How GRIT storage and automounts work

 

Storage of data is saved in manageable bite-sized pieces for easier backup and relocation. These are typically done in this hierarchy:

• Research Group
  • Users
    • user1
    • user2
    • etc.
  • Projects
    • project1
    • project2
    • etc.
  • Data
    • dataset1
    • dataset2
    • etc.

The deepest level (ex. user1 or project1) are filesystems (meaning they are only navigable down from that level.  You cannot navigate from project 1 up to the main Projects folder.  This is true with the Users and Data folder as well.

The deepest levels of this tree are made into automounts (essentially shortcuts) that are accessible from any server on the GRIT ecosystem, with path that looks like this: /home/user1   or  /home/project1  even though the actual path would be something like:
servername.grit.ucsb.edu/storage-area/backup-level/group-name/Users/user1

We do not create access to the User, Project, or Data level because these can grow to unmanageable sizes and make relocating or backing up unwieldy.