Example Applications

You can run a wide variety of applications in your Oakestra cluster. Here, you’ll find some examples applications with their respective SLA template that you can use right away to test out Oakestra. You can also use these examples as a starting point to build your own applications.

header

I. Nginx Client-Server with Load Balancing

Minecraft Preview

To test out the balancing capabilities of Oakestra, you can deploy a simple Nginx server and a client that sends requests to a Round-Robin balanced semantic IP assigned to the server. This means that when scaling up your Nginx service, the client requests will automatically be balanced across the service’s instances without you having to change anything in application code.

Nginx Client-Server Example

Read more about how to create load-balanced microservices

II. Cloud/Edge Gaming: Minecraft

Minecraft Preview

Features:

  • 🎮 Play Minecraft from a browser, no client installation needed (thanks to WebMC)
  • 👭 Play multiplayer locally or remotely
  • 🖥️ Host your Minecraft server (Openhack) and proxy.
  • ⚙️ Scale your server instances to handle more users
  • 🛠️ Customize your deployment
Self-hosted Minecraft Demo

See GitHub repository for installation details.

III. Augmented Reality (AR)

Object Detection Preview

You can try out our custom distributed AR application, which is composed of three microservices.

1. Preprocessing: The preprocessing microservice collects the frames and adapts them for the model.

2. Object Detection: This service detects the bounding boxes inside the image. If Object Recognition is up and running, it forwards the frames there. Otherwise, it sends the bounding boxes back to the client.

3. Object Recognition: This service receives the frames from object detection. For each bounding box of type “Person” it detects the face features and sends them back to the client.

Camera
Pre-Processing
(
Pre)
Object Detection
(
Obj)
Object Recognition
(
Rec)
Person
Display
Person,

Male, 32yo
Person
Male, 32yo
Camera
Pre-Processing
(
Pre)
Object Detection
(
Obj)
Object Recognition
(
Rec)
Person
Display
Person,

Male, 32yo
Person
Male, 32yo

Augmented Reality Pipeline Setup

See GitHub repository for installation details.

Good to know

Read our research on how Augmented Reality and video analytics applications can be accelerated using Oakestra here.

IV. Unikraft Web Server

Similarly to the regular Nginx deployment in Oakestra, we can deploy nginx using Unikraft Unikernels. This will allow us to have a more lightweight and isolated version of nginx for the machines supporting unikernel virtualization.

Nginx Unikernel Deployment

Read more to learn how to create unikernel microservices.