Supplychain Diagnostics

AWS Supply Chain Diagnostics

When to use

Any AWS Supply Chain investigation — instance creation, data integration, demand planning, supply planning, inventory visibility, insights generation, connector configuration, or S3 integration issues.

Investigation workflow

Step 1 — Collect and triage

aws scn list-instances --query 'instances[*].{Id:instanceId,Name:instanceName,State:state}'
aws scn get-instance --instance-id <instance-id>
aws scn list-data-integration-flows --instance-id <instance-id>

Step 2 — Domain deep dive

aws scn get-data-integration-flow --instance-id <instance-id> --name <flow-name>
aws scn get-data-lake-dataset --instance-id <instance-id> --namespace <namespace> --name <dataset-name>
aws scn get-bill-of-materials-import-job --instance-id <instance-id> --job-id <job-id>

Step 3 — Detailed investigation

aws cloudtrail lookup-events --lookup-attributes AttributeKey=EventSource,AttributeValue=scn.amazonaws.com --max-results 20
aws scn list-data-lake-datasets --instance-id <instance-id> --namespace <namespace>
aws s3 ls s3://<supply-chain-bucket>/<prefix>/

Read references/guardrails.md before concluding on any Supply Chain issue.

Tool quick reference

Gotchas: AWS Supply Chain

• Supply Chain instances are the top-level resource. All operations (data integration, planning, visibility) are scoped to an instance.
• Data integration flows connect external data sources to the Supply Chain data lake. Flows can fail silently — always check flow execution status.
• The data lake uses predefined namespaces and schemas. Custom data must conform to the expected schema or transformation is required.
• Demand planning and supply planning require sufficient historical data. Models need weeks to months of data before producing accurate forecasts.
• Inventory visibility aggregates data from multiple sources. Discrepancies often indicate data integration timing issues, not bugs.
• Connectors (SAP, EDI, etc.) have specific configuration requirements. Connector failures often stem from credential or network issues.
• S3 is the primary data exchange mechanism. Bucket permissions, file formats, and folder structures must match Supply Chain expectations.

Anti-hallucination rules

1. Always cite specific instance IDs, flow names, or job IDs as evidence.
2. Data integration flows and data lake datasets are separate concepts. Never conflate them.
3. Demand planning requires historical data. Never claim instant forecast generation.
4. Inventory visibility aggregates data — discrepancies may be timing issues, not errors.
5. Spend no more than 2 minutes on any single hypothesis. Pivot if inconclusive.

12 runbooks