Schema Conversion Operations Reference

Command Summary

Command	Purpose	Async	Poll via	Waiter
start-metadata-model-import	Import metadata from source or target DB	Yes	`describe-metadata-model-imports`	`metadata-model-imported`
start-metadata-model-conversion	Convert source schema to target	Yes	`describe-metadata-model-conversions`	`metadata-model-converted`
start-metadata-model-assessment	Assess conversion complexity	Yes	`describe-metadata-model-assessments`	`metadata-model-assessed`
start-metadata-model-creation	Create statement-based model	Yes	`describe-metadata-model-creations`	`metadata-model-created`
start-metadata-model-export-as-script	Export DDL to S3	Yes	`describe-metadata-model-exports-as-script`	`metadata-model-exported-as-script`
start-metadata-model-export-to-target	Apply converted DDL to target DB	Yes	`describe-metadata-model-exports-to-target`	`metadata-model-exported-to-target`
export-metadata-model-assessment	Generate PDF/CSV report to S3	Sync	N/A	sync
describe-metadata-model-children	Navigate tree structure	Sync	N/A	sync
describe-metadata-model	Get object definition (DDL)	Sync	N/A	sync
cancel-metadata-model-conversion	Cancel running conversion	Yes*	N/A	`metadata-model-conversion-cancelled`
cancel-metadata-model-creation	Cancel running creation	Yes*	N/A	`metadata-model-creation-cancelled`

* The cancel API call returns synchronously, but the cancellation state transition is async (CANCEL_RECEIVED → CANCELING → CANCELED). Use the corresponding waiter to confirm the operation reached the CANCELED state.

Note: All waiters require --filter 'Name=schema-conversion-operation-id,Values=<RequestId>' in addition to --migration-project-identifier.

For full parameter details and examples, refer to the linked CLI documentation for each command.

DMS Waiters

Use DMS waiters to wait for async operations to complete:

Operation	Waiter command
Import	`aws dms wait metadata-model-imported`
Conversion	`aws dms wait metadata-model-converted`
Assessment	`aws dms wait metadata-model-assessed`
Creation (statement)	`aws dms wait metadata-model-created`
Export as script	`aws dms wait metadata-model-exported-as-script`
Export to target	`aws dms wait metadata-model-exported-to-target`
Cancel conversion	`aws dms wait metadata-model-conversion-cancelled`
Cancel creation	`aws dms wait metadata-model-creation-cancelled`
Extension pack	`aws dms wait extension-pack-associated`

Documentation: https://docs.aws.amazon.com/cli/latest/reference/dms/wait/index.html

Operation Statuses

All async operations use the same status values:

Status	Meaning
`RECEIVED`	Request received, queued
`IN_PROGRESS`	Operation is running
`SUCCESS`	Operation completed successfully
`FAILED`	Operation failed — check error details
`CANCEL_RECEIVED`	Cancellation request received
`CANCELING`	Cancellation in progress
`CANCELED`	Operation was cancelled
`RETRY`	Operation is being retried
`PENDING`	Operation is pending

Execution Pattern

Call start-* → extract RequestIdentifier
Wait for completion using the corresponding aws dms wait command (see DMS Waiters table)
If the waiter returns successfully → proceed to next step
If the waiter fails or is unavailable (e.g., Invalid choice due to outdated CLI) → fall back to manual polling with the corresponding describe-* command until status reaches SUCCESS or FAILED

Fallback polling pattern (use when waiter is not available):

aws dms describe-metadata-model-<operation>s \
  --migration-project-identifier <migration_project_identifier> \
  --filter Name=request-id,Values=<RequestIdentifier>

Check Requests[0].Status — repeat every 30 seconds until it reaches SUCCESS or FAILED.

Constraint: You MUST NOT proceed to the next step until the operation has completed. If the waiter fails or times out, you MUST fall back to polling with the describe command. Never assume an operation succeeded without confirming its status.

Key Notes

describe-metadata-model returns TargetMetadataModels with target SelectionRules — use these to query the TARGET tree (do NOT reuse source selection rules for target)
describe-metadata-model-children returns MetadataModelChildren[] with MetadataModelName and SelectionRules — use the child's SelectionRules to drill deeper
export-metadata-model-assessment is synchronous — returns S3 links immediately (PdfReport.S3ObjectKey, CsvReport.S3ObjectKey)
start-metadata-model-creation only supports SQL Server → PostgreSQL/Aurora PostgreSQL

/dms-schema-conversion

referencesschema-conversion-operations.md