...
Please visit the following website to monitor real-time system status.
Code Block |
---|
language | bash |
---|
title | HAL Real-time System Monitor |
---|
|
https://hal-monitor.ncsa.illinois.edu:3000/ |
...
Connect to HAL cluster
There are two 2 methods to log on to the HAL system. The first method is to SSH via terminal,
Code Block |
---|
|
ssh <username>@hal.ncsa.illinois.edu |
...
Submit Interactive Job to HAL
There are 2 ways to submit interactive jobs to the HAL system. The first one is to use the Slurm Wrapper Suite,
and the second method is to submit with Slurm directly.
Code Block |
---|
|
srun --partition=gpux1 --pty --nodes=1 --ntasks-per-node=12 \
--cores-per-socket=3 --threads-per-core=4 --sockets-per-node=1 \
--gres=gpu:v100:1 --mem-per-cpu=1500 --time=2:00:00 \
--wait=0 --export=ALL /bin/bash |
Submit Batch Job to HAL
There are 2 ways to submit batch jobs to the HAL system. The first one is to use the Slurm Wrapper Suite,
Code Block |
---|
|
swbatch run_script.swb
sbatch run_script.sb |
...
Manufacturer | Software Package | Version |
---|
IBM | RedHat Linux | 7.6 |
NVidia | CUDA | 10.1.105 |
NVidia | PGI Compiler | 19.4 |
IBM | Advance Toolchain | 12.0 |
IBM | XLC/XLF | 16.1.1 |
IBM | PowerAI | 1.6.1 |
SchedMD | Slurm | 19.05.2 |
OSC | Open OnDemand | 1.6.20 |
Job Management with Slurm
...