Context: Some simulations occupy tens or hundreds of GB (e.g. JOREK cases), and store vast amounts of data across several IDSs. It happens quite regularly that some particular field of a given IDS needs to be corrected (e.g. a sign correction in the definition of one quantity).
Problem: At the moment, it seems that one needs to re-upload the full data into a new simulation entry to do such corrections. This creates unnecessarily a huge duplication of memory, and users may hesistate reuploading their full data because of this reason. It would also be good to be able to correct the metadata without touching the uploaded files.
Proposed solution: If the uploaded data is identical to the "replaced" simulation, re-use the old simulation data and only store the difference.
Context: Some simulations occupy tens or hundreds of GB (e.g. JOREK cases), and store vast amounts of data across several IDSs. It happens quite regularly that some particular field of a given IDS needs to be corrected (e.g. a sign correction in the definition of one quantity).
Problem: At the moment, it seems that one needs to re-upload the full data into a new simulation entry to do such corrections. This creates unnecessarily a huge duplication of memory, and users may hesistate reuploading their full data because of this reason. It would also be good to be able to correct the metadata without touching the uploaded files.
Proposed solution: If the uploaded data is identical to the "replaced" simulation, re-use the old simulation data and only store the difference.