elastic-AI.runtime.cloud

Table of Contents

• Setup
• Tests
  • Test execution
• Runtime
• Applications
  • Exit Codes
• Monitor
• Distribution
  • Build Docker Container
  • Run Docker Container

This repository aims to provide the cloud backend for the Elastic-AI implementation. It uses MQTT as the backend for the pub-sub-based messaging protocol and is build to be used in combination with the Elastic-AI Node v5.

Setup

For compiling the project, the required tools are:

• Java Development Kit (JDK) version 22

• Docker (Container Runtime)

💡	You can check if your local machine satisfies the required dependencies by executing test script in the root folder.

We recommend using pre-commit to check that the commit messages align with the conventional commit message standard enforced by the merge workflows!

💡	Setup Guide

❗

HOST_IP The elastic-AI.runtime.cloud:monitor target requires the Host-IP to be addressed. The application will retrieve this information from a system environment variable called HOST_IP. Therefore, it is recommended that you execute # MacOS export HOST_IP=$(ipconfig getifaddr en0) # Linux export HOST_IP=$(hostname -I) before starting the monitor (with docker or locally).

Tests

Unit tests and integration tests can be executed independently. Both use jacoco to finalize the tests with a report that shows the test results and the code coverage. The reports can be found in the location build/reports/ relative to the corresponding build file.

Test execution

Command	Task
./gradlew test	Run all unit tests
./gradlew :subproject:test	Run unit tests from specific subproject (i.g. gradle :runtime:test)
./gradlew integrationTest	Run all Integration test
./gradlew :subproject:integrationTest	Run integration tests from specific subproject (i.g. gradle: runtime:integrationTest)

Runtime

The runtime is meant to provide the necessary functions to implement a backend for the elastic-ai ecosystem. It provides the functionality to operate the ecosystem, like the implementation of the Twin concept or the MQTT Client implementation together with the necessary functions to handle the interactions with the MQTT Broker.

Applications

The Elastic-AI ecosystem provides two applications.

• Monitor (General purpose to monitor all clients)

• BallChallenge (Specific implementation for the Ball Challenge example)

To start a application locally via Gradle run

./gradlew :applications:<target>:run

Exit Codes

Exit Code	Description
0	No error
10	Argument Parser failed

Monitor

The monitor is used to provide an external interface for the user to interact with the Elastic-AI ecosystem. This interface is provided via a SpringBoot web application, which can be accessed with your default browser like

• Google Chrome

• Firefox

• Safari

• …

for example. To start the monitor locally via Gradle with the default parameter, execute

./gradlew :monitor:run

if you want to change the parameter for the broker ip and port you can pass these arguments by appending --args"-b <broker_ip> -p <broker_port>" to the run command.

The monitor can then be accessed from your local machine at http://localhost:80.

🔥	If you want to access the monitor from external devices, you have redirected your local endpoint to a port accessible through your firewall!

Distribution

Build Docker Container

A docker container for a subproject can be created with:

./gradlew  :<subproject>:jibDockerBuild

This container can then be used in a docker-compose file or started manually.

Run Docker Container

A Container should be run with:

docker run \
    --rm \ #(1)
    --network=runtime-network \ #(2)
    --publish 8081:80 \ #(3)
    --name <subproject> \ #(4)
    <subproject:tag>

1. Remove the container after shutdown

2. Required to communicate with the docker container running the MQTT broker

3. Port mapping for the webserver port, which allows the webinterface to be accessible from your browser (host:container)

4. Specify the name of the container

❗	Monitor The Monitor requires the additional flag --env "$HOST_IP". If the container is executed via docker-compose, the value needs to be set as an environmental variable.