Datazone Diagnostics

DataZone Diagnostics

When to use

Any Amazon DataZone investigation — domain creation, projects, data assets, subscriptions, governance, business catalog, or data quality.

Investigation workflow

Step 1 — Collect and triage

aws datazone list-domains --max-results 20
aws datazone get-domain --identifier <domain-id>
aws datazone list-projects --domain-identifier <domain-id> --max-results 20

Step 2 — Domain deep dive

aws datazone get-project --domain-identifier <domain-id> --identifier <project-id>
aws datazone list-environments --domain-identifier <domain-id> --project-identifier <project-id>
aws datazone search --domain-identifier <domain-id> --search-scope ASSET --search-text "*"
aws datazone list-subscriptions --domain-identifier <domain-id> --max-results 20

Step 3 — Detailed investigation

aws datazone get-asset --domain-identifier <domain-id> --identifier <asset-id>
aws datazone get-subscription --domain-identifier <domain-id> --identifier <sub-id>
aws datazone get-environment --domain-identifier <domain-id> --identifier <env-id>
aws cloudtrail lookup-events --lookup-attributes AttributeKey=EventSource,AttributeValue=datazone.amazonaws.com --max-results 10

Read references/guardrails.md before concluding on any DataZone issue.

Gotchas: DataZone

• DataZone is a data governance and catalog service, not a data processing service.
• Domains are the top-level organizational unit containing projects and assets.
• Projects contain environments where data producers and consumers work.
• Data assets are metadata entries in the catalog — they reference data, not store it.
• Subscriptions grant access to data assets between projects.
• Environments connect to data sources (Glue, Redshift, Athena) for actual data access.
• Business glossary provides standardized terminology across the organization.

Anti-hallucination rules

1. DataZone manages metadata and governance — it doesn't store or process data.
2. Domains must be created before any other DataZone resources.
3. Subscriptions require approval workflow — they're not instant access grants.
4. Environments connect to actual data sources — DataZone doesn't replace them.
5. Data quality rules are defined in DataZone but executed by underlying services.
6. Spend no more than 2 minutes on any single hypothesis. Pivot if inconclusive.

14 runbooks