It is important that researchers not spend too much time and effort transferring datasets from one place to another. This session will introduce the basics of data transfer so that participants can identify bottlenecks that slow down a transfer job and learn how to overcome them. The session will also introduce various data transfer tools that researchers can use in their daily workflow. The hands-on part will focus on the Globus transfer tool.
The session assumes no prior experience or knowledge about networking or data transfer practices.
Attendees will be able to select the best data transfer tool for different transfer jobs. Attendees will also learn the best way to transfer large and small datasets from and to the Research Computing clusters (Perseus, Della, Tiger, etc) at Princeton.
Basic facility with Linux and the Bash command-line
If you would like to participate in the hands-on activity, you will need to enable your Princeton Unix account and make sure that you can connect to Princeton’s Nobel machine via SSH. Here are the steps:
Enable your Princeton Unix account (if you have not already done so), as described in the PICSciE Workshop Requirements page.
Make sure you can connect via SSH to Princeton’s Nobel computers. For instructions on how to do so for your operating system, visit the PICSciE Workshop Requirements page and see the section titled “Connecting to Adroit” (you need the instructions under “SSH Connection”, not the shorter ones under “Web Portal”), making the following changes:
everywhere you see “Adroit”, replace it with “Nobel”
ignore any references to VPN (you don’t need it to connect to Nobel)
Finally, verify that you can connect to Nobel (follow the steps on the PICSciE Workshop Requirements page in the section titled “SSH into Adroit”, but put “Nobel” everywhere you see “Adroit”.
Lecture and hands-on
All presentation materials are here.
A recording of the session is here (requires active Princeton NetID to view).