Do you really mean
data warehouse or something else?
About the lowest level of cloud computing resources is BLOB storage, more specifically "block BLOB storage." This provides programs with remote data stored as chunks of bytes within contianers. You can think of the BLOBs or chunks as "files" since they might contain a Wrord document, an image file, binary data, etc. The containers are similar in concept to "directories" or "folders."
Block BLOB storage often allows ranges of bytes to be read or changed, but normally programs would transfer entire BLOBs.
One of the oldest forms of this is WebDAV, an early "web service" protocol. There are WebDAV client redirectors that even allow a WebDAV root to be mounted as a network share, mapped to a Windows drive letter, etc. There are also WebDAV APIs for programmatic use that bypass such (often a little flaky) redirectors.
One easy to use WebDAV host I like is
MyDrive. They offer an access-control model based on an admin user and sub-users (guests) with their own passwords. The admin can set guests' access rights to "folders" and these impact the contents of the container ("folder"). You can allow writes/uploads on one, set read-only access on another, etc.
They offer a basic free account level with limited space and one "guest" sub-user account.
Interaction can be through MyDrive's Web user interface or through a WebDAV redirector or client API.
Beyond WebDAV there are other block BLOB and page BLOB storage services from many cloud computing vendors. These tend to use more proprietary APIs and protocols though, making a vendor switch more effort than it needs to be.
Of course there are also more structured offerings out there, from key-value datastores to queues to full RDBMSs.
