Data Governance Catalogs and Discovery
Parent: Data Analysis · Published reference · snapshot 2026-09-08 · skill da-data-engineering-platform/references/da-30-data-governance-catalogs.md
↓ Facts as markdownall context files
Data governance is the discipline of exercising authority, control, and shared decision-making over the management of data assets — who can take what action, on which data, under what circumstances, u
These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.
Overview
- Data governance is the discipline of exercising authority, control, and shared decision-making over the management of data assets - who can take what action, on which data, under what circumstances, using what methods. Catalogs and discovery are the operational layer that makes governance executable: a metadata platform that inventories assets, attaches meaning (glossaries, tags, classifications), traces movement (lineage), assigns accountability (stewardship), and exposes it all through search so people can find and trust data. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#overview]
- This skill covers governance as a practice and an architecture, not the adjacent disciplines. Privacy law and ethics live in da-11; operational reliability/freshness monitoring lives in da-19; pipeline construction lives in da-13. Focus here: frameworks → metadata → discovery → lineage → meaning → roles → classification → access → products/mesh → tooling. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#overview]
- Two macro-shifts define the 2024–2026 landscape: [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#overview]
- Passive → active metadata. Catalogs stop being static inventories and become bidirectional orchestration layers that push metadata back into the stack to drive automation (Gartner; Atlan). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#overview]
- Centralized → federated governance. Data mesh reframes governance as federated computational governance - global rules enforced computationally, local ownership by domain teams (Dehghani; Fowler). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#overview]
1. Governance frameworks: DAMA-DMBOK and DCAM
- DAMA-DMBOK (Data Management Body of Knowledge, DAMA International) organizes data management into 11 knowledge areas rendered as the "DAMA wheel" with Data Governance at the hub: governance, architecture, modeling, storage & operations, security, integration & interoperability, document & content management, reference & master data, data warehousing & BI, metadata, and data quality. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#1-governance-frameworks-dama-dmbok-and-dcam]
- DCAM (Data Management Capability Assessment Model, EDM Council) is a maturity-assessment standard organized into eight core components (data strategy, business case & funding, governance, architecture, technology, quality, operations, control environment). DCAM v3 (released 2024) is the current standard. DCAM measures how mature the program is; DMBOK describes what the disciplines are. CDMC (Cloud Data Management Capabilities) extends DCAM-style assessment to cloud + sensitive-data controls. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#1-governance-frameworks-dama-dmbok-and-dcam]
2. Metadata management & active metadata
- Metadata = data about data: technical (schemas, types, partitions), business (definitions, glossary terms, ownership), operational (run logs, freshness, query frequency). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#2-metadata-management-active-metadata]
- Passive metadata sits in a static catalog read by humans. Active metadata is continuously analyzed, curated, and pushed back into tools to drive automation - e.g., auto-propagating a PII tag from a source column to every downstream table. Metadata moves both directions. Gartner projected ~30% of orgs adopting active metadata by 2026 with up to 70% faster time-to-delivery, and reframed metadata management as foundational to AI readiness in its 2025 Magic Quadrant (first refresh in five years, Nov 2025). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#2-metadata-management-active-metadata]
3. Data discovery & search
- Discovery is the consumer entry point: search across assets ranked by relevance, enriched with ownership, quality, popularity, and lineage so a user can judge trustworthiness. Modern catalogs add natural-language/conversational search (Atlan) and query-log ingestion (Alation) that mines actual execution patterns to rank assets. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#3-data-discovery-search]
4. Data lineage — table-level and column-level
- Table-level lineage answers "which datasets feed which." Column-level lineage maps dependencies field-by-field, enabling precise impact analysis ("if I drop this column, what breaks?") and root-cause analysis. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#4-data-lineage-table-level-and-column-level]
- Lineage is derived by SQL parsing; parser choice matters: DataHub uses SQLGlot (schema-aware, highest correct-lineage rate); OpenMetadata uses sqllineage; OpenLineage/Marquez uses openlineage-sql. Native column-level support covers Snowflake, BigQuery, Databricks, and BI tools (Looker, Power BI, Tableau). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#4-data-lineage-table-level-and-column-level]
5. Business glossaries
- A business glossary is the controlled vocabulary of agreed business terms (e.g., "active customer") with definitions, owners, and relationships, linked to physical assets so technical columns inherit business meaning. Distinct from a data dictionary (technical, schema-level) and a taxonomy/ontology (semantic relationships). Stewards own glossary curation. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#5-business-glossaries]
6. Data stewardship & ownership roles
- Data Owner - accountable (senior business role) for classification, protection, use, and quality of a domain; results-focused; signs off on glossary and access policy. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#6-data-stewardship-ownership-roles]
- Data Steward - responsible for quality, definitions, documentation, glossary, and lineage; task-focused; day-to-day governance operator. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#6-data-stewardship-ownership-roles]
- Data Custodian - IT role; implements and maintains storage/security controls the Owner specifies; handles access provisioning, incident review, platform monitoring. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#6-data-stewardship-ownership-roles]
- RACI: Owner = Accountable, Steward = Responsible (meaning/quality), Custodian = Responsible (technical controls). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#6-data-stewardship-ownership-roles]
7. Data classification & tagging
- Classification assigns sensitivity levels (public / internal / confidential / restricted) and compliance categories (PII, GDPR, HIPAA). Modern platforms auto-discover and tag sensitive data - Unity Catalog uses an agentic/AI classifier for continuous PII discovery. Governed tags enforce a controlled tag vocabulary (vs. free-form) so policies key off them reliably. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#7-data-classification-tagging]
8. Access governance & policy enforcement
- Move from per-object grants to policy-as-data. ABAC (attribute-based access control) evaluates tag-based conditions and applies row filters (which rows you see) and column masks (what values you see) automatically across catalogs/schemas - e.g., mask any column tagged PII unless the user is in pii-readers. Unity Catalog made ABAC row filters, column masks, governed tags, and data classification GA in 2025. Pair classification (find) + tags (label) + ABAC (enforce) for scalable, declarative governance. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#8-access-governance-policy-enforcement]
9. Data products, data mesh & federated computational governance
- Data mesh (Zhamak Dehghani, ThoughtWorks 2019) is a sociotechnical approach to analytical data at scale on four principles: domain-oriented ownership, data as a product, self-serve data platform, and federated computational governance. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#9-data-products-data-mesh-federated-computational-governance]
- Federated computational governance = a decision model led by a federation of domain + platform product owners with local autonomy, adhering to global rules enforced computationally (encoded into the platform, not by committee). A data product is the smallest architectural unit encapsulating everything needed to share data (data + metadata + code + access + SLOs), owned by the domain team. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#9-data-products-data-mesh-federated-computational-governance]
10. Data contracts & specifications
- A data contract is an enforceable agreement between producer and consumer covering schema, semantics, quality, and SLAs. ODCS (Open Data Contract Standard, v3.x, governed by Bitol, a Linux Foundation AI & Data project; originated at PayPal) defines schema-level executable contracts. ODPS (Open Data Product Specification) is broader - design, publish, discover, monetize, govern data products as business-value units, and can reference ODCS contracts inline or by URL. Use ODCS for the interface; ODPS for the product wrapper. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#10-data-contracts-specifications]
Tools & Frameworks
- Frameworks: DAMA-DMBOK (scope/vocabulary), DCAM v3 + CDMC (maturity assessment), data mesh (federated operating model), ODCS/ODPS (contracts & product specs). [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#tools-frameworks]
Methodology — standing up governance + a catalog
- Frame the operating model. Centralized vs. federated (mesh). Map domains. Assign Owner/Steward/Custodian per domain (RACI). Use DCAM to baseline maturity and set a roadmap. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Pick the scope that delivers value first. Highest-value or highest-risk domains, not boil-the-ocean. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Ingest technical metadata. Connect sources (warehouses, lakes, BI, dbt); auto-harvest schemas + lineage; verify column-level coverage for key dialects. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Layer meaning. Build the business glossary; link terms to physical assets. Stewards curate. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Classify & tag. Run automated sensitive-data classification; apply governed tags, not free-form. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Enforce access declaratively. Define ABAC policies keyed on tags (mask PII, row-filter by region). Test propagation across schemas. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Activate the metadata. Wire automation: tag propagation along lineage, freshness/popularity into search ranking, push-back to source tools. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Operationalize. Stewardship rituals, glossary review cadence, data contracts (ODCS) on critical interfaces, product specs (ODPS) for shared data products. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
- Measure. Coverage (% cataloged/owned/classified), adoption (search usage, time-to-find), trust (% certified), policy compliance. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#methodology-standing-up-governance-a-catalog]
Practical Patterns
- Certify, don't catalog everything. A "verified/certified" badge on trusted assets beats 100% coverage of unmanaged junk. Discovery is about trust, not census. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Tag-driven policy. Classify → governed tag → ABAC. One policy ("mask PII") covers thousands of objects and auto-applies to new ones. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Propagate along lineage. Use column-level lineage to auto-inherit classifications/tags downstream. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Glossary terms as the bridge. Bind business terms to physical columns so non-technical users search in business language. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Federated rules, central platform. In mesh, encode global rules computationally in the self-serve platform; let domains own products within those rails. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Contracts on the boundaries. Put ODCS contracts on cross-domain/producer-consumer interfaces where breakage is expensive; don't contract everything. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
- Query logs for relevance. Rank search and recommend assets by actual usage (Alation-style), not alphabetical or last-modified. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#practical-patterns]
Anti-Patterns
- Catalog as a graveyard. One-time bulk ingest, no stewardship, no owners, stale within months. Governance is a continuous program, not a project. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Passive metadata only. Treating the catalog as a read-only wiki - no automation, no push-back. Metadata that doesn't act decays. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Free-form tag sprawl. Uncontrolled tags (pii, PII, personal, sensitive) make policies unreliable. Use governed vocabularies. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Governance-by-committee bottleneck. Central team must approve every change - kills velocity. Federate ownership; enforce computationally. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Owner/steward/custodian conflation. One overloaded "data person" can't be accountable, responsible for meaning, and run the platform. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Table-level lineage where column-level is needed. Impact analysis on a schema change is guesswork without field-level lineage. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Tool-first, model-last. Buying Collibra/Atlan before defining domains, roles, and policies yields shelfware. Operating model first. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
- Boil-the-ocean rollout. Cataloging every asset before any are governed. Start narrow, prove value, expand. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#anti-patterns]
Troubleshooting
- Lineage incomplete / missing columns. Check parser/dialect support (SQLGlot vs sqllineage), ensure schema context, confirm the connector ingests query history (not just DDL). Dynamic SQL and SELECT * degrade column-level resolution. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- Sensitive data slipping through. Automated classification missed it - re-run/expand classifiers, add custom patterns, propagate along lineage so derived columns inherit the tag. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- ABAC policy not applying. Verify the object carries the governed tag the policy keys on, that the policy is at the right catalog/schema scope, and that classification ran before policy evaluation. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- Low catalog adoption. Usually a trust/relevance problem: no owners, no certification, poor search ranking. Add ownership, certify key assets, rank by query-log popularity, link glossary terms. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- Purview ↔ Unity Catalog drift. Schema/lineage/classification out of sync - confirm connector/API sync cadence; Purview is discovery/technical-metadata, UC is the enforcement plane. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- Glossary nobody uses. Terms not linked to physical assets, or no steward cadence. Bind terms to columns; put glossary review in the stewardship ritual. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
- Mesh governance chaos. Global rules defined but not computational - encode them into the self-serve platform; agreement docs don't enforce. [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#troubleshooting]
References
- DAMA International - DAMA-DMBOK: https://www.dama.org/cpages/body-of-knowledge ; https://www.damadmbok.org/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Snowflake - "DAMA-DMBOK Explained": https://www.snowflake.com/en/fundamentals/data-governance/framework/dama-dmbok/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- EDM Council - "Announcing DCAM v3": https://edmcouncil.org/announcement/announcing-dcam-v3-meet-the-new-standard-for-your-data/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- EDM Council - DCAM framework: https://edmcouncil.org/frameworks/dcam/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Gartner - Magic Quadrant for Metadata Management Solutions (Nov 19, 2025): https://www.informatica.com/metadata-management-magic-quadrant.html ; https://atlan.com/gartner-magic-quadrant-for-metadata-management/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Gartner - Market Guide for Active Metadata Management: https://www.gartner.com/en/documents/4004082 (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- OvalEdge - "Active Metadata Management": https://www.ovaledge.com/blog/active-metadata/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- DataHub - "How DataHub's Column-Level Parser Works": https://datahub.com/blog/extracting-column-level-lineage-from-sql/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- DataHub Docs - Lineage feature guide: https://docs.datahub.com/docs/features/feature-guides/lineage (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- OpenMetadata Docs - "How Column-Level Lineage Works": https://docs.open-metadata.org/latest/how-to-guides/data-lineage/column (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Atlan - "Alation vs Collibra vs OpenMetadata vs Atlan": https://atlan.com/alation-vs-collibra-vs-openmetadata-vs-atlan/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Atlan - "16 Best Data Catalog Tools": https://atlan.com/data-catalog-tools/ (2026) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- DQOps - "Data Owner vs Data Steward vs Data Custodian": https://dqops.com/data-owner-data-steward-data-custodian-roles/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- EWSolutions - "Data Stewardship Roles": https://www.ewsolutions.com/data-stewardship-roles-a-complete-guide/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Databricks - "Find Sensitive Data at Scale with Data Classification in Unity Catalog": https://www.databricks.com/blog/find-sensitive-data-scale-data-classification-unity-catalog (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Databricks - "ABAC row filtering and column masking GA in Unity Catalog": https://www.databricks.com/blog/abac-row-filtering-and-column-masking-policies-governed-tags-and-data-classification-are-now (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Microsoft Learn - Unity Catalog Data Classification: https://learn.microsoft.com/en-us/azure/databricks/data-governance/unity-catalog/data-classification (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Zhamak Dehghani - "Data Mesh Principles and Logical Architecture" (martinfowler.com): https://martinfowler.com/articles/data-mesh-principles.html (2020) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Dehghani - Data Mesh: Delivering Data-Driven Value at Scale, O'Reilly (2022) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Starburst - "Federated Computational Governance": https://www.starburst.io/blog/data-mesh-book-bulletin-principle-of-federated-computational-governance/ (2024) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Bitol / Linux Foundation - Open Data Contract Standard (ODCS) v3.x: https://github.com/bitol-io/open-data-contract-standard ; https://bitol-io.github.io/open-data-contract-standard/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Open Data Product Specification (ODPS) - "ODPS vs ODCS": https://blog.opendataproducts.org/when-standards-collide-clarifying-odps-and-odcs-in-the-data-product-landscape-c2978f9c13d9 (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- DataHub Docs - Architecture Overview: https://docs.datahub.com/docs/architecture/architecture (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- Atlan - "OpenMetadata Explained": https://atlan.com/openmetadata-explained/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- TheDataGuy - "Open-Source Data Governance Frameworks": https://thedataguy.pro/writing/2025/08/open-source-data-governance-frameworks/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
- OpenMetadata Standards: https://openmetadatastandards.org/ (2025) [archived source: https://llms-explorer.com/sources/mdb-context-hub/da-30-data-governance-catalogs/#references]
Project ideas
- Stand up a metadata catalog such as DataHub or OpenMetadata and wire column-level lineage so a schema-change impact-analysis question can actually be answered. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Build an ABAC access-control layer that keys row filters and column masks off governed classification tags, so a single mask-PII policy automatically applies to new tables as they're tagged. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Set up an automated sensitive-data classifier, such as Unity Catalog's PII discovery, and propagate its tags along column-level lineage so derived downstream tables inherit the classification. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Draft an ODCS (Open Data Contract Standard) data contract for one high-value producer-consumer interface, covering schema, semantics, quality, and SLAs, rather than trying to contract every interface at once. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Where this helps
- Deciding whether to centralize or federate governance for an organization, using DCAM to baseline current maturity and set a realistic roadmap rather than guessing at scope. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- A schema-change impact-analysis question — what breaks downstream if this column changes — that only column-level, not table-level, lineage can actually answer. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Standing up self-serve data discovery where users need to judge an asset's trustworthiness from ownership, quality, popularity, and lineage signals, not just its existence in a catalog. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Enforcing access control across many tables and schemas using tag-based ABAC policies instead of maintaining per-object grants by hand as the data estate grows. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Common mistakes
- Bulk-ingesting an entire data estate into a catalog once, with no ongoing stewardship or ownership assigned, producing a catalog that goes stale within months. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Treating the catalog as a passive, read-only wiki instead of wiring active metadata to automatically propagate tags and push signals back into the stack. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Letting tags sprawl in free-form variants such as pii, PII, personal, and sensitive instead of a governed vocabulary, which makes any tag-keyed access policy unreliable. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Buying a catalog tool before defining domains, ownership roles, and policies, which reliably produces expensive shelfware instead of a working governance program. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Known issues
- Column-level lineage accuracy depends heavily on SQL parser and dialect support, and dynamic SQL or SELECT * usage degrades resolution regardless of which parser is used. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- A federated, data-mesh-style governance model still fails if global rules are only written down and agreed to rather than actually enforced computationally in the self-serve platform. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Owner, steward, and custodian are meant to be distinct roles, but a single overloaded data person covering all three is a common and structurally weak setup in smaller teams. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Passive vs. active metadata is a real architectural distinction, not just a marketing term — a catalog that never pushes metadata back into the stack to drive automation stays a static inventory no matter how complete its coverage is. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Children
- DAMA-DMBOK and DCAM governance frameworks (frontier)
- Metadata management and active metadata (frontier)
- Data discovery and search (frontier)
- Table-level and column-level data lineage (frontier)
- Business glossaries (frontier)
- Data stewardship and ownership roles (frontier)
- Data classification and tagging (frontier)
- Access governance and policy enforcement (ABAC) (frontier)
- Data products and data mesh federated computational governance (frontier)
- Data contracts (ODCS and ODPS) (frontier)
- Catalog tooling landscape (DataHub, OpenMetadata, Amundsen, Atlan, Collibra, Alation, Unity Catalog, Purview) (frontier)
Frontier under this node: Access governance and policy enforcement (ABAC), Business glossaries, Catalog tooling landscape (DataHub, OpenMetadata, Amundsen, Atlan, Collibra, Alation, Unity Catalog, Purview), DAMA-DMBOK and DCAM governance frameworks, Data classification and tagging, Data contracts (ODCS and ODPS), Data discovery and search, Data products and data mesh federated computational governance, Data stewardship and ownership roles, Metadata management and active metadata, Table-level and column-level data lineage