We need to finalise how we want to handle the post-processed data.
My preferred option would be to delete the raw data (which is available on CASDA if needed) and store the post-processed data in /data/VAST/vast-data-post-processed/. This would then be symlinked into /data/VAST/release/ using the validation pipeline release command.