In Amazon WS yes, because you can mount S3 as a folder disk, but seems this is not your case and you have bought some kind of hardware. For it should be necessary create a new Datastorage class.
Can you share with us the paper of this hardware. I have read something in the past about Object storage, where documents have it's own UUID ( not path ) and seems might be useful for scalability ,anyway are other solutions for large disk (
https://cloudian.com/blog/object-storag ... e-storage/ ), but not much else, really do you have a PB scenario ?
For what I read in the past you must access from API. In terms of a DMS have some disadvantages does not watch the file directly, because for all the operation must make an API call, meanwhile in the disk we have the path ready to be used.
OpenKM have a configuration to use database as a repository, in this case should be something similar writing a news class for working with this API.