This is a hopefully more coherent follow-up to an earlier post. Still struggling to retrieve files via s3cmd.

Has anyone else experienced this problem?

I can list the contents of a bucket using s3cmd but not recursively get all its contents (although the dry-run of this reports no errors).

I.e. I can traverse through the pseudo-directory structure in Cyberduck but can see certain keys as both file and DIR (e.g. 'assets' and 'assets/'). However still getting these weird 'ERROR: Skipping' messages.

> s3cmd lists
s3cmd ls s3://<bucket>/videos/
                       DIR   s3://<bucket>/videos/assets/
2009-11-24 03:51         0   s3://<bucket>/videos/assets

> s3cmd get -r  s3://<bucket>/videos/
s3://<bucket>/videos/assets -> ./assets  [1 of 267]
 0 of 0     0% in    0s     0.00 B/s  done
ERROR: Skipping ./assets/10: Not a directory
ERROR: Skipping ./assets/10/original.flv: Not a directory

Any thoughts?


Michela Ledwidge