Access denied copying files using S3 CLI

Question

Attempting to pull down the contents of an S3 bucket using the AWS CLI, I'm getting the following:

aws s3 cp --region us-east-1 s3://s3.amazonaws.com/my-bucket . --recursive
A client error (AccessDenied) occurred when calling the ListObjects operation: Access Denied
Completed 1 part(s) with ... file(s) remaining

Using aws s3 sync similarly fails.

The user policy is:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": "s3:*",
      "Resource": ["arn:aws:s3:::*"]
    }
  ]
}

(I've tried various less restrictive policies too, but to no avail).

I've tried an empty bucket policy, and also this bucket policy:

{
    "Id": "Policy1357935677554",
    "Statement": [
        {
            "Sid": "Stmt1357935647218",
            "Action": [
                "*"
            ],
            "Effect": "Allow",
            "Resource": "arn:aws:s3:::my-bucket",
            "Principal": {
                "AWS": [
                    "arn:aws:iam::XXXXXXXXXX:user/my-user"
                ]
            }
        },
        {
            "Sid": "Stmt1357935676138",
            "Action": [
                "*"
            ],
            "Effect": "Allow",
            "Resource": "arn:aws:s3:::my-bucket/*",
            "Principal": {
                "AWS": [
                    "arn:aws:iam::XXXXXXXXXX:user/my-user"
                ]
            }
        }
    ]
}

Interestingly, this does work:

aws s3api list-objects --region us-east-1 --bucket my-bucket

Solution

The format for specifying an s3 location is s3://bucket/key so instead of s3://s3.amazonaws.com/my-bucket you would use s3://my-bucket/.

Comments

[KevinD]

Thanks, that worked! "Access denied" isn't the most obvious error for a malformed URL!

[Chris]

@KevinD to be clear the URL is not malformed, "s3.amazonaws.com" is a valid bucket name per naming conventions.