Using a production clone for AEM Assets migration analysis with Content Transfer Tool
Running migration analysis directly on the production AEM author instance can affect performance when planning a migration to AEM as a Cloud Service using the Content Transfer Tool. Use a clone of the production environment for migration analysis while considering the limitations of using a clone for delta (top-up) extractions.
Description description
Environment
- Adobe Experience Manager (AEM) on-premises as the source
- AEM as a Cloud Service as the target
- Shared datastore (for example, S3)
- Large repository (millions of assets, tens of TB)
Issue/Symptoms
When migrating large-scale AEM Assets repositories from on-premises to AEM as a Cloud Service, running migration analysis directly on the production author instance can impact performance. Using a clone of the production environment for analysis tasks can help reduce this impact, especially for large repositories with a shared datastore.
- Production author instance performance is degraded or sensitive to additional load.
- Migration set size, asset counts, versions, renditions, and repository structure need to be analyzed without impacting production.
Cause
Migration analysis and extraction processes can be resource-intensive and may affect production author performance. Using a clone can offload analysis tasks, but there are important limitations for delta extraction workflows.
Resolution resolution
Follow the steps below to resolve the issue:
-
Prepare a clone of your production author environment for migration analysis:
- Ensure the clone uses its own node/segment store and does not connect to the live production node store.
- Confirm that the clone and production have different repository identities. This can be checked in the AEM Web Console under
/system/console/topology. - Do not run BlobGC, datastore cleanup, or any process on the clone that could delete objects from the shared datastore.
- If using a shared datastore (e.g., S3), configure the clone with read-only permissions to reduce risk, but always validate that the clone starts and operates correctly with these permissions.
-
Use the clone for migration analysis only:
- Run migration set sizing, asset counts, version and rendition analysis, and repository structure review on the clone.
- Be aware that the analysis reflects the state of production at the time the clone was created; any changes made in production after cloning will not be included.
-
For actual migration and delta (top-up) extractions:
- Run the Content Transfer Tool (CTT) migration and all top-up extractions from the live production author instance, not from the clone.
- CTT’s delta extraction relies on checkpoints in the source repository. Recreating or refreshing a clone resets these checkpoints, making it unsuitable for delta extraction from multiple clones.
- Schedule migration and top-up extractions from production during low-usage periods to minimize performance impact.
-
Verification:
- After configuring the clone, verify that it starts successfully, does not interfere with production, and that analysis tasks complete as expected.