Querying AWS Sagemaker Catalog

Query AWS SageMaker Catalog System Tables

Overview

Works best with the AWS MCP server (opens in a new tab) for sandboxed execution and audit logging. All commands below use the AWS CLI and work in any environment with configured AWS credentials.

Amazon SageMaker Unified Studio (whose catalog feature is referred to below as SageMaker Catalog) exports asset metadata as a daily-snapshot Apache Iceberg table in the AWS-managed aws-sagemaker-catalog table bucket. This enables SQL queries over your entire data catalog inventory — asset counts, governance gaps, ownership audits, and historical comparisons — without building custom ETL.

Data is partitioned by snapshot_time and exported once daily (around midnight per region). The table is read-only.

Decision Tree

Common Tasks

1. Check If Configured

aws datazone get-data-export-configuration \
  --domain-identifier <DOMAIN_ID> \
  --region <REGION>

• If no domain exists: aws datazone list-domains --region <REGION>
• If export not enabled: guide user to enable.
• One domain per account per region.

Verify table bucket exists:

aws s3tables list-table-buckets --region <REGION> \
  --query "tableBuckets[?name=='aws-sagemaker-catalog']"

2. Enable

With KMS encryption (recommended for production):

aws datazone put-data-export-configuration \
  --domain-identifier <DOMAIN_ID> \
  --region <REGION> \
  --enable-export \
  --encryption-configuration kmsKeyArn=<KMS_KEY_ARN>,sseAlgorithm=aws:kms

Note: Encryption cannot be changed after creation. Always specify KMS for sensitive catalog data.

Without encryption (for quick testing only):

aws datazone put-data-export-configuration \
  --domain-identifier <DOMAIN_ID> \
  --region <REGION> \
  --enable-export

First data available within 24 hours. See: Exporting asset metadata (opens in a new tab)

3. Verify Permissions for Querying

Requires:

• S3 Tables federated catalog registered in Glue (s3tablescatalog)
• Lake Formation SELECT + DESCRIBE grants on the table

Grant access:

aws lakeformation grant-permissions \
  --principal DataLakePrincipalIdentifier=<ROLE_ARN> \
  --resource '{"Table": {"CatalogId": "<ACCOUNT>:s3tablescatalog/aws-sagemaker-catalog", "DatabaseName": "asset_metadata", "Name": "asset"}}' \
  --permissions DESCRIBE SELECT \
  --region <REGION>

4. Query

Query syntax:

"s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"

Constraints:

• You MUST always filter by snapshot_time — without it, the query scans all historical snapshots and returns duplicates

• You MUST confirm workgroup and output location before executing

• Default to DATE(snapshot_time) = CURRENT_DATE for current state

• You SHOULD use the key columns documented in this skill to build queries. If you need the full schema, run get-tables once:

  Copy

  aws glue get-tables --catalog-id "<ACCOUNT>:s3tablescatalog/aws-sagemaker-catalog" --database-name "asset_metadata" --region <REGION>

Key columns:

Current catalog state:

SELECT resource_type_enum, COUNT(*) as count
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"
WHERE DATE(snapshot_time) = CURRENT_DATE
GROUP BY resource_type_enum
ORDER BY count DESC;

Assets without business descriptions:

SELECT asset_name, resource_name, resource_type_enum, account_id
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"
WHERE DATE(snapshot_time) = CURRENT_DATE
  AND business_description IS NULL;

Asset growth over last 30 days:

SELECT DATE(snapshot_time) as date, COUNT(*) as total_assets
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"
WHERE DATE(snapshot_time) >= CURRENT_DATE - INTERVAL '30' DAY
GROUP BY DATE(snapshot_time)
ORDER BY date DESC;

Time travel — compare current vs 7 days ago (new descriptions added):

SELECT t.asset_id, t.resource_name,
       p.business_description as before,
       t.business_description as now
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset" t
JOIN "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata".

Assets by owner:

SELECT extended_metadata['owningEntityId'] as owner, COUNT(*) as count
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"
WHERE DATE(snapshot_time) = CURRENT_DATE
  AND extended_metadata['owningEntityId'] IS NOT NULL
GROUP BY extended_metadata['owningEntityId']
ORDER BY count DESC;

Filter by metadata form field:

SELECT *
FROM "s3tablescatalog/aws-sagemaker-catalog"."asset_metadata"."asset"
WHERE DATE(snapshot_time) = CURRENT_DATE
  AND extended_metadata['<metadata-form-name>.<field-name>'] = '<field-value>';

Key Behaviors

• Daily snapshots — exported around midnight per region
• Always filter by snapshot_time — without it you get all history (duplicates, slow)
• One domain per account per region — to switch domains, delete config first
• No additional charge beyond S3 Tables storage + Athena queries
• Read-only — to update asset metadata, use Glue Discovery APIs or SageMaker Unified Studio

Troubleshooting

Security Considerations

Data sensitivity: Catalog metadata exposes organizational structure including asset names, ownership, account IDs, naming conventions, and internal resource identifiers. Treat query results as sensitive by default.

Encryption at rest: Always enable KMS encryption when creating the export configuration. Encryption cannot be changed after creation. Additionally, configure SSE-KMS on your Athena workgroup output bucket.

Least-privilege access: Grant Lake Formation SELECT + DESCRIBE only on the specific asset_metadata.asset table to roles that need catalog analytics. Avoid granting access to the entire aws-sagemaker-catalog bucket.

Audit trail: Enable CloudTrail logging for DataZone (PutDataExportConfiguration, GetDataExportConfiguration), Athena (StartQueryExecution, GetQueryResults), and S3 Tables API calls to track who queries catalog metadata.

Credential hygiene: Use IAM roles with temporary credentials for querying. Avoid long-lived access keys for users accessing catalog metadata. Scope down or rotate principals when access is no longer needed.

Additional Resources

• Exporting asset metadata (opens in a new tab)
• Asset table schema (opens in a new tab)
• Integrating S3 Tables with analytics services (opens in a new tab)
• Lake Formation permissions (opens in a new tab)