Storing Data

In Notebook, PrimeHub provides several places of storing various data.

User Volume

Each user has own space called User Volume which is only accessible to the user to store data. The default volume capacity for everyone is 20 GB. The capacity is able to be enlarged, please contact Admin to learn the what capacity it is or to enlarge it if required. See System Settings.

In Notebook, it is home folder a.k.a /home/jovyan.

Group Volume

Each user must be associated with one group at least. If Shared Volume of a Group is enabled with a specified capacity by Admin, all of group members are able to access the space called Group Volume. please contact Admin to learn what shared capacity it is or to enable a group volume for a group. See Group Management

In Notebook which is launched under a group <GroupId>, users can see a mounted folder groupid under home folder. E.g. InfuseAICat -> infuseaicat/.

Dataset Volume

Admin is able to create a volume dedicated to a dataset via Dataset Management for users, the capacity is specified by Admin.

If one of user's groups is associated with a private dataset or global datasets, user can see the dataset folder from the Notebook which is launched under the same group.

The writable/read-only permission is varied with the combination of dataset type, global and associated groups. See Dataset - Group Access Control.

In Notebook, the dataset folders are under datasets/<dataset_name>. Please contact Admin to create dataset volumes for users with the proper permission.

PHFS Storage

PHFS Storage is based on PrimeHub Store technology. This storage is shared and accessed among the same group members, group members can share user data here. it seems similar to Group Volume, however, there are differences between PHFS and Group Volume.

In addition, PrimeHub features also store relative group-context data in the storage, such as Job stores artifacts under /phfs/jobArtifacts/. Since the limitation of the storage, we don't recommend storing performance-sensitive data such as datasets, please use Dataset Volume instead.

In Notebook, the storage is under /phfs.