Using SFTP

From a terminal
If you created an ~/.ssh/config file, for example: ~/.ssh/id_rsa
​
Host mybucket
sftp mybucket
Otherwise, use your username@hostname
If you've named your SSH key something different than id_rsa, change this to
sftp -i path_to_my_private_key username@sftp.mybucket.org
​
Connected to sftp.mybucket.org.
sftp> ls
SFTP, like AWS S3, uses PUT and GET.



With Cyberduck or FileZilla
Cyberduck will read your SSH configuration from your ~/.ssh/config file, and needs no other configuration.
​
​

With rclone (recommended) or rsync
rclone and rsync are exceptionally good tools to use.
​
rclone is especially good at volume transfers as it is heavily multithreaded, and rsync is simple to use for small jobs.
​
Note: When using rclone, set_modtime must be set to false, owing to issues with the underlying s3 storage.
Here's an example of an rclone configuration file:
> cat /Users/jcabraham/.config/rclone/rclone.conf
[mybucket]
type = sftp
host = sftp.mybucket.org
user = jabraham
key_file = ~/.ssh/id_rsa
md5sum_command = none
sha1sum_command = none
set_modtime = false
With this configuration, one can do things like this:
> rclone sync ./my-giant-dir/ mybucket:/my-giant-dir