HPC User Documentation

Welcome on UNIGE's High Performance Computing (HPC) Baobab service documentation. You will find here all the information you need to use the unige's clusters Baobab, Yggdrasil, Bamboo and get started.

In the documentation, when we talk about Baobab we may refer as the HPC service or one of our cluster.. Baobab.

Before accessing our HPC clusters, you need to :

Have an account - if you don't have one yet, request an account here.
Read and agree to the Terms of Use

HPC clusters at UNIGE

HPC for dummies, an introduction to HPC

Use the HPC resources

How to use Linux - for Linux beginners
How our clusters work - description of the HPC environment
Access the clusters - how to connect with SSH or X2Go, transfer files
Applications and libraries - how to use module to load the available applications on our clusters
Storage - where to store your data on the different storage available on the clusters
Data Life cycle - How to manage your data
Slurm and job management - how to request computing resources with Slurm
Resource utilization and accounting - cpu hours used, utilization reporting
Best practices - best practices and smart use of the HPC resources
HPC team - HPC team and local HPC contact
FAQ - Frequently Asked Questions

How to use OpenOndemand

RoadMap

gantt title HPC Roadmap (Baobab) dateFormat 2025-01-01 section Maintenance section Maintenance February 2025 18-21: M1, 2025-02-18, 3d May 2025 18-21: M2, 2025-05-18, 3d section Data Retention Policy Deploy on Baobab :B1, 2025-03-17, 30d Deleted old files - 1 year :B2, after B1 , 31d Deleted old files - 6 month :B3, after B2 , 31d Deleted old files - 3 month :B4, after B3 , 31d

gantt title HPC Roadmap (Yggdrasil) dateFormat 2025-01-01 section Data Retention Policy

gantt title HPC Roadmap (Bamboo) dateFormat YYYY-MM-DD section Maintenance March 2025 25-27: M1, 2025-03-25, 3d section Data Retention Policy

Other resources

VDI Infrastructure - for GPU computation that cannot be done on the clusters

Support - get help

FAQ : First of all, check the FAQ for general questions and troubleshooting.

Still couldn't find what you need ?
- use the search engine (use @hpc to only search HPC related pages)
- check the detailed Table of content.
- The Glossary if everything sounds like gibberish.

Forum/Community (mandatory for non-private requests): The HPC community forum is the required platform for asking questions if you didn't find a solution here. The forum is quite active and is the best way to ask for help, share, or discuss with the community.
- Please use the Support>issue category and follow the template for any problems you encounter.

HPC-Lunch : First Thursday of each months, you can join our zoom session to discuss with us about HPC. More here

Email : For any other requests (confidential questions, meeting request, etc.) you can contact the HPC team by email (see below)

Contact the HPC team by email

README CAREFULY

WELL WELL WELL, if you read this message is that you have run out all available documentation ressources, isn't it ?

That is to say:

You have googled your issue and found nothing
You have read or search in the official documentation (slurm/linux centos|redhat/etc…)and found nothing
You have post your issue on hpc-community forum section hpc-support>hpc-issues , and you get no answer.
You have read or search in the Unige HPC documenation, and guess what ? nothing neither
You have a specific request

Here he is! You are The one and only winner of the Gemini Croquette Contest, 
and earn the righ to contact the HPC-team. what luck!

You can contact the HPC team by email :

When you reach out for help, we always do our best to get back to you as soon as possible, but please bear in mind that :

Hundreds of other people use our clusters and almost each of you use some special combination of softwares (compilers, libraries, using specific version, etc.). We don't have experts for each software.

The best way to get help quickly is to read this documentation or search/post on HPC community forum.

When you ask for help, please give us enough details so we know exactly what you did, such as :
- What did you try, what didn't work, what is the expected result, what error message you have, etc.
- If this information is missing, most of the time, we will just reply and tell you we need this informations… !

Respect the following mail template, use the mail balise in the subject to be more specific like
- Subject: [baobab][yggdrasil][slurm] error launching jobs with srun

<Hello formula >

Username: $USER
Cluster: $CLUSTER
Subject: <slurm|software|system|connexion| (an explanatory keyword) >
jobid: <separate by space if you have more than one>


<

   <Here the mail body explaining ans describing  everething with the MAXIMUM details.  the MAXIMUM!>

    - Explain us your experience like we were newbies. 
      we don't spy your work so we don't know what is going wrong :/
    - Explain the purpose of your mind, sometimes what we want is not what we think/do/are doing.
    - Give us the command you executed, your sbatch, and all relevant informations
    - Give us the path to the relevant files in Baobab (sbatch, error logs, etc.) and/or add them as attachments.

>


<Respect formula (it's always appreciated)>

<Bye Bye formula>

--
$USER

Here our contacthpc@unige.ch

For more general information about the HPC at UNIGE and eResearch : unige.ch/eresearch/hpc (Digital infrastructures and services for research)

Table of Contents