Talend / MuleSoft — private Maven repository

1. Design the authorized distribution

1. Confirm the customer's Talend Studio/runtime release, Spark runtime, Java/Scala compatibility, approved Nexus or Artifactory service, and repository URLs.
2. For MuleSoft, document whether the Mule flow will submit/poll a Spark job or read the committed results through an existing database/API connector.
3. Approve the customer, version, repositories, users/workloads, environments, commercial agreement/order, and support contacts.
4. Separate evaluation, release, and snapshot repositories. Production DataAI version 1.0.0 belongs in a release repository, never a snapshot repository.

2. Publish the DataAI Maven modules

1. Obtain the authorized Maven production bundle and production checksum manifest.
2. Compare each JAR and POM hash with the approved manifest. Inspect group ID, artifact ID, version, dependency graph, license/documentation metadata, and included notices.
3. Create a publisher identity with deploy-only rights and customer/build identities with read-only rights. Keep credentials in approved secret stores.
4. Deploy POMs and JARs in dependency order using the repository's supported Maven deployment method. Do not deploy a different file under an existing release coordinate.
5. Open the repository UI and confirm that every required DataAI 1.0.0 coordinate is present and anonymous access is disabled.
6. From a clean Maven cache and a read-only test identity, resolve the documented top-level dependency and confirm all required DataAI modules download from the intended private source.
7. Run the packaged Spark smoke test with the resolved artifacts and record the runtime versions, hashes, test result, publisher, and date.

3. Configure Talend consumption

1. In Talend Studio, open the artifact repository configuration used by the customer's governed build process.
2. Select the supported Nexus 3, Nexus, or Artifactory repository type and enter the customer's repository/group URL exactly as supplied by its repository administrator.
3. Configure authentication through the customer-approved credential facility. Do not embed a password in a Job, exported archive, screenshot, README, or source repository.
4. Add the approved DataAI dependencies to the Talend Spark Job through the supported module/dependency management interface.
5. Build/export the Job and inspect its dependency output to confirm only the intended DataAI version is present and Spark/Hadoop platform libraries have not been duplicated incorrectly.
6. Run a clean isolated job with fictional data, validate the DataAI output contract, and then promote through the customer's normal Talend release gates.

4. Configure MuleSoft orchestration or output consumption

1. Keep the DataAI computation in the approved Spark/Talend job. Expose only the customer-approved job-control API, queue, scheduler, or governed output table.
2. Configure the Mule flow to submit work with a correlation ID, poll or receive completion status, enforce a timeout, and route non-success status to an error path.
3. Use an existing supported MuleSoft database/API connector to read committed results. Apply least-privilege credentials and the customer's retry/idempotency policy.
4. Test success, DataAI validation failure, Spark failure, timeout, duplicate request, and unavailable output. Confirm logs do not include credentials or protected source rows.
5. If publishing documentation or a custom asset to the customer's private Anypoint Exchange, use the current Exchange Maven publishing process and clearly state that the asset does not contain the proprietary DataAI Spark JARs.

5. Fulfillment and lifecycle controls

• Provide repository URL, exact coordinates/version, checksum record, credential-enrollment instructions, production installation and usage guides, entitlement reference, and support path.
• Issue a new version for changed bytes; retain or remove older versions according to the agreement and customer's rollback policy.
• Revoke read access when authorization ends and retest after Talend, Mule, Spark, or Java upgrades.

Official references

• Qlik Talend: configure artifact repositories
• MuleSoft: publish Exchange assets with Maven

Back to DataAI ETL Markets