Uploading files to datateam

We’ll use the large dataset method if a submitter has a dataset with around 700+ files, as the uploading would take too long through the website. When we confirm with a PI that their submission has 700+ files, we will begin to give instructions for how they can directly upload all their files to our datateam server. Eventually, we’ll move these files to a web-accessible folder so that viewers can access their files directly from our server instead of through metacat.

Instructions for uploading files to datateam

To have submitters upload their files to our server, we will give them access through SSH (secure shell). Please reach out to a Data Coordinator for the credentials of the visitor account in the datateam server.

If the submitter is familiar with SSH, we can give them the credentials and tell them to create a folder under the home directory of /home/visitor to add all their files. If they are not familiar, we will give them the following instructions to use a GUI client for SSH:

Cyberduck is a free software that you can download at https://cyberduck.io/download/, and follow the directions below:

  1. Open Cyberduck.

  2. Click “Open Connection”.

  3. From the drop-down, choose “SFTP (Secure File Transfer Protocol)”.

  4. Enter “datateam.nceas.ucsb.edu” for Server.

  5. Enter your username and password (ask Data Coordinators)

  6. Connect.

  7. Create the folder /home/visitor/submitter_name and upload your files.

Moving files within datateam

Eventually, we will be moving the files from /home/visitor to the /var/data shoulder. While we are in the process of curating the dataset, we will move the files to var/data/curation.

Then, when we are ready for the files to be in a web-accessible location (i.e. when the dataset is ready to be published), we’ll move the files to var/data/10.18739/preIssuedDOI/. Before we can create the preIssuedDOI directory, we’ll need to mint a pre-issued DOI for this dataset.