I have a portable drive and several workstations.
Each workstation is set up to sync folders from my portable drive to the Seafile server.
This setup seems working but has one tradeoff: when I change data on some workstation, sync changes and move my drive to another one, the same changes are synced from that second workstation too, and in the end I have several identical versions of files on the server.
As I understand this happens because each seafile client keeps its own file index and indexes are kept outside of data directories. And when a client wants to sync data, the only source of information on what to sync is its local file index.
Is there any way to avoid data duplication when syncing the same portable drive from different workstations?
Because data is being deduplicated this normally should not happen (even if data is being uploaded twice).
Yes, I understand that deduplication saves space. But files need to be uploaded many times and such uploads are also reflected in history which is inconvenient.
So as I understand, there is no workaround for the moment.
Hope in future releases Seafile developers will work out a solution for portable drives.
Thank you for the reply anyways.
Have you verified this? The client should check via head request whether the block is being available on the server first and only upload it if this is not the case.
No I have not (don’t know how).
But the thing I noticed is when I move the drive that’s already been synced to other workstation, the Seafile client starts sync process and reports about successful sync of the files that have already been synced from the first workstation. And history on the server is populated with records about such identical copies from different workstations.
If I see that records in history does it mean that file transfer took place or only history record was created?