Skip to main content

​
Benefits

  • Zero-copy sharing: BlockDB exposes tables through Delta Sharing so you query them from any Databricks workspace (or other compatible client).
  • Unity Catalog integration: Access is enforced via share-level permissions, keeping audits centralized.
  • Supports both archive snapshots and incremental change data (table_changes()), ideal for lakehouse pipelines.

​
Getting Connected

  1. Provide your Unity Catalog metastore ID and share invite email to [email protected].
  2. BlockDB sends a Delta Share invitation (e.g., blockdb_archive).
  3. Accept the share in Databricks:
CREATE SHARE IF NOT EXISTS blockdb_archive SHARE_ID 'xxxxxxxx-xxxx';
CREATE RECIPIENT IF NOT EXISTS blockdb_team USING IDENTITY 'your-uc-principal';
GRANT USAGE ON SHARE blockdb_archive TO RECIPIENT blockdb_team;
  1. Create a catalog in your workspace to expose the shared tables.

​
Table Layout

  • Each dataset (blocks, transactions, pricing, etc.) appears as a Delta table with the same schema as /BlockDb.Postgres.Tables.Public. -, _tracing_id, _created_at, _updated_at, and verification columns are preserved.
  • Change data can be accessed via:
SELECT * FROM table_changes(blockdb_archive.blocks_0101, 'latest');

​
Operational Guidance

  • Use Auto Loader or streaming jobs to merge the shared tables into your production catalogs if you need custom partitioning.
  • Monitor _updated_at and compare against Data Freshness commitments.
  • Keep an eye on Unity Catalog grants to ensure only intended principals can read the shared datasets.
If you also rely on cloud buckets (S3/Azure/GCS), Delta Share can coexist—use it for analytics while buckets feed downstream archival storage.