The safest way to handle renditions and derivatives is to make the work restartable, traceable, and testable before production cutover. The accountable owner should produce rendition lineage table; restartable export and accept the result only when approved outputs regenerate from the intended master.
This guide treats originals, crops, previews, and transcodes as part of the migration contract. The problem is not complete when bytes arrive; it is complete when identity, metadata, permissions, versions, relationships, and delivery behavior meet documented acceptance rules.
Make extraction restartable
Partition renditions and derivatives into deterministic batches using a stable cursor or manifest. Each row should include the source identity, requested operation, attempt number, response, destination identity when created, byte count, checksum, and final disposition. A completed batch can be replayed safely: already accepted items are verified or skipped, never duplicated.
Separate binary transfer from metadata assignment when that makes recovery clearer. Limit concurrency deliberately, observe latency and rate responses, and use exponential backoff with jitter for transient failures. Authentication failures, invalid metadata, unsupported files, and missing source objects are permanent until a person or rule changes something; endless retry only hides the defect.
The exception contract
Every exception needs a reason code, raw response, first and latest attempt, current owner, next action, and deadline. Preserve the source artifact until the destination record passes validation. The operating metric is masters identified and required outputs reproducible; report accepted, skipped, retried, quarantined, and failed items separately.
Finish the run with rendition lineage table; restartable export. Then select random and risk-based samples and prove approved outputs regenerate from the intended master. This closes the gap between an API returning success and the migrated capability working for users.
Control batches, checkpoints, and replay
Give every batch a deterministic identity and immutable manifest. Checkpoint after the destination response, not before it, and verify an existing result before skipping it. A safe replay produces the same destination state while adding a new audit event; it never creates a second asset merely because a worker restarted.
For renditions and derivatives, the media platform owner should review masters identified and required outputs reproducible by migration wave and risk class. The review must connect each exception to originals, crops, previews, and transcodes, the controlling rule, and a disposition. This is how the team prevents mistaking generated outputs for masters from being hidden inside an aggregate completion number.
Record the decision in rendition lineage table; restartable export. Include who approved it, which evidence was reviewed, what remains open, and the trigger for another review. The release gate stays closed until approved outputs regenerate from the intended master. If the threshold is not met, reduce the wave, repair the rule, and repeat the same query so the improvement is measurable rather than anecdotal.
Before approval, ask a reviewer who did not build the migration rule to reproduce one ordinary case, one restricted or high-risk case, and one known exception. Store those results with the batch identity. Independent reproduction catches hidden assumptions in credentials, local files, cached responses, and operator knowledge. It also leaves the next team a usable diagnostic path instead of a success note that cannot be repeated.
A minimum evidence pack
Keep the source inventory and query, immutable batch manifests, mapping or policy version, API responses, exception ledger, reconciliation output, access-test results, downstream screenshots or probes, and the signed go/no-go decision. Use timestamps and stable IDs so an auditor can connect each destination record to the exact migration run.
Avoid three shortcuts: treating a successful request as business acceptance, averaging away high-risk failures, and deleting the source before rollback and restore paths are proven. These shortcuts make a migration look fast while moving the risk into production.
Related DAM migration guides
- 7 Best DAM Migration Tools: Cloudinary Leads the Shortlist
- Map originals, crops, previews, and transcodes Into the New DAM Without Losing mistaking generated outputs for masters
- Design a Representative Pilot for renditions and derivatives
Primary implementation reference: cloudinary.com documentation.