connect to UF Research Computing HPG3 Cluster from terminal on macOS

Type the following command in the terminal:

ssh urgatoruser@hpg.rc.ufl.edu

Then, follow the prompts to enter your password and Duo two-factor authentication.

(base) % ssh urgatoruser@hpg.rc.ufl.edu
Password: xxxxxxxx
Duo two-factor login for urgatoruser@ufl.edu

Enter a passcode or select one of the following options:

 1. Duo Push to XXX-XXX-1809
 2. Phone call to XXX-XXX-1809

Passcode or option (1-2): 1

Works through the UF Single-sign on DUO app similar to for Canvas and gatormail.

a single node 2 CPU core job with 2gb of RAM for 90 minutes can be started with the following command

srun --account=psb4934 --qos=psb4934 --ntasks=1 --cpus-per-task=2 --mem=2gb -t 90 --pty bash -i

[mygatoruser@login1 ~]$ srun --pty -p hpg2-compute -n 1 -N 1 -t 90 --mem=2gb /bin/bash

a single node 4 CPU core job with 28gb of RAM for 120 minutes can be started with the following command

srun --account=psb4934 --qos=psb4934 --ntasks=1 --cpus-per-task=4 --mem=28gb -t 60 --pty bash -i

srun --account=psb4934 --qos=psb4934 --ntasks=1 --cpus-per-task=4 --mem=28gb -t 60 --pty bash -i
srun: job 4572441 queued and waiting for resources
srun: job 4572441 has been allocated resources

Now you are on the compute node. You can check the hostname and the node list with the following commands:

[mygatoruser@c0711a-s6 data]$ echo "Hello from $(hostname)"
Hello from c0711a-s6.ufhpc

echo $SLURM_JOB_NODELIST
c0711a-s6

Navigate to group storage on Blue drive then activate conda environment to have access to pipeline

cd /blue/psb4934/share/mears/misophonia/miso-project-pipeline

module load conda
conda activate mne_1_4_2

Random Tips:

Create a symbolic link to your group storage directory on Blue drive:

ln -s /blue/your_group/urgatoruser/ blue

Using OpenOn Demand for file management

Troubleshooting HPG3 Open OnDemand

Using rsync to copy/move files on HPG3

macOS Development Environment: iTerm2, oh-my-zsh, and VS Code