Hello,
We’ve run into a data management issue during the transfer of data from an old server to a new one. The old cluster operated on cryoSPARC v3.2.0, while the new cluster is running v4.4.1.
Following the transfer of some datasets using the command ‘rsync -avrP old_path new_path’, we successfully attached the old project to the new cryoSPARC. Subsequently, as each project finished transferring, we began cleaning up the dataset from the old server, including deleting the project in the old cryoSPARC. However, we overlooked that cleaning up the old dataset would generate new management files, potentially overwriting the old (and intact) ones in the new server, as we repeated rsync command due to the intermittent internet connectivity.
Upon comparison between “broken projects” and “intact projects,” we observed changes not only in the workspaces.json file but also in all job.json files within each Job folder, presumably due to our operational error. Consequently, it seems impossible for us to restore the project in the new cryoSPARC:
We attempted to replicate a workspaces.json file, but after attachment, all workspaces appeared empty, likely because each job.json is corrupted. Attempts to import individual jobs or .csg result files also failed, as they require an export step which we cannot fulfill since the old dataset has been deleted.
We believe that all real data files remain complete, but accessing them is hindered by the collapse of management files. For these datasets, we only require access to specific information (e.g., the remaining particle number after 2D classification, the representative 2D classification results). Is there any possible methods to retrieve this information given the circumstances?
Thanks for your assistance,
Xiaoxiao