Awesome

copper-modular-demo

A demo project that shows how to build a modular application with Java 11 and COPPER 5.0. The Gradle script allows creating a custom runtime image of this application.

The code also illustrates how COPPER 5.0 workflows support Java 10 features such as local-variable type inference (look for the var keyword).

We tried to keep the application as simple as possible: no Spring, no persistent workflows, no SOAP services. Have a look at copper-starter for an application using these features.

So, what is our application good for? It creates project teams consisting of a leader and a few members. To this end, it uses a recruiting service, which takes as input a list of constraints and provides a person that satisfies them.

A person is characterized by name, gender, and location. To keep things simple, we consider that each person offered by the recruiting service is qualified to be part of the team, so we only need to impose constraints on gender and location. For example, we can ask the service to find us a woman from Australia. It's not mandatory to specify all constraints. We could also ask for any person from Australia, any woman, or simply any person.

We use uinames.com to emulate the recruiting service.

For each team to be created we specify its size and the gender of its leader. Therefore, the workflow data contains two fields: int teamSizeand boolean femaleLeader.

The application implements the following workflow:

• appoint a team leader with the gender implied by femaleLeader
• recruit (in parallel) teamSize members from the same region as the leader
• display the team

A COPPER workflow will be created and executed for each team to be created. The number of workflows (teams) to be created can be configured in application.properties.

The uinames.com server cannot handle a large number of requests simultaneously. Under heavy load, it sends an error response or sometimes even no response at all. For our demo application this is actually a good thing, because it allows us to test error and timeout scenarios.

The parameter delayMillis in application.properties allows throttling the recruiting requests. In this way, you can control the error frequency. But please do not abuse the uinames.com server by running the application with low values of delayMillis for long periods!

Quick start

From the releases page download the archived custom runtime image for your operating system. Unpack the archive, go to the copper-modular-demo-image/bin directory and start the copper-modular-demo script.

The program will display the teams it creates. You can change the runtime parameters by editing the bin/application.properties file.

Click on the image below to see a sample run of the application.

<a href="https://github.com/copper-engine/copper-modular-demo/raw/master/doc/copper-modular.gif"><img src="https://github.com/copper-engine/copper-modular-demo/raw/master/doc/copper-modular-thumb.png"></a>

As mentioned before, some of the calls to uinames.com may fail. As a consequence, the program creates fewer teams and/or teams with fewer members than requested. By default, the application doesn't print information about these failed calls, but this can be changed by editing the bin/logback.xml file. (Set, for example, the level of org.copperengine.demo to debug.)

Creating a custom runtime image

Gradle must use Java 11 in order to be able to build the project. To create the custom runtime image execute:

./gradlew jlink

The runtime image will be available in the build/copper-modular-demo-image directory.