Databricks — Databricks Marketplace or Unity Catalog volume

1. Prepare the immutable Databricks release

1. Verify customer/listing authorization, license/order, cloud and Region, workspace, Unity Catalog metastore, compute access mode, Spark/Scala/Java runtime, users, and support owner.
2. Obtain DataAIETL-Databricks-1.0.0.zip and ARTIFACTS.sha256. Recompute hashes and inspect JARs, notebooks, configuration examples, documentation, license references, and sample data.
3. Run the packaged local smoke test when compatible, then test on the claimed Databricks Runtime with an isolated catalog/schema and fictional data.
4. Run 00_INSTALL_AND_VERIFY.py before quality, analytics, matrix, and BI examples. Record output, runtime, access mode, hashes, and approver.

2A. Fulfill through a customer Unity Catalog volume

1. Have the customer create or select a governed catalog, schema, and volume for licensed DataAI artifacts. Use managed storage or an approved external location according to customer policy.
2. Grant the uploader only the required catalog/schema/volume permissions. Grant production job principals READ VOLUME and avoid broad workspace-user access.
3. Upload the immutable 1.0.0 JARs, notebooks/configs where appropriate, checksum manifest, and license certificate reference. Use a versioned directory and do not overwrite it.
4. Compare hashes after upload/download. Add the approved volume JARs as Databricks Job or compute libraries using the current supported UI/API.
5. For standard access mode, configure any required library allowlist and confirm permissions with the actual job service principal.
6. Import or reference the notebooks, replace sample locations with governed customer locations, keep secrets in Databricks secret management, and run the complete acceptance sequence.
7. Record volume path, object hashes, grants, job/compute ID, runtime, customer acceptance, license/order reference, and removal date.

2B. Publish through Databricks Marketplace

1. Apply to become a Databricks Marketplace provider and complete the business/provider profile, public organization name, support contacts, legal terms, privacy information, and Databricks account/workspace prerequisites.
2. Have an account/workspace administrator grant the required Marketplace provider and data-sharing privileges to the publishing identity.
3. Choose Free and instantly available only if the listing can be delivered through a compliant share without a separate agreement. Otherwise choose Requires approval so Yanbor can validate the customer and complete commercial terms before access.
4. Create the governed provider-side share and add only eligible reviewed assets. For this package, use the current Files/Volumes and notebook capabilities only if enabled and accepted in the provider workspace.
5. Do not include credentials, customer data, unrestricted proprietary source, or unapproved dependencies. Use a versioned artifact directory and a fictional sample/verification notebook.
6. Create the listing and complete name, summary, detailed description, categories, industries, provider contact, support/documentation links, terms, privacy, sample notebooks, regions, and access model.
7. Explain precisely how the recipient obtains access, attaches the JARs to a job/compute resource, runs verification, confirms commercial rights, and requests support. State all runtime and Unity Catalog prerequisites.
8. Preview and test as a recipient in a separate clean workspace/account where possible. Validate request approval, share access, file/notebook visibility, hash verification, library attachment, and revocation.
9. Submit for Databricks review and resolve all findings. Publish only after Databricks and Yanbor approvals, then verify the live listing and consumer acquisition path.

3. Approve request-access customers

1. Verify the requester's organization and intended environments, complete the commercial agreement/order, and approve only the entitled recipient identifier.
2. Grant access to the correct share/assets and send the version, hashes, license certificate/order reference, installation and usage guides, and support path.
3. Record activation and acceptance. Revoke the recipient/share permission when the term ends or access is withdrawn.

4. Release lifecycle

• Publish changed bytes under a new versioned path and update the listing/release notes through Databricks review where required.
• Retest all claimed Databricks Runtime and compute access modes.
• Never imply Databricks endorsement beyond the actual status of the exact live listing.

Official references

• Databricks: get started as a Marketplace provider
• Databricks: create and manage a listing
• Databricks: work with files in Unity Catalog volumes

Back to DataAI ETL Markets