Overview

Globus is a secure and reliable service for research data transfers. It handles interruptions, retries automatically, and enables high-performance movement of large datasets. SeaWulf does not operate as a permanent endpoint, so users create a temporary one as needed.

Setup Your Endpoint

module load globus/3.15.0

globus login

You will be prompted to open a URL in your browser, select “Stony Brook University,” sign in with your NetID, then copy the authorization code back to your terminal.

globus endpoint create --default-directory /gpfs/scratch/$USER/ --personal seawulf

Record the Endpoint ID and Setup Key returned by this command.

globusconnectpersonal -setup SETUP_KEY

Replace SETUP_KEY with the key you were given.

1. Load the Globus tools on SeaWulf (works on nodes milan1 and milan2):
2. Log in using your Globus account:
3. Create your personal endpoint (default directory in scratch):
4. Complete setup using the key:

Using Your Endpoint

globusconnectpersonal -start -restrict-paths rw/gpfs/scratch/$USER,rw/gpfs/home/$USER &

The ampersand lets the endpoint run in the background.

globus ls $ep1

globus transfer $ep1:/gpfs/scratch/$USER/myfolder $ep2:~/otherfolder --recursive --label "CLI single folder"

globusconnectpersonal -status

globusconnectpersonal -stop

1. Start the endpoint (allow access to your scratch and home directories):
2. Check if it is working by listing files on your scratch directory (replace $ep1):
3. Transfer files between two endpoints (replace $ep1 and $ep2):
4. Check endpoint status:
5. When done, stop the endpoint:

Summary Table