# OCI Data Catalog

> Managed metadata catalog for Oracle Cloud that harvests Oracle databases, Object Storage and other sources, with a business glossary, lineage from OCI Data Integration and Data Flow, and a Hive metastore for Spark.

Source: https://toolsfordata.com/tools/oci-data-catalog/ · Updated 2026-11-19

[Website](https://www.oracle.com/big-data/data-catalog/) · [Docs](https://docs.oracle.com/en-us/iaas/Content/data-catalog/home.htm)

Best for: data professionals and data engineers running workloads on Oracle Cloud Infrastructure who need a native catalog integrated with OCI Data Flow and Oracle Autonomous Database.

## Overview

- Core Strength: Free catalog and Hive metastore for Oracle Cloud, with column-level lineage from OCI jobs
- Lineage Depth: Partial — automated, table, column, impact analysis…
- Catalog Curation: Partial — glossary, custom fields
- Stewardship Workflows: Not supported
- Policy & Classification: Not supported
- Compliance Depth: Partial — audit trail
- Marketplace & Data Products: Not supported
- Built-in Data Quality: Not supported
- AI & Agent Interface: Not supported
- AI Governance: Not supported
- Developer Surface: Partial — API, Python SDK, webhooks
- Deployment Model: Partial — SaaS, data residency
- Security & Access: Full — SSO, SCIM, RBAC, audit log…
- Notable Caveat: Few releases since 2024, Oracle-centric connectors, and Oracle's newer catalogs get the investment

## What it's used for

OCI Data Catalog is Oracle Cloud's managed metadata catalog: teams on OCI use it to harvest metadata from Oracle databases and Object Storage, link it to a glossary, and share a Hive metastore with Spark jobs.

## Catalog

OCI Data Catalog harvests Oracle databases, Object Storage, MySQL, PostgreSQL, SQL Server, Db2, Hive and Kafka, with a glossary whose terms have owners and a review status. There are no Snowflake, Databricks, S3 or BI connectors.

## Lineage

OCI Data Catalog shows table and column lineage for OCI Data Integration and Data Flow jobs, and imports custom lineage through an OpenLineage-compatible API (May 2024).

## AI

Machine learning recommends glossary terms for tables and columns. There's no chat assistant or MCP server.

## Pricing

Oracle says current Oracle Cloud users can use OCI Data Catalog for free, and it has no line on the OCI price list.

## Deployment

OCI Data Catalog is a managed OCI service only, reached privately through private endpoints, and also runs in OCI US Government Cloud.

## Security

OCI Data Catalog uses OCI IAM for SSO, SCIM and fine-grained policies, with OCI Audit logging every API call.

## Integrations

OCI Data Catalog harvests Oracle Database, Autonomous Database and Object Storage, plus MySQL, PostgreSQL, SQL Server, Hive and Kafka.

## Review

Strengths:
- Free for Oracle Cloud customers.
- A Hive metastore shared with OCI Data Flow Spark jobs, alongside the catalog.
- Column-level lineage from OCI jobs, plus OpenLineage-compatible import.

Caveats:
- Few releases since 2024, while Oracle's newer catalogs (AI Data Platform Master Catalog, Autonomous AI Database Catalog) get the investment.
- No sensitive data discovery, policies or data quality: Oracle sells discovery and masking in Data Safe.
- Connectors are mostly Oracle and classic databases, so lineage doesn't reach dashboards.

Recommendation: Use OCI Data Catalog if your data lives on Oracle Cloud and a basic, free catalog is enough. For governance across platforms, consider a cross-platform catalog such as DataHub or Collibra.

## Facts

- OCI Data Catalog became generally available in February 2020.
- Oracle launched the AI Data Platform with a separate Master Catalog in October 2025.
- Oracle still calls it OCI Data Catalog; 'OCI Catalog' isn't an Oracle name.

## Pricing

- Model: included with Oracle Cloud
- Free: Free for Oracle Cloud users
- Entry team price: $0
- Source: https://www.oracle.com/big-data/data-catalog/ (checked 2026-11-12)

## Features

### Data Catalog

- Automated Metadata Harvesting: Yes — Harvest jobs extract schemas, tables, columns and files from Oracle databases, Object Storage, MySQL, PostgreSQL, SQL Server, Db2, Hive and Kafka, on demand or on an hourly to monthly schedule. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/harvest-technical-metadata.htm))
- Search & Discovery: Yes — Keyword search with wildcards, filters, custom-property filters and sorting covers data assets, entities, attributes, glossary terms and jobs. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/use-search-bar.htm))
- AI-Powered Search: No — Search is case-insensitive keyword matching with wildcards; no natural language or semantic search is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/use-search-bar.htm))
- Usage & Popularity Statistics: No — Entity pages show summary, attributes and constraints; no query counts or popularity rankings are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/explore.htm))
- Custom Fields & Templates: Yes — Custom properties (plain or rich text, number, boolean, date, with lists of values) attach to chosen object types and can be shown, filtered and sorted in search. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/custom-properties-create.htm))
- Asset Certification: No — No certified, endorsed or deprecated status for data assets or entities is documented; only glossary terms have an approval status. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/explore.htm))
- Business Glossary: Yes — Glossaries hold categories, subcategories and terms with owners and a workflow status (New, Under Review, Approved, Escalated); only approved terms can be linked to objects. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/create-glossary-term.htm))
- Metadata Propagation: No — Tags, terms and custom properties are applied per object; no inheritance from parent to child objects is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/enrich-metadata.htm))

### Data Lineage

- Table-Level Lineage: Yes — The lineage graph shows entity-level upstream and downstream flow for Data Integration, Data Flow and imported lineage. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- Column-Level Lineage: Yes — Entities in the lineage graph expand to show column-level lineage, except for Data Integration Flatten, Pivot and Function operators. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- End-to-End Lineage: No — Lineage covers Data Integration, Data Flow and Spark jobs; no BI tools are harvested, so lineage does not reach dashboards. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- Impact Analysis: Yes — The lineage graph shows lineage to the left of the selected entity and downstream impact to the right; there is no separate impact report. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- Automated Lineage Extraction: Yes — Lineage is collected from OCI Data Integration tasks and OCI Data Flow Spark applications once lineage generation is switched on in each, so automated capture is limited to Oracle's own engines; other tools push OpenLineage-format payloads through the ImportLineage API. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- OpenLineage Support: Yes — The ImportLineage API accepts lineage payloads in an OpenLineage-compatible format, documented for Spark applications. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))

### Data Stewardship

- Critical Data Elements: No — No critical data element register or flag is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Review & Approval Workflows: No — Glossary terms carry a status field (New, Under Review, Approved, Escalated), but there are no configurable review workflows with assignment or escalation. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/create-glossary-term.htm))
- Governance Campaigns: No — Bulk custom property updates exist (March 2024), but no steward-run governance campaigns are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Governance Scorecards: No — No governance maturity scorecards are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Metadata Automation: Unknown — Harvest and custom property changes emit OCI Events that customers can route to Functions, but no built-in metadata rules are documented. ([source](https://docs.oracle.com/iaas/data-catalog/using/events.htm))

### Data Classification

- Sensitive Data Discovery: No — Data Catalog does not scan for PII; tags are free-form and applied by users. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/link-tags.htm))
- Data Policy: No — IAM policies control who can read or manage catalog objects, not who can query the underlying data. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/policies.htm))
- Native Policy Enforcement: No — No access or masking policies are pushed down to data platforms. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/policies.htm))
- Data Masking: No — No masking of column values; masking is offered by Oracle Data Safe, a separate service. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### AI Features

- AI Chat Assistant: No — No chat assistant is documented for Data Catalog. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- AI Documentation Generation: No — No LLM-generated descriptions are documented; descriptions are entered by users or harvested from sources. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/enrich-metadata.htm))
- Agentic Metadata Enrichment: No — Term recommendations are ML suggestions that users accept or reject one by one, not autonomous agents. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/recom.htm))

### Data Compliance

- Audit Trail: Yes — OCI Audit records Data Catalog API calls, including metadata, glossary and custom property changes. ([source](https://docs.oracle.com/en-us/iaas/Content/Audit/Concepts/auditoverview.htm))
- Compliance Reporting: No — No compliance dashboards or evidence packs are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Regulatory Framework Mapping: No — No mapping of assets or terms to GDPR, HIPAA or other regulations is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Activity Monitoring: No — Data Catalog does not monitor queries on source systems. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- ROPA: No — No GDPR Article 30 processing register. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Retention Policies: No — No retention or deletion rules for data assets are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Subject Rights: No — No DSAR or deletion request handling. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Collaboration

- Alert Notifications: No — No native Slack or Teams governance alerts; harvest and custom property events go to OCI Events. ([source](https://docs.oracle.com/iaas/data-catalog/using/events.htm))
- Discussions & Annotations: No — No comments or threaded discussions on assets are documented; enrichment is limited to tags, terms and custom properties. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/enrich-metadata.htm))
- Ticketing Integrations: No — No Jira or ServiceNow integration is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Data Marketplace

- Self-Service Access Requests: No — Access is granted by administrators through IAM policies; there is no Request Access button. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/policies.htm))
- Data Products: No — No data product objects are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Mesh Support: No — No domain workspaces or federated governance; organization is by compartments and glossaries. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/enrich-business-glossary.htm))
- Subscription Management: No — No subscriptions to datasets or data products. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Data Quality

- SQL Table Tests: No — No data quality tests are documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Health Scores: No — No quality or health score per asset. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Anomaly Detection: No — No anomaly detection on data values or volumes. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Data Profiling: No — No column statistics are computed; entity pages show attributes and harvested primary, unique and foreign key constraints. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/explore.htm))
- Data Contracts: No — No data contracts. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### AI Governance

- AI Asset Inventory: No — No AI model or agent inventory in Data Catalog. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- AI Lineage: No — Lineage covers Data Integration, Data Flow and Spark jobs, not model training. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/view-data-lineage-dis.htm))
- AI Regulatory Compliance: No — No EU AI Act or NIST AI RMF features. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- AI Model Health Score: No — No AI model scoring. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- AI Agent Access Governance: No — No governance applied at AI agent query time. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Master Data Management

- Master Data Management: No — Data Catalog is a metadata catalog with no MDM capabilities. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Workflows

- REST / GraphQL API: Yes — The Data Catalog REST API covers data assets, harvesting, glossaries, custom properties, search and lineage. ([source](https://docs.oracle.com/en-us/iaas/api/#/en/data-catalog/20190325/))
- dbt Integration: No — dbt is not a supported source and no manifest import is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/harvest-technical-metadata.htm))
- Python SDK: Yes — The pip-installable oci package includes a DataCatalogClient for all catalog operations. ([source](https://docs.oracle.com/en-us/iaas/tools/python/latest/api/data_catalog.html))
- Webhooks: Yes — Harvest job and custom property change events are emitted to OCI Events, which can route to HTTPS endpoints through OCI Notifications. ([source](https://docs.oracle.com/iaas/data-catalog/using/events.htm))

### Data Context Layer

- MCP Server: No — Oracle's MCP server repository has no Data Catalog server; only the generic OCI API server, which runs CLI commands. ([source](https://github.com/oracle/mcp))
- Knowledge Graph: No — Metadata is not modeled as a property graph. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Semantic Models: No — No semantic model generation. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Verified Queries: No — Data Catalog has no AI assistant or agent, so there are no curated queries for one to use. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/recom.htm))
- Semantic Layer Interop: No — Harvesting covers databases and files, enriched with glossary terms, tags and custom properties; no metric or semantic model import or export is documented. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/harvest-technical-metadata.htm))
- Unstructured Data Catalog: No — The Include Unrecognized Files option harvests files such as .pdf, .txt and .log from Object Storage, but only their names and paths, without content, classification or access control. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/harvest-technical-metadata.htm))

### Security

- SSO: Yes — OCI IAM identity domains federate with SAML 2.0 identity providers such as Microsoft Entra ID and Okta. ([source](https://docs.oracle.com/en-us/iaas/Content/Identity/federating/federating_section.htm))
- RBAC: Yes — IAM policies grant groups inspect, read, use or manage rights on catalogs, data assets, glossaries and metastores, down to a single glossary. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/policies.htm))
- Audit Log: Yes — OCI Audit logs every Data Catalog API call with the caller, time and request. ([source](https://docs.oracle.com/en-us/iaas/Content/Audit/Concepts/auditoverview.htm))
- Compliance Certifications: Yes — Data Catalog is available in OCI US Government Cloud with FedRAMP authorization (2021) and DISA Impact Level 5 (2022). ([source](https://docs.oracle.com/iaas/releasenotes/changes/5016b026-841a-4db8-a2cd-424828f76037/index.htm))
- SCIM Provisioning: Yes — OCI IAM identity domains accept SCIM 2.0 user and group provisioning from identity providers such as Okta and JumpCloud. ([source](https://docs.oracle.com/en/learn/oci-iam-jumpcloud-sso-scim/index.html))

### Deployment

- SaaS Deployment: Yes — Data Catalog is a fully managed, serverless OCI service. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))
- Self-Hosted Deployment: No — There is no self-hosted version; it runs only in Oracle's cloud regions. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/create-instance.htm))
- Hybrid Deployment: No — No on-premises agent; harvesting reaches on-premises sources over private endpoints connected to a VCN. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/pe-create.htm))
- Data Residency: Yes — Each catalog instance is created in a chosen OCI region, with a limit of two instances per region. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

### Open Source

- Open Source: No — Data Catalog is proprietary. ([source](https://docs.oracle.com/en-us/iaas/data-catalog/using/overview.htm))

## Integrations

15 integrations, each sourced from the vendor docs.

- Data Integration: OCI Data Integration
- Data Platform: OCI Big Data Service, OCI Data Science
- Data Storage: Oracle Cloud Object Storage
- Data Transformation: Apache Spark, OCI Data Flow
- Data Warehouse: IBM DB2, Oracle DB
- Databases: Azure SQL Database, Microsoft SQL Server, MySQL, Oracle Autonomous Database, PostgreSQL
- Query Engine: Hive SQL
- Streaming: Apache Kafka

## Alternatives

- [AWS Glue Data Catalog](https://toolsfordata.com/tools/aws-glue-data-catalog/): The AWS equivalent: a serverless metastore for Iceberg and Hive, free for the first million objects.
- [Google Cloud Knowledge Catalog](https://toolsfordata.com/tools/google-cloud-knowledge-catalog/): The Google Cloud equivalent, included with Google Cloud and priced by usage.
- [Databricks Unity Catalog](https://toolsfordata.com/tools/databricks-unity-catalog/): The Databricks equivalent, if your data lives on Databricks.
- [Microsoft Purview](https://toolsfordata.com/tools/microsoft-purview/): The Microsoft equivalent, with compliance-grade governance.
- [Atlan](https://toolsfordata.com/tools/atlan/): Consider Atlan for a managed catalog that covers Oracle and other clouds.
- [OpenMetadata](https://toolsfordata.com/tools/openmetadata/): Consider OpenMetadata for an open-source catalog with lineage, governance and data quality tests built in.
