Atlan Lakehouse
Changes to Atlan Lakehouse—namespaces, schema and API behavior, query engine integrations, and setup requirements.
Gold namespace available natively in the catalog
The Gold namespace is now built directly into the Lakehouse catalog. Pre-materialized tables in the gold namespace are kept in sync with the underlying metadata in real time—no customer-managed compute, refresh schedules, or setup scripts required. A new BI_ASSET_DETAILS table has been added for PowerBI, Tableau, Looker, and Sigma assets. Breaking changes: the LINEAGE, TAGS, CUSTOM_METADATA, and README tables have been removed from the Gold namespace. For tags, custom metadata, readmes, and lineage, query the raw entity_metadata tables directly (TagRelationship, CustomMetadata, Readme, Process). Additional column-level changes include: TAGS column removed from ASSETS, DESCRIPTION split into DESCRIPTION and USER_DESCRIPTION, SCHEMA_DATABASE_NAME renamed to DATABASE_NAME, PROCEDURE_SCHEMA renamed to SCHEMA_NAME, MC_MONITOR_RULE_LAST_EXECUTION_TIME renamed to MC_MONITOR_RULE_LAST_EXECUTION_AT, and 13 common columns removed from GLOSSARY_DETAILS (use ASSETS table instead).
Network requirements for private tenants documented
If your Atlan tenant uses private networking or IP allowlists, your query engine's egress IPs must be added to the tenant's load-balancer allowlist before connecting to Lakehouse. Open a Support ticket with your egress IPs or NAT range to get started. Prerequisites for Snowflake, Athena, and PySpark setup guides now link to this requirement.
AI coding agent skill available
The atlan-lakehouse Agent Skill lets AI coding agents like Claude Code query Lakehouse using natural language. Ask about metadata completeness, lineage gaps, glossary coverage, or user adoption and the agent writes and runs the SQL for you. The skill auto-detects your platform (Snowflake, Databricks, or Python) and adapts queries accordingly.
Credential rotation supported
Atlan can now rotate the OAuth credentials that your query engine uses to authenticate to the Lakehouse catalog. Contact Atlan Support to request a one-time rotation or set up a recurring schedule. After rotation, update your query engine connection with the new credentials from the Lakehouse setup page.
observability namespace available
Lakehouse now exposes workflow and job execution metrics in a new namespace. Track data quality scores over time, monitor job success and failure rates, analyze retry patterns, and measure job duration across Lakehouse pipelines using standard SQL. The job_metrics table includes a flexible custom_metrics JSON column with workflow-specific metrics that vary by job type.
Lakehouse Solutions repository published
The Lakehouse Solutions public repository contains production-ready scripts and tools for extending Lakehouse—including Databricks and BigQuery setup scripts and table maintenance utilities. Use these as starting points for platforms that don't yet support federated Iceberg REST catalogs natively.
Amazon Athena integration available
Connect Amazon Athena to Lakehouse using AWS Glue Catalog Federation. Lakehouse tables are exposed through the AWS Glue Data Catalog, making them queryable from any Athena-compatible engine without additional infrastructure.
entity_metadata tables are per-asset-type
Each asset type has its own table in the entity_metadata namespace. Querying a table for an asset type with no assets returns zero rows—this is expected behavior, not an error.
Snowflake setup: specific privileges required
ACCOUNTADMIN isn't the only valid role. The specific privileges required are CREATE INTEGRATION and CREATE DATABASE. Grant commands for custom roles are now documented, along with a troubleshooting note for the OAuthTokenResponse error.
Snowflake on GCP: table initialization fix documented
Snowflake users on GCP-hosted tenants hitting a SQL compilation error when querying Lakehouse Iceberg tables stored on GCS can now follow a step-by-step guide to configure the required GCS external volume and Polaris catalog integration.
usage_analytics namespace available
Lakehouse now exposes product telemetry from the Atlan UI in a new namespace—including page views, user actions, and identity snapshots. Query active users, measure feature adoption, identify dormant groups, and join usage data with organizational data using standard SQL.
Lakehouse available on App Marketplace
Atlan Lakehouse is now activatable directly from the App Marketplace as an Apache Iceberg-based data lakehouse. Query Atlan metadata—technical, business, governance, and operational signals—through any Iceberg REST-compatible engine (Snowflake, Databricks, Athena) without custom export pipelines.