Data Transfer Server is a dedicated high-speed data transfer head node, running on the Linux operating system, with 10-gigabit network connections to the public Internet and the HPC private network. hpc-transfer is used solely for data transfers. Job submissions and other standard head node activities are not provided on this data transfer node. For job submissions and compiling, please utilize one of the standard head nodes.

Please keep in mind that hpc-transfer is a shared resource. If you launch too many processes and these processes negatively impact either the data transfer server or the shared file system, your processes will be terminated.

Instructions for transferring your data to hpc-transfer using scp or sftp can be found on our Transferring Files Between Your Computer and HPC page.

Note: Additional settings are required to use hpc-transfer with Filezilla. Use the Site Manager to create a new connection or edit existing settings for hpc-transfer.

  • General tab: select “Interactive” for “Login Type”
  • Transfer Settings tab: check “Limit number of simultaneous connections”
  • Transfer Settings tab: Maximum number of connections: 1

This will result in a pop-up box with DUO instructions when you log in. The Transfer Settings options will prevent you from having to authenticate again once the session is established.