deployment-config

This repository provides a deployment model for deploying the DeMAF.

Prerequisites

Installation of Docker and Docker Compose are required, see https://docs.docker.com/compose/install/

Deploy Locally

The DeMAF can be deployed on the local machine by executing

docker-compose pull && docker-compose up -d

Shutdown with

docker-compose down

Deploy on Kubernetes Cluster

The DeMAF has two available Kubernetes deployment configurations (in the k8s directory). One with memory and CPU bounds and one without.

First, take a look at the ConfigMap and configure the DeMAF deployment to your liking.

Then, to deploy use the following command:

kubectl apply -R -f ./<withBounds|noBounds>

To shut down the deployed system you can use this command:

kubectl delete -R -f ./<withBounds|noBounds>

For more details and issues regarding the current deployment, see the README.md in the k8s directory.

Using the DeMAF

There are currently three ways for using the DeMAF:

• WebUI
• DeMAF-shell
• REST API

WebUI

Once the DeMAF is deployed (wither using Docker Compose or Kubernetes) you can access it through the WebUI.

The WebUI is available at:

• http://localhost:80 when deployed locally using Docker Compose.
• https://try.demaf.de (or your own domain) when its deployed on a Kubernetes cluster.

DeMAF-shell

The DeMAF-shell is a service that provides a CLI for interacting with the DeMAF.

Clone it from https://github.com/UST-DeMAF/demaf-shell and start it as described.

REST API

The DeMAF provides a REST API for exposing its functionality.

After deploying you can access the OpenAPI definition under http://localhost:8080/swagger-ui/index.html#/

Transform a Deployment Model

The deployment model uses a volume to exchange data between the DeMAF services running in Docker containers and the local file system. It is currently configured to mount the directory under ./volume onto the Docker containers under /usr/share .

Input

If you want to transform a technology-specific deployment model located on your local file system, either

• copy the files to ./volume , or
• change the path to the local folder that should be mounted in the docker-compose.yml file in line 9.

You can then input the technology-specific deployment model by providing the location URL pointing to /usr/share/{yourDeploymentModel}, which is the path on the Docker container where the deployment model is mounted on.

Output

The DeMAF outputs the generated technology-agnostic deployment model under ./volume/tadms or the path that you configured.

Example

• Go to the volume folder
• Clone the Example Deployment Model: git clone git@github.com:Well5a/kube.git
• Go back to the base directory and start the DeMAF: docker-compose pull && docker-compose up -d
• Start the DeMAF Shell
• In the DeMAF Shell, execute the command transform --location file:/usr/share/kube/azure-start.sh --technology bash --commands ./azure-start.sh