/MinimumViableDataspace

Guidance on documentation, scripts and integration steps on using the EDC project results

Primary LanguageJavaApache License 2.0Apache-2.0

Minimum Viable Dataspace

The Minimum Viable Dataspace (MVD) is a sample implementation of a dataspace that leverages the Eclipse Dataspace Connector (EDC). The main purpose is to demonstrate the capabilities of the EDC, make dataspace concepts tangible based on a specific implementation, and to serve as a starting point to implement a custom dataspace.

The MVD allows developers and decision makers to gauge the current progress of the EDC and its capabilities to satisfy the functionality of a fully operational dataspace.

As a fully decentralized dataspace is hard to imagine, the MVD also serves the purpose of demonstrating how decentralization can be practically implemented.

Documentation

Developer documentation can be found under docs/developer, where the main concepts and decisions are captured as decision records.

Create Dataspace Deployment

To be able to deploy your own dataspace instances, you first need to fork the MVD repository and set up your environment.

Once your environment is set up, follow these steps to create a new dataspace instance:

  • Go to your MVD fork in GitHub.
  • Select the tab called Actions.
  • Select the workflow called Deploy.
  • Provide your own resources name prefix. The prefix must be 3 to 7 lowercase letters and digits, starting with a letter. This name prefix ensures the resources name's uniqueness and avoids resource name conflicts. Note down the used prefix.
  • Click on Run workflow to trigger the deployment.

Destroy Dataspace Deployment

Follow these steps to delete a dataspace instance and free up the corresponding resources:

  • Go to your MVD fork in GitHub.
  • Select the tab called Actions
  • Select the workflow called Destroy
  • Click on Run workflow
  • Provide the resources prefix that you used when you deployed your DataSpace.
  • Click on Run workflow to trigger to destroy your MinimumViableDataspace DataSpace.

Local Development Setup

The MVD backend and MVD UI (Data Dashboard) can be run locally for testing and development.

  1. Check out the repository eclipse-dataspaceconnector/DataDashboard or your corresponding fork.
  2. Set the environment variable MVD_UI_PATH to the path of the DataDashboard repository. (See example below.)
  3. Use the instructions in section Publish/Build Tasks system-tests/README.md to set up a local MVD environment with the exception to use the profile ui. (See example below.)
    • In order to verify your local environment works properly, also follow section Local Test Execution in system-tests/README.md .

Using the profile ui will create three MVD UIs (Data Dashboards) for each EDC participant in addition to the services described in system-tests/README.md.

Bash:

export MVD_UI_PATH="/path/to/mvd-datadashboard"
docker-compose --profile ui -f system-tests/docker-compose.yml up --build

PowerShell:

$Env:MVD_UI_PATH="/path/to/mvd-datadashboard"
docker-compose --profile ui -f system-tests/docker-compose.yml up --build

In Windows Docker Compose expects the path to use forward slashes instead of backslashes.

The profile ui creates three Data Dashboards each connected to an EDC participant. The respective app.config.json files can be found in the respective directories:

  • resources/appconfig/company1/app.config.json
  • resources/appconfig/company2/app.config.json
  • resources/appconfig/company3/app.config.json

That's it to run the local development environment. The following section Run A Standard Scenario Locally describes a standard scenario which can be optionally used with the local development environment.

Tip: The console output from the services spun up by Docker compose can be noisy. To decrease the output from the services on the console set EDC_CATALOG_CACHE_EXECUTION_PERIOD_SECONDS to a higher value, e.g. 60, for each EDC participant in system-tests/docker-compose.yml.

Note: The container cli-tools will turn into the state healthy after registering successfully all participants and will keep running as an entrypoint to the services created by Docker compose. This is useful for local development in order to manually check commands against the participants (e.g. company1, company2, company3).

Sample how to enter the container cli-tools and test a command manually.

Host:

docker exec -it cli-tools bash

Container:

java -jar registration-service-cli.jar \
>    -d=did:web:did-server:registration-service \
>    --http-scheme \
>    -k=/resources/vault/company1/private-key.pem \
>    -c=did:web:did-server:company1 \
>    participants get

Output (container)

{
  "did" : "did:web:did-server:company1",
  "status" : "ONBOARDED"
}

Run A Standard Scenario Locally

Prerequisite: create a test document manually:

  • Connect to the local blob storage account (provided by Azurite) of company1.
    • Storage account name: company1assets, storage account key: key1.
    • Microsoft Azure Storage Explorer can be used to connect to the local storage account on localhost:10000.
  • Create a container named src-container. (Container name is defined for Postman request Publish Master Data in deployment/data/MVD.postman_collection.json)
  • Copy deployment/terraform/participant/sample-data/text-document.txt into the newly created container.
    • N.B.: it does not have to be this exact file as long you create a file which has the name text-document.txt.

All this can also be done using Azure CLI with the following lines from the root of the MVD repository:

Bash:

conn_str="DefaultEndpointsProtocol=http;AccountName=company1assets;AccountKey=key1;BlobEndpoint=http://127.0.0.1:10000/company1assets;"
az storage container create --name src-container --connection-string $conn_str
az storage blob upload -f ./deployment/terraform/participant/sample-data/text-document.txt --container-name src-container --name text-document.txt --connection-string $conn_str

PowerShell:

$conn_str="DefaultEndpointsProtocol=http;AccountName=company1assets;AccountKey=key1;BlobEndpoint=http://127.0.0.1:10000/company1assets;"
az storage container create --name src-container --connection-string $conn_str
az storage blob upload -f .\deployment\terraform\participant\sample-data\text-document.txt --container-name src-container --name text-document.txt --connection-string $conn_str

This should result in a similar output as follows. Via the Microsoft Azure Storage Explorer it would be possible to review the new container and the uploaded blob.

{
  "created": true
}

Finished[#############################################################]  100.0000%
{
  "etag": "\"0x1CC7CAB96842160\"",
  "lastModified": "2022-08-08T15:14:01+00:00"
}

The following steps initiate and complete a file transfer with the provided test document.

  • Open the website of company1 (e.g. http://localhost:7080) and verify the existence of two assets in the section Assets.
  • Open the website of the company2 (e.g. http://localhost:7081) and verify six existing assets from all participants in the Catalog Browser.
    • In the Catalog Browser click Negotiate for the asset test-document_company1.
      • There should be a message Contract Negotiation complete! Show me! in less than a minute.
  • From the previous message click Show me!. If you missed it, switch manually to the section Contracts.
    • There should be a new contract. Click Transfer to initiate the transfer process.
    • A dialog should open. Here, select as destination AzureStorage and click Start transfer.
    • There should be a message Transfer [id] complete! Show me! in less than a minute. (Where id is a UUID.)
  • To verify the successful transfer the Storage Explorer can be used to look into the storage account of company2.
    • Storage account name and key is set in system-tests/docker-compose.yml for the service azurite. Default name is company2assets, key is key2.
    • There should be new container in the storage account containing two files .complete and test-document.txt.

Contributing

See how to contribute.