Missing Chunks, Manually Locating Chunks Show Duplicate Chunk Files?

[gabrielramos]

Hello, I’m a bit confused with how recovering from a Missing Chunks error is performed. I’ve been browsing around the form and have also read Fix missing chunks. There was also mention about changing the ID of a repository?

When I tried locating the some of the files it shows duplicates, one with a file size of 0 and the other with varying file sizes. Would deleting those 0 size files resolve the error? And Im not sure if it’d be practical deleting them manually as there are 100+ of these missing chunks.

I have 3 backup locations; a local NAS, Wasabi, and B2. The Missing Chunks are only occurring on my B2 bucket.

Daily Storage Check Schedule:

This is from the B2 check log:

Here are a couple of those duplicates:

This is my entire Schedule:

[saspus]

Likely missing chunks are from ghost snapshots left behind as a result of failed/interrupted prune.

All you need to do is find them and delete manually from the storage.

1. Run check -persist
2. Parse output to get a list of bad snapshots
3. Delete those from snapshots folder on the b2 target
4. Clear local cache.

This is all trivially scriptable.

Until the prune bug is fixed this is unavoidable. Another solution going forward is to not prune in the first place.

Duplicacy should have a built-in healing facility that would do just that.