I noticed that to get filesystem to S3 migration work, file:// prefix in dvobject.storageidentifier needs to be replaced to configured storage ID instead of s3://as currently instructed in https://guides.dataverse.org/en/latest/developers/deployment.html . So, the protocol part in dvobject.storageidentifierwill look like <storage_id>://
We are using CEPH S3 instead of Amazon S3, so perhaps this is related to that (?). Dataverse version is 6.7.1.
Hmm. It looks like @Don Sizemore added those docs in #6789. Any comment, Don?
the prefix you're seeing isn't the protocol but the identifier of the datastore in use. you're essentially copying the files outside of Dataverse, then updating the location pointer within the database.
Would it make sense to change the example SQL update queries so they donβt use s3:// as the datastore identifier, and use something like <datastore_identifier>:// instead?
absolutely. I think what you're seeing is that, in the early days, everybody called theirs s3 :smile:
@mjlassila great idea. If you'd like to make a pull request, I'm happy to help!
I made a pull request, I'm happy to tweak it more if it needs improvement: https://github.com/IQSS/dataverse/pull/11795
Looks great! Merged! Thanks, @mjlassila! ![]()
Last updated: Nov 01 2025 at 14:11 UTC