I’ve no problem with using an SQLite database for this, however the ever-growning file is creating various problems for me [well, my staff & students]
What’s the current method for configuring the file in a different location?
Is there a something.SQLiteYStore.db_path = '/some/path/juypter_ystore.db' configuration option, or do we need to subclass ala Berkeley [Merge pull request #4269 from ryanlovett/stat159-ystore-class · berkeley-dsep-infra/datahub@01b59ca · GitHub]
Yes, SQLiteYStore.db_path is configurable, so placing it in a jupyter_server_config.json or jupyter_server_config.py file in an appropriate location (Configuring a Jupyter Server — Jupyter Server documentation) or passing it directly via CLI should work:
jupyter lab --SQLiteYStore.db_path = '/some/path/juypter_ystore.db'
however the ever-growning file is creating various problems for me [well, my staff & students]
@perllaghu I am trying to fix it right now with my colleagues and would love to get more traction, feedback and basically pick some brains on the solutions we are exploring, your feedback or just a 
would be appreciated on:
As another mitigation for frequent writes of the large database file we are looking for ways to respect the autosave preference of users (i.e. potentially disable it if user wishes to), and this turned out to be more controversial, but the more feedback the better:
Again, I would really appreciate any feedback on any of these!