Note: These instructions apply if you are directly creating buckets, such as via the command line or Cyberduck. If you are using SFTPGo, you are not creating buckets directly; instead, you are creating folder structures within -private and -public buckets that have already been created for you. |
To create a bucket, you must create a text file (.txt) containing the names of up to five buckets you want to create.
Naming rules
- Bucket names are automatically prefixed with the stack name — do not include the stack name in the file.
- Each bucket name must be entered on its own line.
- Bucket names can only contain alphanumeric characters and
-. - Bucket names must not begin or end with
-. - Bucket names must be no more than 63 characters total, including:
- the stack name prefix (
duracloud-$ID) - 5 reserved characters for the
-repl suffix
⚠️ Practically, this means your names should be no more than:
63 - 5 - (length of your duracloud-$ID)
*** To create a publicly accessible bucket, the name must end with -public.
- This subtracts an additional 7 characters from the max length.
Reserved prefixes and suffixes
The following cannot be used in bucket names:
duracloud- : already included as the foremost prefix-bucket-requested : used for bucket request files-logs : used for access logging buckets-managed : used for system-managed buckets (you will find reports, logs, and other data here after you start creating buckets and uploading content)-repl : used for replication target buckets (sending your data to Amazon Glacier)
Steps
- Open a text editor, such as Notepad or Notepad++, and decide on your bucket names.
- Upload your text file to the
duracloud-$ID-bucket-requested bucket. - The file will be processed in the background and an attempt will be made to create each bucket.
- Processing normally takes 0–2 minutes.
- A report file will be uploaded to the
logs folder inside the -managed folder, providing details about the outcome. - Review the log when it becomes available.
- Refresh your client view or reconnect to S3.
- Successfully created and configured buckets will now be visible.
- Each new bucket will have an associated replication bucket with a
-repl suffix. - Remember that replication buckets are list-only (files cannot be downloaded).
- The newly created buckets are now usable, and files can be uploaded.
- To create more buckets:
- Re-use and re-upload the same file with new bucket names, or
- Create and upload an entirely new file. Both approaches work.
If you're using Cyberduck or another GUI client, the client may ask you to confirm you wish to override the existing .txt file if using the first option in step 8. |
Troubleshooting:
- If you don't see any new buckets created, check the logs in the
$ID-managed folder for any error messages. - If you try to create multiple buckets at one time but have an error with one bucket (for example, you accidentally gave it too long a name or tried to create more than 5 buckets at one time), none of your buckets will be created. You will need to start again for all buckets.