nSights Talks

Amazon Appstream 2.0

Tutorial Highlights & Transcript

00:00 - Beginning of Video
Hi my name is Moeez ul Haq, I recently joined nClouds for the role of DevOps Engineer, it’s been two weeks. Today the topic I’m going to be presenting is AppStream 2.0.
00:22 - What is Appstream 2.0?
What is AppStream? Basically, we use AppStream to stream our applications to the user to know they can be accessible by the browser in the backend. Actually, AppStream uses EC2 fleets to host applications. And these applications can be accessed and used by multiple users at a single time. Also, we can integrate Google Suite and other applications to access AppStream.
00:48 - When to use Appstream 2.0
When do you use AppStream? Some of you might have a question in your mind that why use AppStream when you have EC2? Basically, we use AppStream, when we don’t want to give a user SSH access, or our user just wants to use the application for a certain period of time and move on. Another benefit of AppStream is that it has high availability, high performance, and most of the users that require AppStream are because AppStream has issues that are GPU intensive. And most people want applications to run on AppStream because they require a lot of GPU processing. That’s why we use AppStream.
01:31 - Appstream Use Case
How did I get to know about AWS AppStream? We have certain clients who just want to use the application and carry it out, and they want to render their images and save them under S3 and access it whenever they want. And they don’t want to get into EC2s and stuff like that. That is why I configured an AppStream for them.
01:59 - Appstream Demo Use Case
And this is the basic case of mine. Here’s an engineer. For example, take me, I invented an image folder and created a Linux-based image, and then I installed Klayout on it. And then there are a few commands, I think there are two commands that need to be said. One is to add the application into the AppStream so it knows that these are the applications which are going to be configured, when to configure the image and then run the command to actually create the image. These are EC2 fleets in which we are going to run the image that we created. Later on we have a stack, we can associate the EC2 fleet with a stack. The stack has the list of users which can access the EC2 fleet and so on.
02:47 - Demo of Appstream
Let me show you how it works. This is over AWS and we go to AppStream 2.0. Then we go to images. And we’ll find the image below over here. And I’ve already started Image Builder because this takes about 10 to 15 minutes to start up. Just to save time I’ve created – I’ll show you how to create an Image Builder if you want. Just as you can see, there are different types of EC2s, select the graphic, and g4 and these are really GPU intensive. Just to show you how to create an Image Builder, I’m just going to use the simple image and save it as testing. Set the type of EC2 I want. In my case, I’ll select Small and next. Okay, if you want your Image Builder to have internet access in which you would want because you want to download applications you’re going to select Enable internet access, select your VPC and give it a public subnet. Then next, and you’re going to launch Image Builder if you create Image Builder for you. It takes 10 to 15 minutes. I’ve already created it and I’ll show you how it works. Click connect, I selected Image Builder. It’s going to open up a browser and show me the interface at the next step.

Over here, to save time, I’ve also downloaded the application which is going to be Klayout. I’ve already downloaded it so we don’t waste any time. I’ll place my commands over here so it can add duplication to the Image Builder. I’ve already created an image, it just takes quite some time. I’ve created the image over here clearly. This is the version which I’m using. Then I have created a fleet with the image. I’ll go to the fleet sector, type the fleet that I want. Give my fleet name. Select the type of fleet instances I want and set up the minimum and maximum capacity of the fleet that I want and give it a VPC and a public subnet so it can be accessed. This also takes about 10-15 minutes to start up. This is a fleet that I have already created, and it’s running and it’s attached with a stack over here. And if I want to connect to the fleet, or the application, I’ll go over here and create a URL. This will generate a URL which will let me access the fleet.

And this is the application that I wanted in my fleet to be configured. I’ll go over here and if it opens an application, this application is quite GPU intensive that the client wants to be configured in AppStream.

I’ve also saved some files in case to show you how it actually works. Yeah, this is how it works and renders the image.

Jasmeet Singh

Moeez ul Haq

DevOps Support Engineer


Moeez is a DevOps Support Engineer at nClouds.