Then, you have to set the fleet capacity and policy at which they need to scale; both are really important since if they are configured wrongly, you can incur a service outage or a massive bill.<\/p>\n\n\n\n
The next step will be to create and assign an IAM Role<\/strong> so that the users connecting to the instances will have a predetermined set of permissions (we\u2019ll talk later about the security). <\/p>\n\n\n\n After setting the IAM Role, the main configuration of the fleet is done. At this point, we just need an image and the network configuration.<\/p>\n\n\n\n The image can be selected from the predetermined list of images that AWS provides, or we can customize our own, and that\u2019s the main focus of Amazon AppStream 2.0: configuring an image and streaming applications from it.<\/strong><\/p>\n\n\n\n The last step is the network.<\/strong> <\/p>\n\n\n\n Since the underlying infrastructure is an EC2 with an AutoScalingGroup, you will need to assign Amazon AppStream a VPC, 3 Subnets (they don\u2019t need to be dedicated to Amazon AppStream), and a security group. <\/p>\n\n\n\n You can implement it with AWS\u2019 Directory Service if you find it easier to manage the permissions from there.<\/p>\n\n\n\n Once you have configured all these parameters, Amazon AppStream\u2019s Fleet will be ready to be used\u2026 but what if you want to create a specific image with a specific program?<\/p>\n\n\n\n First, you will need to launch an Image Builder instance. Its configuration is a smaller version of the whole Fleet, so no worries! <\/p>\n\n\n\n Once you have deployed the Image Builder, you\u2019ll need to connect to the instance and then personalize your image. At this point, you\u2019ll have an image with your preferred applications configured on it.<\/p>\n\n\n\n You can change the default Fleet image as you like since the changes don’t disrupt the users who are connected to an active session; only the new connections established after the change will have the updated image.<\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n Not everything is always made up of only good stuff, even Amazon AppStream 2.0! <\/p>\n\n\n\n These are a couple of aspects that can make somebody not choose this service.<\/p>\n\n\n\n One question that comes to mind is \u201cWhy should I use Amazon AppStream instead of AWS Workspaces or any other service provider\u2019s alternatives?\u201d<\/p>\n\n\n\n On the surface, they might seem pretty similar, but after a deeper look, you can clearly identify the differences and the use cases for each.<\/p>\n\n\n\n Amazon AppStream lets you stream an application within an enclosed environment without the possibility of tampering with the underlying OS, even without any ulterior steps. <\/p>\n\n\n\n On the other hand, AWS Workspaces streams a whole desktop environment, and as such, you can have all the default applications. This offers more freedom, but it\u2019s less secure without more complex configurations.<\/p>\n\n\n\n If you need different applications or if you need to modify, for example, some registry keys in Windows, you will need to work with Workspaces. <\/p>\n\n\n\n This is due to the inability to modify anything once the image has been completed on Amazon AppStream 2.0<\/p>\n\n\n\n The way you connect to Amazon AppStream is more streamlined and direct since you can do that with a single URL. On the other hand, with Workspaces, you will need to connect through an RDP connection.<\/p>\n\n\n\n To summarize, Amazon AppStream solves a more specific problem; as the name suggests, streaming specific applications and custom ones, not whole desktops environments (even though it can do that too, but not as well as competitors).<\/p>\n\n\n\n <\/p>\n\n\n\n Why do we recommend using AppStream when working with CAD softwares?<\/p>\n\n\n\n Since CAD applications are among the most GPU-heavy, and the average price per GB of GPU is relatively high, they lend themselves extremely well to the use of Amazon Appstream 2.0 since it allows you to select the most cost-optimized instance type and size for that kind of software needs. This makes the whole solution cheaper than buying directly the hardware needed to run some of the most demanding computing tasks in the IT world. As mentioned above, it\u2019s quite useless to configure and reserve a whole desktop instance if you only use an application. Overall Amazon AppStream is a solid option that can be considered when approaching the need to distribute applications easily, scalably, securely, and affordably, all while respecting an End-user environment.<\/p>\n\n\n\n I found its use relatively simple; it doesn’t require much effort in setting up the environment and the application, and managing the images and fleets is streamlined and well-documented.<\/p>\n\n\n\n There are definitely some alternatives that are similar to Amazon AppStream, like Citrix, VMWare, and Microsoft. Still, they all have their preferred use case, so I recommend carefully evaluating each provider\u2019s pros and cons!<\/p>\n\n\n\n If you are interested in deepening this aspect, feel free to leave a comment below! We’ll be happy to discuss this further in our following articles. <\/p>\n\n\n\n See you again in 14 days on Proud2beCloud! <\/p>\n\n\n\nImage Builder<\/h4>\n\n\n\n
![\"Amazon](\"https:\/\/blog.besharp.it\/wp-content\/uploads\/2023\/07\/image1-1.png\")
<\/figure>\n\n\n\nAmazon AppStream: the good stuff<\/h2>\n\n\n\n
\n
\n
\n
\n
\n
\n
The bad stuff<\/h3>\n\n\n\n
\n
\n
\n
\n
What to keep in mind before implementing Amazon AppStream<\/h3>\n\n\n\n
\n
Amazon AppStream 2.0 VS Workspaces<\/h2>\n\n\n\n
Real-World use case: CAD<\/h3>\n\n\n\n
Well, you can select the most cost-optimized instance type and size for your software needs, making it cheaper than buying directly the hardware needed to run some of the most demanding ones.<\/p>\n\n\n\n
You can seamlessly save your outputs inside an Amazon S3 bucket, where, if configured correctly, you can use a VPC endpoint and move the heavy CAD files only inside the AWS region, reducing costs even more!<\/p>\n\n\n\nWrap Up<\/h2>\n\n\n\n
\n\n\n\nAbout Proud2beCloud<\/h4>\n\n\n\n