`duplicacy check -all -fossils -resurrect -persist` not skipping errors

Bug
#1

Please describe what you are doing to trigger the bug:
Running duplicacy check -all -fossils -resurrect -persist

Please describe what you expect to happen (but doesn’t):
According to the man page: -persist continues even if errors occur

Please describe what actually happens (the wrong behaviour):

Chunk 770cb52b491ec57a08cf2e2c84755a8e0b6bbaa1ccd82d1970baabbaddae0718 can't be found
Failed to load chunks for snapshot laptop at revision 46: unexpected end of JSON input

CLI version 2.7.2

#2

That chunk is likely a metadata chunk that stores the list of referenced chunks for revision 46. Without this chunk Duplicacy doesn’t know which chunks revision 46 needs so it bails out even with the -persist option.

Can you follow this guide to figure out why this chunk is missing?

#3

Thank you for your quick reply. I went through the guide, it doesn’t exist in the repo, and I don’t have the prune logs anymore. Would be great if the logs where also stored on the server.

Now when I try to duplicacy prune -id laptop -r 46 it fails with the same error. I understand why it’s happening – it’s the same problem, that the metadata chunk is missing, but for prune, prune should then fall back to deleting the revision so that the chunks can be fossil collected?

Just found the solution: Can't recover from Chunk not found error during prune

1 Like