WENDELIN combines Scikit Learn machine learning and NEO distributed storage for out-of-core data analytics in python
There is no concept of "global storage" in Rapid.Space. Storage is considered as local to each compute node.
It is the responsibility of the developer to deploy distributed storage services on each VPS or VM.
It is also possible also to implement "storage services" as custom buildout profiles that can be deployed on some Rapid.Space servers. This is how Rapid.Space "Big Data" service is implemented based on NEO distributed transactional object storage.