Getting Started

This page provides a TL;DR introduction to getting started with SCITAS so that you can start working as quickly as possible.

Step 1: connecting to the clusters

You will need to connect to the machines: ssh -l <username> <cluster name>

Account

Ensure you have the appropriate account type for the cluster you intend to use. For more detailed in formations you can read the page Connecting to the clusters.

Step 2: submitting my first a job

Once you are connected you can write a job script and submit using the sbatch command or run an interactive job via Sinteract

You can find out how to write scripts and submit them in the Running Jobs page.

More information on the limits that will be applied on your jobs by the different partitions and QOSs can be found on the Slurm QOS and partitions page.

If you want to understand more about how Slurm manages the job's priorities, we refer you to our Slurm job priorities documentation.

Step 3: more complex jobs need data and generate data

Our documentation on transferring data will give you all the details on how to transfer data between the clusters and your local machine.

Be sure to give a look at How to use the file systems in order to understand where to store your different data

Step 4: I need a specific software

We provide you an exhaustive software stack in the form of modules. More information can be found here Software stack.

If your software/library is not installed, you can try to install it yourself using Spack.