Welcome¶
Welcome to the TP-Stack HPC cluster documentation. This site covers everything you need to get started — from accepting your invitation to running jobs on the cluster.
-
TP Stack
Key directories, monitoring, and rules of the road.
-
Connecting
Accept your invitation, register, set up SSH, and transfer files.
-
Running Jobs
Submit CPU and GPU jobs with Slurm.
-
Environments
Set up shared Python environments for your project.
-
Troubleshooting
Common problems and how to fix them.
Quick Reference Card¶
Access¶
Connect: ssh YOUR_USERNAME@hpc.tp-stack.co.uk
Portal: https://waldur-portal.tp-stack.co.uk
Monitoring: https://grafana-monitoring.tp-stack.co.uk
Files¶
Upload: scp file.csv YOUR_USERNAME@hpc.tp-stack.co.uk:~/
Download: scp YOUR_USERNAME@hpc.tp-stack.co.uk:~/result.csv ./
Upload folder: scp -r project/ YOUR_USERNAME@hpc.tp-stack.co.uk:~/
Directories¶
Jobs¶
Submit: sbatch my-script.sh
Status: squeue --me
History: sacct --me
Cancel: scancel JOB_ID
Nodes: sinfo
Python¶
Create venv: python3 -m venv /projects/X/venv
Activate: source /projects/X/venv/bin/activate
Install: pip install numpy pandas torch