Currently, when a document is sent to cold storage, Nuxeo moves the binary from the standard S3 bucket to the S3 GLacier bucket.
Then, it is AWS who change the S3 storage class from "Standard" to "Glacier" thanks to the lifecycle rule (to be set into the bucket).
This system involves 2 things:
- We have to wait until 24 hours (minimum delay available on S3) before the binary is really on Glacier (so no retrieve and restore available)
- We are dependant of the S3 bucket configuration as the bucket has to be configured with a lifecycle rule to handle it (extra work for our cloud teams)
That would be great that Nuxeo forces the storage class on his own following the move to the glacier bucket.