Connecting to Argon (Linux/Mac terminal)
You can connect to the Argon cluster on a Linux or Mac machine (or using Cygwin or WSL on Windows) by opening a terminal and using ssh with your HawkID and password:
Note that if, on the machine you’re using, your username is not the same as your HawkID, you’ll have to specify your login username:
$
ssh hawkid@argon.hpc.uiowa.edu
By default, ssh uses port 22, which is blocked from outside connections by the University of Iowa network. So if you are connecting from off-campus, you will have to specify port 40:
$
ssh -p 40 argon.hpc.uiowa.edu
Note that you do not have to set up a VPN to connect to Argon; however, you will be prompted for two-factor authentication via Duo.
To set up a password-less login using RSA keys, follow the instructions here.
Connecting to Argon (Windows)
On a Windows machine, you can use PuTTY to connect to Argon in essentially the same way as above, only with a GUI interface. Just be sure to specify:
- Connection: SSH (default)
- Host Name: argon.hpc.uiowa.edu
- Port: 22 (default) is fine on-campus, but you need to specify 40 if you’re connecting from off-campus
You will then be prompted for your username (HawkID) and password.
Connecting by Remote Desktop
It is also possible to connect to Argon by remote desktop using FastX2, which you can download from ITS. I have never personally done this, but instructions are available here. If anyone has tried this out and has any tips to share, let me know and I will make them available here.
Connecting through IDAS
This tutorial is organized around learning how to submit batch jobs to the HPC by logging into Argon directly. Another option is the University’s Interactive Data Analytics Service (IDAS). As the name indicates, this is intended to support interactive use. Rather than submitting a job and walking away, you will open an RStudio (or Jupyter) session and run things on the HPC interactively.
There are pros and cons to each approach. It is easier to get up and running with IDAS, but there is also power in learning how to submit, say, 100 automated jobs and walk away. Each approach is useful in various cases. See here for documentation and an FAQ comparing IDAS and Argon.