going to use MVC. But once again, once it's dotnet Core, the concepts will be transferable. Know what we're going to do is use Azure Table Storage to store the register and click Submit. We're going to store that data in this table storage. We're going to allow them to upload an image. That image upload, we'll use the Azure Blob Storage. Okay, So usually you would store it locally, but we don't want it stored locally on the server. We're going to offload that as your Blob storage. We're going to use azure Queue storage because we want to send off an e-mail. So I used that example earlier in this segment of the course. We're sending an email is something that requires the up to freeze, get that done, and then move on. So we're going to do is just hand it off to the queue. And then we're going to create a console app that is going to simulate a client that would read the queue and then process the message or the email data from the queue for dispatch. So we're going to do all of those things coming up. 9. Create and Setup .NET Core Project: So using the same solution that we've been using for this entire course, I wanted to create a new project. Of course, this is optional. If you want to separate each one into its own solution, then you can go ahead. And once again, I'm using Visual Studio. You can also follow along if you're using Visual Studio Code. So go ahead and create a new project. And this time I'm going to use an MVC web app. So I'm just going to type in the letters MVC in the search, and then choose ASP Core Web up with Model View Controller. Next, I'm going to call this one MVC Storage Account Demo. Next, everything can remain as default. And then we create, if you're using the terminology for Visual Studio on Linux or Mac OS, you can go ahead and navigate to your desired directory. And then you can say dotnet, new MVC. And then you can output it to a directory with the name of the project. And then you can go ahead and hit Enter. So now that the project has been created, let us go ahead and add some of the libraries that we'll need to work with. And of course we know by know the different ways that we can add libraries. But I'm just going to use the terminals since the terminal will work for everyone, whether you're in Visual Studio or Visual Studio code. So I'm going to say dotnet add packages or add package rather. And then the first package we're adding is Azure Storage Blobs. So we can just hit Enter. Alright, and then once that is successfully inserted, I'll just clear the screen. And then this time we're adding a zero dot storage queues. And then finally, we're going to add a zero dot data dot tables. Alright, so dotnet add package as your data dot tables. Now that we have all of those setup, let us go ahead and add our storage connection string, storage or colon connection string. I'm just going to jump over to the app settings that Jason, let me hide it, the terminal so we can get some more real estate. And I'm just going to call this storage connection string. And then the value for our storage connection string is going to come from our portal. So let's jump over to our portal. And then we jumped down to the access keys bleed. And then we're going to select the connection string. So I'll just show Copy to Clipboard, jump back over to our code, and then we can paste that. In. There we go. So my clipboard was not prime just yet. So that is our connection string. And I'm going to create a new folder. And this new folder is going to store our data model for the storage tables. So I'll just call it data. In keeping with our already established naming convention, we're going to add a new class. And I'm just going to call this one attendee. Know, because we are going to be using Table Storage, we need to go ahead and implement the IT entity interface. So attendees going to inherit off the bat from i table entity. And this comes to us from our data tables class. And then by using Control dots, I can implement the interface. So this now gives us these specific columns, right? Remember that these columns were there before we started adding other attributes. Alright, so of course I can just clean this up and let they get B I get and the set, set. And remove all of these through new throw nuts implemented exceptions. Alright, so after we've cleaned that up, let's go ahead and add our own properties. So this is helping me out here. So I have name and I would probably want firstName separate from last name. And let's put in an e-mail address. And what else would you need a bolt on attendee? Let's work with this for now. So as our requirements, cms will mean putting more. But let's start off with this. 10. Add Table Storage Service: Now let's configure our Table Storage Service. So I'm going to create a new folder, and I'm going to call this one services. And in here I'm going to create a new class. And this class is going to be called Table Storage Service. And while I'm here, I'm just going to make a quick adjustment here. I think I'm just going to qualify this a bit and call it attendee entity. Just so it's not just that ten d. That's just my my little adjustments. If you feel the need to make it as well, then do so. I'm just renaming it here and renaming the file to attend the entity soul for our methods here, I'm going to have, and I'm just going to write the public. Sorry, I'm going to write all the method names first and then we fill out the blanks. So I went to have a task that is going to return attendee entity. And this task is going to be called get attendee. Alright? And it will take and ID string ID because I intend to use goods for the ID. Good values. So we're going to have good attendee get attendees, which will take an old parameter. And then this will return a list of attendee entities. And then two more methods, one to upsert and one to delete. So between these methods we have full crud functionality. I'm just going to extract this into an interface so that we have that abstraction and that inheritance. There we go. Now, we can start wiring this up. So I'm going to create a constructor CT OR and Tab twice, degenerate that constructor. And I'm going to inject my icon configuration object into this service so that we can get access to that configuration file and the values there in. Let me just rename that. I'm also going to have a private, let's say const string for our table name. And I'm just going to call this attendees. Alright, so that's what we're going to be calling the table in our table storage. And we have the configuration. Now I need a client. Alright, so I'm going to create method, I'll make it a private method down here. So private task that is going to return a table client object. Alright? And of course, as you go along, you add the missing using statements. We're going to say get table client. And this client is going to firstly define a client, I'm just calling client. This client is going to be equal to new table. Well, actually this is a service client. Apologize, I'm rename this to service client. So this is new table service client. There we go. Then we're going to pass in that configuration value for our connection string. And just in case you don't remember the key, you can always jump back over here. I don't remember how much it jumping over here to get it. And that is what we're going to be using for the configuration key value. Alright, so now that we have the service client, we're going to save our table client is going to be equal to service clients. Dots get table, table client via the table name. The table name is the same value that we just stored up top here. So we can do that. Then we're going to await or table clients to create if it doesn't exist. So look at this null. If the table doesn't exist when we are trying to communicate with our storage account, we have the option to just create if not exists. So I can just choose that. And of course I'll choose the asynchronous option and then we can return the table client. So if the table doesn't exist, then it will go ahead and create it and connect to it. And then our client now embodies that connection object. So let us figure out all we're doing this. Get attendee by ID. Now here we're going to have to get unstable client instance, and then we're going to go ahead and do I get Entity Sync? Know, it just looks like this. And of course if we're awaiting than the methyl and must be async. So var table client is equal to two. We'd get table client to choose the method we just made. And then the second line you notice has an error. So it's get entity a sync and all of those put the ID and we have an error. Why do we have an error? Because it requires a partition key. Alright? So we have to give a partition key and row key in order to get the entity. So attendee entity does not have any partition key value here, right? I mean, yes, we do have the built-in field for a partition key, but we don't have a dedicated field that will act as our partition key. So as part of good design, we would actually give it another property. And let us see that each attendee to our conference would be professional in a certain industry, for instance. So what we can do here is say industry, we'll just call it the name of the industry that you're in, whether it's IT, education, et cetera. And then industry here can be our partition key to partition all the attendees by the industry therein. It makes sense, right? So as you build out your application for your purpose, you always want to make sure that you're making the best decisions possible for your design industry here will be or partition keys. So back to our method and what I'll do is that we should pass in string industry and string ID. And then we will pass in the industry and the ID to go ahead and fetch that attendee. For the next method, which is get attendees. We of course start off with the table client and then we're going to do a little different bits of code here. So I'm going to say page table. I tend the entity object called attendee entities is going to be equal to the table Client dot query attendee entity. And then we can return attendee entities D2 list, and then that will satisfy that return type. Now let's look at our absurd. As usual, we start off with our fishing off or table client object. And if we await, the method must be a sink. And then we're just going to go ahead and await our table client dots. Let me try that again. So oh, I'm sorry. That's async. Await table client. There we go. Not upsert entity async. We're just going to go ahead and give it the attendee entity. So we already saw hold upsert works from when we were doing Azure Cosmos in case you're not so familiar with it. Absurd is basically a portmanteau or a combination of the words update and insert. So if the entity exists based on the information provided, especially with the key and partition key, then it will just replace the details. If it doesn't exist, then it will create the new record. Alright, so that's this whole upsert works for or delete attendee. Of course, we make it async. We go ahead and fetch our table client. And we're just going to go ahead and await table client Dot Delete a sink. And what I didn't do for this one is provide parameters that we need to provide to delete async. So for delete async, we need to pass in what would be the partition key and the row key values. And those are the values that we pass into our delete operation. Now, if you're ever in doubt as to what you should pass in, you can always look at overlords of the methods and that will help to clue you in. So go ahead and pass in the parameters and that should be delete entity a sink. I apologize not delete async, but delete entity async. Async could have deleted the table. Alright, so delete entity will take those two parameters. Now we have our service for our Table Storage created. We made a slight adjustment to our entity here. And of course we need to update our contract, which is the interface, because I changed the signature for this method. So I'll just jump over there and change the Delete to take the two parameters as it's expected to. And for it to get attendee, we also need to update it with this. Alright? So once the contract is happy, then the service is happy. And we have no no conflicts and record. And then as usual, we have to register this. So I'm just going to jump over to our program.cs and see a builder dot. Let me try that again. Builder dots services. Dots add scoped service for the IT was stored service implemented by Table Storage Service. And that's it. So when we come back, we're going to go ahead and create the controller and corresponding views that then allow us to facilitate that registration process. 11. Create Controllers and Views: Now let's create our controller. We're going directly to our controllers folder select controller, and we'll just do an MVC controller with readwrite actions. Click Add, and let us call it the attain. This attendee registration controller, of course, needs to be appended. And there we have it attendee, her dissertation on controller. Let us inject our IT service. So I Table Storage, I Table Storage service. Right? And we can go ahead and initialize this as a private read-only field. Alright? And now we can start with our actions. Now, generally speaking, you'd want to use view models to handle what's happening on the front and different from the backend. But once again, this is a quick demo, so forgive the lack of best practices and abstractions were just doing a simple demo to show the connectivity. Afterwards you can go all out with all of the best practices and separations of concerns for our index view, first thing that we'll want to do, of course, is get the data. So I want to say var data is equal to underscore table storage dot get attendees. Now, this is a synchronous I need to await. And then of course, convert this method into an asynchronous method as well. Then I can pass that data over to the view. Now let's generate the view. So right-click the word View, click Add view. And I'm just going to go to razor view. And I'm going to use a list template. And the modal class is going to be attendee entity. And I can leave everything else. I don't need to reference script libraries and then click Add. And in no time our index view has been generated. So from here, I won't bother to modify what we're looking at yet. What I will modify is this section where we pass over the primary keys so we can use the role row key. And of course we need the party chunky, so let me name these appropriately. So we need the id and the partition key, which is going to be called industries. I can say industry is equal to item dot industry comma and id is equal to rho KI. Alright, so that's what we're going to have to pass over to our edit, to our details, to our delete methods. Why? Because all of these need both the partition key and the id value, because we need to fetch in order to edit, we need to fetch in order to look at the details. And of course the delete, we also need that data. So that is how we're going to do the index. Now let's move on to the next one, which is the details. And the details would follow a similar operation. But remember that here we're taking the int id and we're taking the string industry value. And actually id is a string because we're going to be using a grid. And we need to make some adjustments here. So this needs to be an async task. Of course, I could have just used the IntelliSense to convert this. So let me just make the method Async to spirits that typing. And then we're not getting attendees, but we're getting our ten d, which is not going to expect us to pass over the industry value and the id value. So we pass those to over we return the view with the data. And then we can generate this data, this view. And we'll detail, so we'll use the detail template against the same entity and go ahead and add that. Right now we have our details view created. So let me close anything that we don't need at this point. For Create, we can leave that because we don't need any data upfront or Louis. So let us look at this. If we want them to provide the industry value, we could theoretically give them a drop-down lists are not theoretically based on your needs. You could give them a drop-down lists so that they can choose from a select curated list of industries that you would like to have as your partition keys. For this simple demo, we'll just go ahead and type it out. But of course, in future, in the future, you may want to have a curated list that you recognize as your party shall or potential partition key values, right? So let us leave the Create view alone. Well, let us generate this view. So I'm just going to right-click Add view, razor view. And then we'll use our create templates against the same entity. Go ahead and add. After that creators has been added, It's Notice it's not saying async, we didn't need anything asynchronous in the create, but we will need to do some asynchronous stuff inside of the post method. So in the post method, we're expected to get back on attendee. Entity object. There we go. Then what we're going to do is try. So before we even get to the inside of the try, before we even call the service, we need to put in some data. So here I wanted to say that the partition key, attendee entity partition key is going to be equal to and then that's the attendee entity. Dots. What did we agree on? I think it was industry. That's correct. And then I'm going to state that the row key, so we add the row key. There we go, is going to be equal to Boyd dots, new GUID. To string. Those two, we can go ahead and call our Table Storage Service dot upsert attendee and give it the attendee entity object. Because all the other fields would have been set through the filling out of that Create form. I have an error here because the conversion into async was not completed. So let's go ahead and meet. That's a task and that error goes away. So now we can move on to the edits. So the edit will have to do a fetch similar to the details. So we can just go ahead and do that. And of course the parameters need to be similar. So why are you type when you can copy and paste? And afterwards, oh, we need this to be asynchronous, so let's convert that. And then in the post operation, we're actually going to do something similar to what we did in the create. So we have to make sure that we reset a partition key and row key. And then we go ahead and absurd. Alright, so you'll see that creates and edits basically follow the same kind of procedure, but we have to be a bit more careful with OER. So I'm just going to change all this parameter and it will go through and make sure that everything will operate the way we expect. So once again, we're setting the industry. What are the industry is to be the partition key. Do I want to have a new row key though? I wouldn't want to have a new row key. So I'm going to remove that one because the row key is already coming with this entity and we can always make it hidden on the form, right? So I want the setting the row key again, I'll just be resetting the partition key just in case that value changed when it was edited. Now, I need to meet this, of course, is synchronous. There we go. We can move along. Now for the Delete, I'm going to forgo the need for the delete registration. I'm just going to delete confirmation at least for now. So I'm going to remove the getMethod and then we implement the post method. I've already converted it to an asynchronous method, and I placed in the string ID and industrial parameters much like we had for the edit. And what we'll do is, I'll wait, delete attendee after it receives the industry and ID values. So what we'll do in the index is re-factor this from a link to an actual form submission button. Just place the forum here. And what we have is a simple form tag with the ASB action as delete, so it will hit the post method. And then we have two inputs, both as hidden. And the value here will tie in. And let me make this correction. So one should be item dot roll key and the other one is item dot industry, that's value. And then the names of course correspond with the parameter names that we outlined in our axon parameters. And then we have an input that is a submit button, which just looks like a danger buttons. So I can comment on this link because we don't need it. So that is how we're going to handle the delete. Now I have one more thing that I want to point out, and this is a little bit of our refactor and it's kind of a criticism that I have for visual Studio and the way it refractors methods when it does them as a sink. But appending that async keyword to the end, I don't like it's purse. Also. Like you'd notice that all of these are async methods and I did not say get attendee async. So if you want to keep that naming convention with your methods, that's fine. It's not something that I necessarily practice beyond my personal preference. I've noticed that MVC does not do a good job handling navigation when async is added to the name of the axon and the view. So be on the safe side. My suggestion is if it got renamed to async, just remove that async naming. That's one. So do that for all the actions. You'll see that I already went through. And all of those that were edit async and index async and everything. It's renamed to just the original name without async appended. That's one to rename the views, right? So all of these views will actually try to go to the async version or they're named async. So you can just go ahead and remove that async keyword. I'm just clicking it and pressing F2 on my keyboard to then rename. And the action inside of the ones that are forms that it, it should just go to Edit, right? So this is just a quick refactor that I'm suggesting just to make sure that we don't have any problems. I've never really sat down and said, Let me try and make the async naming convention on work. I just prefer it without, so it's something I tried to avoid. Right. And are generated this one before. If you wanted to have the confirmation page, we could have kept the get generated the view accordingly. So this is the actual form that I just used on the index to make sure that we can just go ahead and delete to the post. The final activity is to go over to our layout and let us change our menu to have the option to go to the attendee registration controller, and it will go to the index view, and then you can give it a name. So I just remove the privacy link and replaced that with the attendee registration link. So when we come back, we can go ahead and test. 12. Testing table Storage CRUD: Alright, so I'm running the application and just in case you're wondering how to run. Once again, make sure that the new project is your startup project. And then you can just run without debugging if you want to debug it. And by all means, go ahead. So I'm running without debugging and I'm just going to go ahead and hit attendee registration. So we see that our index page loaded, we got no errors. So that means it's connected to our storage or cones. The table saw nothing came back and it's showing us blank. That's fine. Lets us create new. If I create a new test record, let's say test, test and test at ti.com and industries IT, and of course I won't provide a partition key, row key, and timestamps so that spare parts of the clean up that would need to perform on our UI. But that's fine. Let's hit Create. And here we see that it came back with the record. So the industry and I am partitioning key of the same values and we got a new good for or 0 key and a timestamp value. If I go to details, we see everything about that. So we know that the fetch one by ID slash partition key is working. If I go to Edit, let's try and change something here. So this is test 123, and I'm going to save, and that has edited that file. Alright. Now let's take note or something. If I go to Edit again, I changed the industry this time to ITU one-two-three or let's try something real education, right? So it wasn't, IT was education. And then I tried to save nodes that I get an entirely new record. It's the same rule key, but at different partition key. So you need to be very careful with that. Because we did specify that the table storage is using a composite key. It's looking at both partition key and row key. So apart, Deshawn could have several records with the same sorry, we can have several partitions with the records with the same rural key value. And that's what we're seeing here. This is a different partition and we are getting by the row key value here, right? So it just went ahead and created a brand new record. So that's something you need to be very careful with when managing that partition key value. And for that reason, you could probably just use the row key as both parties, Sean Andrew key, to make sure that it's the same record all the time. Now, I'm going to go ahead and delete, and I'm not bothering to fix up the aesthetics and anything. This is really just to have a demonstration with the Cloud integration, not necessarily to meet the stylish, right? So I'm just going to go ahead and test the delete and remove one of these records. And there we go. The delete worked. So right there we just validated that our Table Storage crowd works. Know what you can do for further consideration now is go ahead and clean up things like the create and edit forms so you don't need timestamp row key and partition key displayed. We can remove those from the Create, and we can remove those from our timestamp. So let's go ahead and clean that up. Then for our index listing, we probably don't need to see those things either. So I don't need to see the partition key, row, key, probably row Keegan, stay. But then since it's more of an idea, I'd probably want to move it to the front. And even then, do you really want to see that grid? Now? Let's remove that as well. So none of those three items exists here as well. So we just keep it nice and clean. And right there we just implemented credit for Table Storage. Now let us see that our requirements have shifted. And no, we need to add some some form of file upload to this form because the attendees need to upload an avatar or some image that we'd need to store securely and then retrieve for display. So now we're going to look at how we can leverage Blob storage.com or something like that. 13. Add Blob Upload Service: Alright, so our requirements have changed. I know we need to store an uploaded image for the user, but we don't want to store it on our servers for several reasons. So what we're going to do is offload that responsibility to an Azure Blob Storage. So I'm going to add a new field to this table entity and I'm going to call it image names. So what we're going to do is rename it to have the same value that the row key would have. And it will just store that image name here. And then that will serve as a point of reference going forward whenever we need to reference the blob. Now let us create a new service, and I've done that already. We're going to call it Blob Storage Service. So just right-click services, add new class and we're calling a Blob Storage Service. And in my Blob Storage Service, I've kind of written notes, some of the code already someone to walk you through so that you don't have to watch me type. Now, what I've done is firstly, create a constructor where I injected the configuration. So you can go ahead and do that. I also have a field called container name, which I'm calling attendee images. Now, the reason I'm using all lowercase is that and I alluded to the naming conventions earlier. With Blob of storage, you cannot use uppercase characters in your knee. So anything that you're naming, you have to name it lowercase, you might use some special characters. Footnote, I'm just keeping everything lowercase attendee images. And then I have my methyl and that is going to create a container clients. Now you can wrap it in that truck yet so that you can catch any exception is especially regarding the naming because what we're doing here is we're seeing try to create a container, Blob container client to be specific and pass in that connection string from configuration and that container name. Now, if your name doesn't match up, then you will definitely end up with an exception. So you can just do that try catch and put a breakpoint here just so you can see what the exception is, you'd probably get a 400 error at that point. No, we're going to go ahead and see a weight creative NOT exists. So when it tries to do this, it would hit the exception. And then if it gets this far, we want to return the container. So that's our first method in our new service. Now the next thing that we want to do is upload a blob. So we are going to have a method called upload blob. It's going to be an async task that returns a string and it's taking iPhone file as the parameter type. So when we upload using MVC or dotnet Core, in general, I form file is the datatype that receives that uploaded file. And then we're going to pass in an image name. The first thing I'm doing is establishing what the Blob name will be. And the image name here represents whatever name I want to store a file as the reason. It's a good idea to rename your file before uploading it is that if you have multiple users who are trying to upload files with the same names, you're going to end up with collision. So it's better to rename it to something that is unique to the record. Then we can go ahead and establish that blob names. So I'm taking the image name value, I can go to pass over a unique value. And then I'm appending that name to the original path of the file that was loaded. So path that gets Extension Form file dot Filename will look at the entire file name, get the dot, JPEG, PNG, whatever it is, and then we append it to our unique name. We get to our country and our client. Then I'm going to establish a memory stream object where I'm going to copy from the form file to the memory stream, reset the position to zero, and then we're going to use the client, upload the blob with the Blob name from the memory stream. Alright? And that default represents cancellation token. You could actually do without it. So since we're not using cancellation tokens that you remove that. And then I'm going to return the Blob name. Now client itself is actually client here is actually going to be a type of response blob, content info. So from that object you could actually look at the value and see different bits of information about a blob that was just uploaded. If you need more information than just the Blob name, feel free to use that. We could also use the container client to get more information afterwards as well. But for now I just wanted to store the name of the file. Now that we have uploaded or file, let us look at how we would retrieve or blob. Now retrieving the blob is going to be a special situation because what happens is that when we set up the Blob container, it's going to default to be private. And we have already established that we use URLs to access the different blobs once they're uploaded, we're going to look more at the URLs later on. But the fact is that if it's private, then you cannot just access it. And you don't want to just make it public because you might not want to ever find anybody to be able to access it. So when retrieving our blobs, we need to generate what we call a SAS tokens are shared access signature tokens and we're going to look at that in this method. So our new method here is get blob URL, which is going to return a string. So it's an async tasks that returns a string. And once again it takes the image name. So we initialize a container client, and then we initialize a Blob client as well relative to that image name. Now we have what we call the blob builder. And this is the builder used to generate the shared access signature token that I just alluded to. So what we're going to do here is blob SAS builder is equal to a new object, which is going to take the container name that we want to access, the Blob name that we'll want to access. Then we're going to set the expiration on this toolkit, which means that after this time passes, and I'm just setting this to 2 min. Alright? But after this time pass is that nobody should be able to access or use this link to access the resource anymore. I was sitting in the protocol to be HTTPS and the resource to be B. If you hover over resource, you will see that B is specific to blob. If it's a different type, you can see, you can see his said C for container and you have different options. But for now we're just doing Blob Storage. So we're setting B as the resource type. And then I'm going to set the permission for the blob SAS builder to be read. So this enum, you may need to include a missing using reference for this enum. But we have several options here. You can generate a SAS token that allows for delete, for ad, etc, are all permissions. Right now we just want read because we want to be able to look at the image when we retrieve it. And then we're going to return blob dot generate a SAS token with the builder options not to string. So this is going to go off, generate that SAS token, give us the full URL as well. And then it's going to return that entire URL to us. Know when we move on, we want to remove that blob whenever the record is deleted. So we're going to have one more method here that is going to remove blob. So public task AsyncTask removed blob and we're taking that image name once again, connecting with the container, creating a blob client. And then I'm just going to say, I'll wait, blob dot delete, if exists, a sink. This is just a feel safe. There is delete async, but delete if exists. Hills with if there's, if it doesn't exist, then there is no error, then we can delete all the snapshots. So remember that with Blob storage we have versioning and each time there might be a collision, it might create a new version and keep that history. We can choose to delete everything once we are removing the blob. So that's it for the coordinator for our Blob Storage Service. I'm just going to go ahead and extract this interface. And with that interface extracted, we jump over to our program.cs and register that service just the same way we did for the Table storage service. No, we can refactor our controllers and views. Let's start with our controllers or controller. I'm going to inject the eye Blob Storage Service and initialize it. We know how to do that already. And then for the index action, I'm going to add this for each loop, someone to say for each item in detail. Because remember this is returning the list of attendees. I'm going to say get the image name. So I attempt on image name is going to be equal to Blob storage service that get blob URL, item dot image name. Why are we replacing that value? Well, I'm not going to be storing the entire URL, I'm just storing the name, right? So that's unique name dot extension. So I'm passing that name, which the blob will have. Then I'm going to get the entire SAS URL, which I'm then going to replace this with for return to the view for display. Of course, in a more cleaner setting would've created that abstraction and Singapore view models so we could massage the data better and so on. But forgoing all of those best practices, this is clean example of how we can just go ahead and get the URL for this paper pulses. Next we have the details. So for details are doing something similar. We've got the record here. And then I'm just going to see a record dot image name is equal to get the blood. You are once again no format create post method. I've added a new parameter and that is the I-Form file parameter. So this I-Form file will capture the File Upload from the form which we will modify and a few. I've also refactor the code of it. So I started off by setting an ID to a new grid. Then I'm sitting the Row Key to be this ID. Now my unique name for naming my blob files is going to be the ID. It doesn't have to be the ID. The ID is going to be unique every time. I think that's an easy way to just look into Blob Storage and identify what's what whenever I need it. So I'm seeing if the form file dot length, meaning something was uploaded, right, is greater than zero, then I'm going to go ahead and set that image name to be equal to whatever this will return. So remember that our upload blob, if I go to the implementation sticking that and the image name, so that's going to be the ID. We're going to go ahead and fit that up, upload it. And then I'm just returning the result of the concatenation of the id and the extension. So that is what we're going to end up with as the image name. And then if nothing was uploaded, we can set up like a default image. Alright, and then we go ahead and absurd the attendee. Let's move down to our edit and edit. We're just doing that same if check, where if the form file is greater than zero, then we go ahead and upload our blob. I'm passing in that row key value and the form file which we added as a parameter as well. Now one thing in the view that I don't think I pointed out, if I go to the View here, makes sure that we remember that we had removed the input values for the row key, partition key, and the timestamp. But to retain the original row key for the record, make sure that you have that hidden ASP for Row Key in the form. Alright, so that's Edit. And then finally we go to the delete. Delete is kind of tricky now because we need the original image name in order to remove the blob. So what I had to do was Fitch, the attendee first. Then we can delete that and then we can remove the blob bypassing in that data, that image name, because we fetched it first, we have it in memory. We can remove it from the table storage and then use what we stored in memory, then parcel with that data for the deletion. Finally, let us look at the different changes we made to the views. So for the create view, I've added a new section that will allow us to upload the avatar. So I just created a new form group. We have the label here and I just typed in upload our tar. And then the input type is file, and the name here is formed file so that it corresponds with the parameter that we're looking for in the post. And that's it for our Create. And my copying and pasting skills got out of hand. So let me remove that validation stuff. And I need this same control inside of the edit. So I'll just add this form group to the edit form as well. So if nothing gets uploaded, that's fine because we keep the same blob and everything is good. However, if something gets uploaded, then we upload and it will do the replacement accordingly. Now in the details view, I've added another DT DD combination to this data list. So here I have Avatar and now I have that image tag, which is going to take it's sourced from model dot image name. And I'm just sitting our width of 150 and a height of 150. Then in the index, I've added another table header. And if you want to label it, you could say avatar or display picture, whatever it is. And then in the TR for the data, I've added a new td, which is a similar image tag. The source is going to be item dot image name, and the width is going to be 50 and the height is 50. Alright? Also have refactored the form of bits to put the HTML links inside of the forum so that they'll display left to right instead of that haphazard display that we had earlier. I know I said we didn't need to fix it, but I'm sorry, I couldn't continue with our fixing it. So I just put those two links inside of the form right next to the Delete submit button. And I've added the new HTML attributes for the class, BTN warning for edit and btn-primary for details. So with all of those changes made, I'm just going to save everything. And when we come back, we will do our tests. 14. Test Blob Service Features: One quick adjustment before we jump into testing. Remember that whenever we want to facilitate form or file uploads in a form, we need to include this attribute to the form. So I neglected to point it out in the previous lesson we're retrofitting or views. So Ink type is equal to multi-part slash form data. So make sure that that is present in the create form, as well as our Edit Form. I'm also going to add this hidden field to the Edit Form to make sure that we can retain the image name value. So remember that we need to retain the original image name. And if the user uploads and new image, then we will change that old. But we need to keep the original image names so that when we do an upsert, that value is not lost. Now let's jump over to our edit. Post action and I just went to adjustments are needed here. One, I want to move this if statement up because what we did was we absurd it before we even tried to check if we should update the file. And we should also update the image field. You mentioned name field, right? So we neglected that part in our own through, but we can correct it now. So we should check if the form file length is greater than zero. And all the things that form file might also be null. So I'm going to make this kind of no safe. Put that question mark there. So if form file has a value, we check the length if it is greater than zero. And of course we could also just check if it's not equal to null because obviously then it would have a length, but whichever check you feel more comfortable with, then we go ahead and say, get the image name to be equal to whatever was uploaded. If we upload something, then we're going to assume that we are new, we need to update that image name. And then we can go ahead and upsert. Now let's go ahead and test our data site, cleaned out all of the existing records because I was testing everything and I didn't create another record for conducting this test. So if you want to clean out all of your existing table storage records, you can always go back over to your storage browser or use the Storage Explorer that you may have downloaded. Go down to tables and you'll see the table that we had previously created and then you can delete whatever record you feel that you need to delete. Now, when I run, run this application and navigate to attend the registration and we see our table. Let me create a new record and I'm going to upload an avatar. And I'm going to choose an image here and give it FirstName, lastname, email address and industry and Create. And there we go. So now I'm seeing an image displayed here in Avatar. Look at that. And I'm seeing all the data that were used to know what is this image. If I click Details, I get a broken link. So clearly we need to do some work here. And what I'm going to do is inspect element. And there we go. That's the error. Hey, it's actually printing the words, model that image. So let's just jump over and fix that where this should actually be at sign. And that shouldn't be little model, that should be big M model to represent the data. So let me go ahead and fix that and rerun it. And when I reload that record here, is the image being displayed properly? No, I still have my Inspect Element open. And if you look at that URL, you will see that this is the URL to the actual blob. So let us review what the whole URL thing is a bolt. If I jump back over to the portal and refresh, we will see our record and or record has image name. That is the same value dot extension, of course, as the row key. That's fine. Once again, we renamed so that we can have uniqueness across every upload. Now, if I jump over to the blob containers, the first thing is that you will see that attendee images was created and it is created as a private blob or container, right? If I go into, attend the images, every block blob that is inside of this container is also going to be private. So I can go here and I can copy the URL. If I tried to go to that URL, I'm going to get this message that the specified resource does not exist. Now we know that that is not true because here's that URL to the Blob storage and then the container and then the name of the file and we just copied it so we know that the file exists. So because it's set to private in the access level, we cannot directly access it via the URL. Once again, based on the access level, we can enable or restrict wholesale access to all of the blobs in our container. Now, this one is private. That's why we generate that SAS URL. And then when I click generate SAS, you see that now we can set up all of those parameters and then we can generate the SAS token and URL. And then once that Tolkien and URL has been generated, if I copy this and then try to go to it in the browser again. Well, my browser is actually downloading the file. All right, so that's what will happen if I'm using, if you're using a certain browsers, in other browsers, it will actually open the file. But let's look at the URL for a bit of pasted it in Notepad Plus, Plus so that we can inspect it a bit more. No, we start off once again with the same parts, the storage of cones, the container, and then the filename. And then you'll notice that you have a query string. So we have Sb, that is the permission, that's it to our Windows. It start. Okay. That's the time. When does it end? That's the time. What is the protocol? Https, and then we have other parameters. The type is blob and the signature is that random degenerated string. And SV there represents a signed version, right? You wouldn't set that, that will be set while the signature is being generated. That is what this SAS token brings to the table where we get this special URL that allows access for as much time as we sit between the start and end time. And you see here, this was set for one day. This was just using the default options. But of course we can set a time span that is specific to the purpose that we need that token for. Now that you have some insight into what is happening behind the scenes when we're getting that URL. Let's jump over to Edit. And when we edit with fetch the record, let me try and change this file. And after clicking Save, I encountered an error. So of course I want to show you the IRS and then we're going to work through them together. So the error here that was caught after tried to upload blob was that the specified blob already exists. Alright? So of course we can't upload a blob that already has that name. So what I'm going to do is remove the blob if it already exists, and then we can go ahead and create. So for upload blob, we're going to jump over to upload blob and we're going to modify what is happening inside of this method. So as it stands, we're just going to go to the container and tried to upload. And if you hover over that, you'll see where it says that it creates a new block blob, which means it will obviously return an exception, as you've seen, where the blob already exists. So that's why I reading documentation is very important. To overwrite an existing block blob, get a blob client, and then upload. All right, we'll take the instructions so we can reuse this line where we got that blob client and replace this original upload line with a new one that says a wet blob dot upload async. And then I'm passing in, I'm using named parameters, so know exactly what we're passing in where the content is going to be the memory stream. And then we're going to overwrite true. So by setting override to true, it will overwrite any existing blobs. And without setting that it would be false. So it end up getting the same exception that we got if we neglect to put in that override parameter. Now there's another twist to this. And it is that what if I uploaded a J peg initially and then I change it to an PNG. At that point, even though the id value, the image name is the same, we are going to end up with two different files. And do we really need to store old files, right? So technically speaking, you could also run a delete before you do the load just to make sure that you don't end up with two different versions of what should be one person's avatar. So you can consider that as well. So to meet that happened in the edit action, I'm going to send over the original image name as a parameter to our upload blob method, which means now we will have to introduce a new parameter. So our new parameter is going to be string, which is not audible original Blob name. And I'm initializing it to not. And of course if we update the interface, then we must also updates the implementation. And we have the same parameter listed here. So now that we have this original Blob name value, I'm going to check if it exists. So I'm going to say, if not string dot is null or empty and pass in that value, then we want to await, remove blob and give it that original Blob name. We're going to remove the existing blob and then we can go ahead and upload. So you see there are many dynamics to distinct in our case because we might end up with different files and extensions and we can't guarantee that this is always going to overwrite the existing one. It is safer for us to remove the old one and then add the new one. Whereas in a situation where maybe the file name is always going to be consistent regardless, then you probably don't have to go through all of that. But once again, options. So now that we've made these adjustments, so let's go ahead and test our edits. Again. I've loaded up the form, I've already chosen my other file, and then I'll see. And there we go. Now I have a brand new Avatar. I've successfully changed the upload or the Blob image associated with that record. So that's nice and easy. Of course, it had some nuances. Know the very example that I gave where the file extension may be different and we will end up with multiple files for the same record I actually had that happened to me. So here's one version with JPEG and another version with PNG, the origin of files PNG, the new file was JP, and instead of replacing it just created a new one I need to dereference. So that is why you may want to remove before you upload to make sure you don't end up with multiple potential files. Even this one was uploaded without an extension earlier. Those are little things that you'll want to be careful, careful about, and pay attention to when handling blobs. Finally, I'm going to test that delete. So let us delete this record and we know that the Table storage required it was deleted. Let's jump back over here and at least one of these files should go. Okay, two of them went because I removed the original one. Then I also removed the one that was attached to the record. This one was attached to nothing so I can delete this one manually. That's fine. But at least now we know that we are successfully writing to our blob storage, reading from our blob storage using an essay, a stalk. And that gives us limited access just for the period. And we are able to remove and upload a new one, Change File associations and everything like that. Now when we come back, we're going to implement our queue service. And the scenarios that wants an attendee has registered. We want to send off an e-mail, but we don't want to connect directly to the email service. Instead, we want to add it to a queue so it can be consumed later on, unprocessed. 15. Add Queue Service: So now we're going to implement our queue mechanisms. So the first thing I'm going to do is create a new model. I'm doing modelling set of data because it's not something that we're going to store. But I do want a class so that we can have a strongly typed message accordingly. So I'm going to call this one email message. And e-mail message is going to have three properties. So the first one is going to be a string, which is the email address. Then we're going to have timestamps. So I'm just going to say date, time called this time stamp. And final thing is the actual message. So that's good for now. And then as usual, we're going to have to sit up on new service. So let's go ahead and add another cluster or services folder. This time, we're going to call it q service. And our queue service is going to have a constructor that's going to inject or our configuration similar to other services, we're going to also have a set queue name, which I'm calling attendee emails. Now I'm just going to have a public that's a task of a public task called send message. And send a message is going to take the email message object. So email message, and I'll just call it email message. And the thing is that I am hard-coding into this queue. But you may also want to keep this kind of flexible to other cues that you may need to interact with. So what you could do for this queue service is instead of setting that queue name, you could actually just say string name as a parameter here. So that you can actually set the queue name when you're passing over the message that you want. But then this is a specific one. So I'm just going to making everything specific to our email messaging operation. Now I'm going to initialize a queue client and I'm not going to bother creating a whole method for that this time because we only have one message, one method which is sent message. So I went to say var acute client is equal to a new queue client where we need the connection string. We get that from the configuration. We get the queue name based on this value. And then we set the queue client options to be message encoding the 64. So remember that we put off, turn this on and off. When we're creating the queue itself. Then as usual, we're going to see Q client or a wait queue clients dot create, if not, exists, a sink. So we go ahead and make sure that it exists before we tried to send a message. To send a message, we can quite simply say Q Client dot send message. And of course there is an async option, but let me go ahead and put in what the message should be. So notice these expects a string message text, but we have an object of email message. So let me go ahead and I'll wait first. Now how do I convert this into something that can be sent as a string? Well, we can serialize it to be JSON. So I can say var message is equal to JSON, convert that serialized object. And then we pass in our email message that will just convert our class into a JSON message or JSON string object. And then we can pass that string over as our message. We can extract our interface. That's the only method we need to extract the interface, go ahead and register it in the program.cs as usual, and we're adding it as sculpt just like the other services before it. And then we can modify or ridges are attend registration control. And after we have injected or new IQ service, IQ service, I like that name. Now we can modify the hotspots. So obviously creates would be a hotspot where after everything is done and we've saved, then we want to create our e-mail message. So I can say var email is equal to new e-mail message. And I'm going to set the e-mail address to be the attendee entity dot e-mail address that was entered. Right. We can set the time stamp be a datetime UTC know, and then we can set our message to be something like hello, FirstName, LastName, hello. And they were just concatenating. So we're using, well, interpolating. So we're using interpolation here. And then I brought the line with the backslash n, backslash r. Thank you for registering for this event. And I have a straight up. And then I brought the line again and then I said, your record has been saved for future reference. So now that I have this e-mail object, I can call on my queue service and C, send message and just give it to that e-mail objects. Of course, we know we can do similar things elsewhere in our application because if it is edited, then we could let them know that it hasn't been modified. So hello, firstName, lastName, your record was modified successfully and then we send that message. And then if it gets deleted, I can also do that because remember, we would've pre-loaded that data here. So after everything has been removed, we go ahead and create our e-mail message and I copied and pasted. So obviously, we would use data as the object here for email address, FirstName, LastName. Then the email message is modified to say it was removed successfully, then we send off. So that is all we would integrate our queue service into our operations. Now when we come back, we'll validate all of this. 16. Test Queue Service: So let us go ahead and test someone to create a new record and upload an image. And of course we're just using test, test, test and IoT industry created. And once we validate that, we have a record and our blob uploaded, we can jump over to our Storage Explorer. And I'm going to refresh all so that I can see if that new queue or tennis racket cues and say Refresh. And now I see attendee emails as my queue and jump in there. And I have several messages because I was testing, right? You can see here that we have several messages and I did a full crud operations to make sure that each one worked. So I have one here that shows when it was created your record hasn't been saved for future reference. That's good. And then we have another one here that says it was removed successfully. We have another one here that says it was okay, I didn't do any modifications, but you see here that all of these messages are actually queuing up. So let us imagine now that the email service was actually don't, right? So these should have been e-mails that are sent immediately. But our email service was dawn. That means that the user would have registered and we might have left The operation with the impression that he's or a distribution on field. Because the e-mail service was done the website field to connect and then throw an exception and probably gave him an error message. We don't want that to happen, right? Another scenario is that okay, maybe we have proper error handling in place. So even though the e-mail services donor we tried to connect, we didn't throw an exception, but then that e-mail is lost forever, right? So IQ is an excellent way to kind of hold our messages for our operations until we can actually process them later on. When we come back, now that we know that our q is working as we expect it to, we're going to build a little utility that will actually pick the messages from the queue. And once we get the messages from the queue, whatever operation on it is that you need to carry out, can be carried out. 17. Logs and Services Calls: Now while we're testing our app, I just wanted to take a few seconds to point out what is happening in the console, just so you get a feel of what is happening behind all of these service calls. So we've wired up our table service or Blob Storage Service and our Queue Storage Service. And if you look at the console or the logs being spit out by the app for each request, you'll see that we're just sending GET and post and different kinds of HTTP requests or HTTPS requests to a bunch of URLs. Really and truly, it's not much more complicated than that. So essentially, our service clients and all of those are really just wrappers for HTTP clients that are communicating via RESTful APIs with each of our services. And we did say at the start of the course that the RESTful APIs do allow us to interact with and manage or resources on here is a live example of holdout works. So when we're talking to our table service, when rp is loads, we tried to get the data. Alright, and later on you'll see that we interact with, okay, here's one where we get a tool for. Right. Here's the one that's I wanted where I deleted the record. So it deleted from the Blob. So if you look at the URL, you'll see that that's the blob service, right? And it deleted that Blob name. And that's a delete request. Alright, here's the output where it's updated something on the Q. Alright, so that's the URL to the queue service. And here's a post request where it created something under Q. And here's I get taken getting from the tables. So I'm just pointing out that everything is really just a RESTful API call happening in the background. 18. Setup Queue Message Reader: Now that we know that our service is working, let us build a service that will subscribe to this queue service and then consume the message. And then whatever it is that you need to do with it you can afterwards. But let us build a utility that will connect to and consume these messages. What does I'm going to build a console app. So I'm going to, in the solution, open up a new or add a new project. And this one is going to be a simple console app. So choose a console app template. And we're going to call this one console dot q consumer dots demo. And then we'll go ahead and hit Next. We'll use the latest.net seven available at the time of recording and Create. Now I need to add a package, and I need to add the same package that we had added earlier for Queue storage, which is zero dot storage queues. So I'm just going to quickly and I think by now we all know a whole to add or packages quickly. So I'm going to choose this route where I'm just going to create a new item group and then just paste the same package reference that was in the MVC app. And save that. And you get will go off in the background and resolve that dependency for me. Now let's write some code. The first thing that we would want to do, of course, is set up the or establish the connection to our queue storage, right? And that is done through our storage account connection string. Know, you have a few options when you're doing this, the easiest option would just be to create a variable here. I'm called the connection string and give it the value. But then for the obvious reason that you don't want to put your connection string directly in code and hard-coded layer that we're not going to take that approach. For this situation, we can use use a secrets, which means that we would have to install a few additional packages and store it in a configuration file and then reference it accordingly. And then by extension, we can also just set a global environment variable on our machine. And then of course, if we're going to deploy this application to any other machine, they will have to make sure environment variable also exists there someone to take that approach for this particular exercise. So to do that, we're going to use our terminal window. And whether you're using, if you're on a Windows machine, we're going to use the following command. And then I'm going to show you the corresponding one for Linux and macOS. If you're on a Windows machine, whether you're using Visual Studio or Visual Studio code, you can say set X, given name, I'm seeing as your storage connection string. And then we place the value here. So I'm just going to jump over to the web project where we had the connection string value from previous exercises. And I'm going to use that here as the value and set. Be sure to get the formatting right. You can see that I had some errors because I copied over and a comma and add extra quotation marks. So just make sure you get just the quotation marks and the value in-between. Now if you're using Linux or Mac OS, the command is basically identical. But instead of saying set X, you're going to see exports. And it's the same format to export the name of the variable and then the value. One downside to this method is that you may need to restart your computer afterwards so you can hit Pause, restart, and when you come back, we'll pick up where we left off. So now that we have set that environment variable, we can set the connection string variable here. So I'm going to say string connection string is equal to environment, get environment variable, and we find that by its name. Alright, then we're going to initialize a new queue clients. And that's going to be similar to what we've done before. Cute client Q is equal to nu q client, referencing the connection string and the name of that Q. I'm going to try and get messages from the queue. So the first thing I want to do is make sure that the QA exists, which is just the rudimentary, right? If I wait, Q dot Q exists. And then inside of this if statement, I'm going to go ahead and check whole many messages are there. So first off to get the property. So I say Q properties is equal to our weight q dot get properties a sink. And then once I have those properties, I can now check if what the approximate, sorry, message cones is. I'm going to say while properties that approximate message Cohen's d is greater than zero. Alright, so that's just a number that indicates how many messages are in the queue. And then I'm setting up a while loop to say, well, while there are messages in the queue, then we want to do something. So I'm going to see go ahead and fetch the messages. So I'm going to say string message is equal to and I'm going to await method that I'm about to create, retrieve. Next message is sink, or we don't necessarily have to follow that async naming convention. But I do want to retrieve, let me get my spelling right. And then I'm going to say console dot, write line, and print that message. And I'm just putting on the word received. Let us implement this method. So I'm just going to degenerate the method below. It's going to be an async method that returns a string. And inside of this method, I'm going to firstly retrieve a message. No, you can retrieve up to several messages at a time, right? So I wanted to say queue message, retrieved message, and this is an array. And then I went to await the Q dot receive messages async. And then we can take a cone. This cone is the max number of messages and the minimum is one and the max is 32. So I can actually say, give me up to 32 messages one at once. But for now I'm just going to do one. So we're going to process them one-by-one. And then I'm going to have to convert from base-64 into a byte array, and then from a byte array into a string. I'm going to explain all of this to you. In our Storage Explorer. When we're looking at our messages, we see the message sticks in plain text and that looks fine. However, when you double-click it, you'll see that it is showing you that it has decoded it as base-64. So it's showing me what's in the message. But really and truly if you were to look at it at rest, if anybody wants to get access to the queue and just try and see the messages are trying to intercept the messages rather than they would just see a base-64 encoded string that would not actually see the texts like you're seeing it. No. So for that reason, when we are processing these messages, they're actually coming over as base-64 encoded messages. And then I have to convert from base 64, right? So this is the base-64 encoded then string. So this is actually binary data, right? And then I can convert back to UTF from that beta RE. So that's whole, that whole flow has to work. If we're using base-64 encoding, then after I have retrieved the message from the queue, I would want to remove it. So I'm going to say a weight q dot delete message async. And we're going to remove using the retrieved message than the message ID and their retrieved message pop receipt. Alright, and if you hover over these properties, you'll see what each one of these means. So essentially after we've processed the message or we've retrieved the message and we can remove it from the queue. And then I'm going to return the message, which we can then process. I don't remember that. Yes, we're going to console that rat land, but this could easily be send the email or store in the database. Because if it's an asynchronous operation and it's something that needs to happen afterwards. Like we said before, especially in a microservices situation, this could have been the agents that node does the anchor leg off whatever process the skew is driving, right? So this is just a simple console app. In a real-life scenario, you'd probably use our work or service that's always on, that you'd have to manually run every time. The code would look pretty much the same way with our work or service. If it was as your function, which we'll look at later on in the course. Then we'll see how we can handle this kind of situation as well. So let us go ahead and test and I'm going to set some breakpoints at some key areas. This is just so that we can track exactly what is happening as the cord goes through. So I'm going to make sure that the console up is my default project and go ahead and run with debugging. And we hit the first breakpoints so we can rest assured that the Q does exist. And then I'm just going to jump down to the while loop. So it went ahead and fetched the properties. Let's look at that object and we see here that the approximate message cone is five. So because of that, we jump into our while loop. And then we'll go ahead and retrieve the next message. And I'll just press F5 so that it jumps down. Lets us assess. First we got one message, no, because this is messages, even though we specify one, it's still going to return an array or collection type. So really and truly, it's going to return an array with that number of messages. So we get an array in this case with one message, that's fine. Now when we look at this retrieved message object, we'll see that we have our body. We have a dequeue cones, meaning how many times was this message red? We have an expires on based on the different configurations that we would have set up and we have message texts, so we have all of those properties on this message object, queue message object. Now, notice that body is an encoded string. So this is what we're talking about. This makes no sense that any human being, unless you are walking computer. So we're going to do is convert that to a string and then convert it to a byte array. And then we're going to convert from the byte array into an actual string. Then we're going to delete the message, because now that we've retrieved the message, we no longer need it on the queue. So this is the message that's that JSON body that we've talked about or that we saw in the preview. So now we can proceed to delete it. And then the message gets returned. And I'm just going to press Continue and remove the breakpoints so that all of the messages will get printed out to the console. And here I have an exception. 19. Code Cleanup and Optimization: All right, welcome back guys. In this lesson we're just going to do some code optimizations and clean up. And the first optimization slash cleanup activity brings us right back to the console app that we just did. So what I've done is I've replaced the while loop with a for-loop. Since I'm getting a definite number of messages, it makes sense to just go through for a different number of times, right? So that will definitely help us to get over that index out of bounds error as we'll only go this far once we are within the cones. So that's our quick refactor. Of course, there are several ways that we could do this and that was just a lapse in judgment to why I ended up doing the while loop in the first place. Hey, we live, we learn. So that's our first correction. No, I want to turn my attention back to the web project and the fact that okay, we have the storage or colon string here. But then in our services like Blob Storage, Queue storage, and table storage, we are instantiating a client every single time. So we can probably refactor that's a little bit. So let us start off by getting a new package. And I'm just going to use the developer console here so that we can all be on the same page, whether you're using Visual Studio or Visual Studio Code. And in the web project, I'm going to dotnet add package. And the package is Microsoft dot extensions that as your. Now that will give us the ability to go over into our program.cs. And while it's installing, I start writing the code. So that gives me the ability to go over into the program.cs and see a builder dots services, and add as your services or your clients rather. Then I can initialize this with a builder and the lambda body and end it with a semicolon. So what are we doing inside of this? Add a zero glands. Well, I can now say builder. Dots, add clients. So you see here, no, I can add blob service client, I can add queue service plans, and I can add table service client, all of that right here. So let's just start off with add blob service client, and we have several overloads. The easiest overload to use is the one that requires a connection string. So I can jump over to the configuration and we just get this storage connection string. And we're going to go to builder dot configuration. I didn't get my spelling right. Configuration dots. Well, not that. Build our configuration. And then the key that we're looking for is our storage connection string. So now that I have that one place, I can reuse it elsewhere. So builder dot add blob, service client, and I'm passing in that connection string value. Alright, I can also build her daughter had to service clients and guess what? It takes the connection string and guess what? I can also add. If you said table, then you are absolutely correct. Now let's zone in on the queue service client a bit. If we go over to the queue service, remember that when we created a new client, we had some options. And that option section included the default encoding that we want. So back in the program.cs, we can actually extend this configuration by adding dot configuration options to the end of this line. And then we're going to create a little configuration body. I'm just using C as my lambda token. Then inside of that configuration section, I'm setting C dot message encoding is equal to the same base-64 encoding. Know that we have all of that done. We can inject into our existing services using these services instead of provisioning them on the fly. So what we can do is start with the Blob Storage client. And what we're going to do is use the injected service to get a container client. And we don't have to call this method every time. So I'm going to inject the blob service client into this constructor and initialize a local field. We know how to do that already. And with that injected, I can actually, instead of saying var container is equal to our weight, get Blob container client. I can actually replace this with a blob service client dot get container client. And this is not a sink. So that's why we're getting that red line. Alright? And I can go through and replace all instances of that manual method to just be using our client. And then if I want, I can erase this method. What I'll leave it there for posterity so that we can see what used to be. We have the table service clients which we inject and initialize. And then we can replace our our way to get table client. I can know just call under service client, get dot, get table client. And then we reference table name and just carton that's spilling. There we go. So now all of these references, and this is probably more similar to the Blob Storage Client initialization. And it is still the Queue storage or the client initialization. So that's fine. And of course this method is here for reference. I did go through and replace all of them, but I just wanted to point out that what we're not doing here is creating if not exists. So there is a possibility that in getting the client, it might return null because it doesn't exist. So another factor operation could have been that we just eliminate this and we still return the service client based on that call. Alright, so we could still retain the method call to this and return the client for each call, right? Or for each method. So it's up to you, Right? We could also have just defined a table client up top here, or a field up top. Then we initialize the field inside of our constructor, and then we just use the fields. So there are several ways to do it. Now on that note, I'm going to jump back over to the program.cs and we're going to look at another way that we could approach this. What we're doing here is injecting the client, the service client. And then because we're injecting the service plan to have to go and create a new instance of the actual client that will connect us to the resource. So what if we wanted to just registered the direct client to the resource? So we would have to take a slightly different approach with what we're just doing our own here. And I'll just do it in another ad as your clients section. Just so that we can have some deal initial. Alright? So let us say that I wanted the queue client directly, not the Queue Service plan, not the service name, but the actual client. I would actually use the builder dot add client. Then this allows me to specify the data types. So I want the queue, client and Q options. And then I'm going to initialize a delegate using placeholders for the expected parameters for this method, if you hover over the AD plant, you see what the parameters are either represent. And then we're going to use a lambda expression for our object body. And then inside of this method were expected to return a new instance of Q client. So that means if I jump over to this queue service where we would have been initializing our clients. I could actually use this code right here and return a new client which is using the configuration or the storage connection string. I could put the queue name in my configuration as well. So I don't have to have it hard-coded in the file. So let's do that. In fact, I'm going to create a whole new section in the configuration and I'm going to call it as your storage. And I'm going to put the connection string there. And I'm going to just take off the word storage is kind of redundant. So now the key and I'm doing the same, the same time because I don't want to forget. So now the key to the configuration for the connection string would be Azure Storage colon connection string, right? Because now it's in that subsection. And I'm doing all of this, no. So that put a comma there. So now I can put the queue name or several queue names, whatever Q names and will need for my application. I can put them here. So the queue name is going to be, don't tell me, there we go. Me too to it. There we go. So the queue name is going to be attendee emails hour and I'll just use the Blob Storage. I just set the keys from null and table storage. Because this technique can be used for all three storage services. So back in the program.cs, I'm going to make a quick adjustment because I'm going to need the configuration values here. So I'm going to change this from Builder to just say B. And then queue name is going to be changed the builder dot configuration. And then we can reference that as your storage section and that the queue name key. Alright? And of course we need to return. So this needs to be returned UQ clients with all of those values. Now that we have this Q clients registered in the queue service, I can know inject that Q client directly. I don't need this direct reference to the queue name. And I can use this cute client to do all the communications accordingly. So I don't need to do that in initialization inside off the send message or inside of this service at all? No, I can do something similar for the Table Storage. I can just take this table name and we can put it inside of our app settings json. And then in our program.cs, I'm just going to copy and paste the queue client in the same section. And we're going to add a client for table clients with table client options. And what we're doing here is returning a new instance of table client with a storage connection string and our Azure Storage Table name, key or Table Storage. Table, Storage table name, whatever it is that you call it table storage. Alright? So now I can inject this table client directly into the Table storage service. So instead of injecting the service client, well, let's just do both. So table client and I'm going to initialize this. So I'm leaving all of this code here so you have future reference, right? It's not necessarily that you need all of this and I think we know that by now. But I don't have to initialize a table client every time I can just use my injected service, much like what we needed to do with the Q clients. So now I can just do table client right through this code base. So these are just some optimizations that you can make to your code to make sure that, oh, well, I don't have to change this one. So this method would now be archived, so to speak, right? So these are some optimizations and cleanup activities that you can due to remove some of these hard-coded references to the configuration. And then of course we discussed the fact that you probably don't want to store all of these keys inside of the app settings. So you could actually use secrets. And to use secrets, you just right-click the CSP image file and you go down to manage users secrets. And then you can add that section there. So it is not one to get checked into source control for one, and it's not immediately visible to prying eyes, but it is still accessible through or configuration in our startup. No one reason I'm not configuring a blob client while I'm closing. I'm not going to configure our Blob client. Because remember that the Blob client gets created directly against the Blob based on the knee. So while this is a bit more efficient to just go ahead and inject the service lanes. We still need to create a client on the fly based on the blob that is about to be processed. Because once again, we don't know which blob is going to be processed in order to start up with that. So you can go ahead and validate that is incentives and these new registrations work. And I'm just going to comment on one of them. And that's it for this activity. Let's close out this module. 20. Conclusion: So we've hit another mask, one and we have completed this module on as your storage. For the duration of this module, we reviewed how to create and manage a storage account. This involved us interacting with Blob Table and Queue Storage Services, all part of this storage account offering from Microsoft Azure. We looked at how we can use the Azure portal or the storage browser that we could download to your computer to interact with the different services. Once again, Blob Table, Queue storage. So we saw that we could manage the contents of those, whether from the Azure portal or locally using the storage browser. We also looked at how we can develop a dotnet Core solution to use the storage account. So right here we looked at storing and retrieving the connection string to the storage or cones whole weekend provision are and write code to interact with the different services and the nuance is in-between. And ultimately we can all talk it up to the fact that each one of these services is being managed through RESTful API calls and encode. All we're doing is provisioning HTTP clients to interact with the underlying services. We also topped it off by looking at how we could facilitate asynchronous communication using queues. And this had us implement something of a Pub Sub pattern where we published a message to the queue. And then we had another app that was subscribed to the queue that would pick the message off, process it, and move on. So that's an example of the pops-up pattern and how we can implement asynchronous communication. I thank you for joining me in this module. I'll see you in the next lesson.