Alternatives

[ad-m]

Hello,

I suggest mention alternatives, so s3cmd support du command. Eg.

$ s3cmd du -s -H
990M     49 objects s3://backup-{CROP}/
6G       37 objects s3://backup-{CROP}/
3G       118 objects s3://backup-{CROP}/
7G       9220 objects s3://cdn.files.jawne.info.pl/
0        0 objects s3://cdn.pilnujemy.info/
28G      612 objects s3://duply-dell/
0        0 objects s3://test_creator/
0        0 objects s3://test_creator2/
0        0 objects s3://test_creator3/
0        0 objects s3://test_creator4/
0        0 objects s3://test_creator6/
0        0 objects s3://test_creator7/
4G       1 objects s3://{CROP}/
--------
51G      Total

s4cmd support du too.

[ad-m]

There is also some bash solution…https://gist.github.com/hrwgc/3fedab87eb937772ca58

[owocki]

As you correctly noted, s3cmd/s4cmd have a du command. For my particular use case (paring down s3 storage costs), I needed to traverse my s3 buckets as specified depths (-d argument, built into above repo).

You're right that the repo should mention the most promising alternatives. I myself would be particularly interested in an alternative that is performant on very large buckets. Want to fork the repo and submit a PR with your suggested edits to the readme?