Custom content objects are not being imported during server to cloud migrations. Custom content objects have no specific prefix in server, but a prefix of 'ac:' in cloud. For example, a server app may have a custom content object app-foo-document-for-confluence:document but the cloud version of the app will have a custom content object ac:app-foo-document-for-confluence:document.
When exporting spaces on server and importing them into cloud (which is what the migration assistant does by default), the ac: prefix is not added to the custom content objects. As a result, the documents are not picked up by the APIs and therefore not available in the cloud version of the app.
Atlassian should take care of adding this prefix as part of the Migration assistant task.
- Install Scroll Documents on Confluence server
- Create a document in any space from the Document Library
- Migrate the space to a cloud instance using the Atlassian Migration Assistant
The Document should be available in the Document Library of the space on Cloud.
The Document is not available in the Document Library on Cloud.
The following workaround exists, but is not scalable:
- Create a space export from server that has a Scroll Document created
- Unzip the space export zip file
- In the entities.xml file manually add the prefix to all occurrences of the custom content instances in the CustomContentEntityObject object
- Repackage the space export archive
- Import to Cloud
Although the fix for this issue will migrate custom app CONTENT data, it does not resolve any permission issues that may be required for the App via Connect. Please watch MIG-303: Add-on users are not migrated with space permissions for updates for a fix for the permissions issue.