Getting Started

If you read nothing else, read this. This page is an insanely simple guide to just barely get you going. The documentation on this page will not be complete. It will not tell you everything. It will tell you just enough to get started with using Research Computing Cluster, SPORC (Scheduled Processing on Research Computing). First off, email rc-help@rit.edu if you encounter any issues with the cluster.

Research Computing also has a documentation site! Besides housing our documentation, it has much of the same resources as this site except it's not wrapped in the updated RIT branding aesthetic. It also has some of our more technical information, like what SPORC is made of and its testing results.

If you are unfamiliar with the topic of research computing or want to know more about what we offer, check out our Services and About page. 

Questionnaire

At Research Computing, we want to know our researchers, their research, and their computational/research computing needs. Before you can use Research Computing resources, you need to fill out this questionnaire. Your responses to the questionnaire help us understand your project and provide you with everything you need to do your research on our cluster.

Read the Rules

If you use the cluster, we ask that you follow a few simple rules to ensure fair access to the cluster for all researchers. Please read our cluster rules in full. If you have questions or need clarification, please don’t hesitate to contact us

Logging in to the Cluster

If this is your first time using Research Computing resources, you need to login to sporcsubmit to create a user home directory. You can do this by following this documentation using your standard RIT credentials.

Notes:

• This user will not be able to do anything on RC resources until it has been added to a Slurm account.
• You only need to do this once. If you already have a Slurm Account and User, you don’t need to do this again.

Once you have a slurm Account and User, you are ready to get onto our cluster, SPORC. To log in to SPORC or other Research Computing servers, you will need to use SSH or FastX.

The first time you access SPORC, we highly recommend you do so using SSH. After you log in, there will be a message of the day that contains helpful tips and import information, so please make sure you read it. You can also view the message of the day at anytime by running the motd command. Remember: The hostname for SPORC is sporcsubmit.rc.rit.edu.

FastX is useful when you need a graphical interface, such as when you are visualizing data or analyzing graphs.

Guides:

• SSH Documentation
• FastX Documentation

Whether you logged in with SSH or FastX, it’s important that you know how to navigate the terminal. Through the terminal, you will do most of of your Research Computing work, including:

• Creating and editing files
• Loading software
• Running compute jobs

If you have never used a terminal before (or need a refresher), we have an Introduction to Unix & Bash Tutorial guide for you

Linux and Mac users have a built-in command for transferring files over SSH, but there are applications available for every platform to do the same thing. We currently have some documentation for transferring files using FileZilla and Globus, but these are not your only options; if you are logged onto a computer with a graphical interface via FastX, you can use web applications like Google Drive or Dropbox.

Please take a look at our File Transfer documentation.

Slurm is the software we use to manage your research computation on SPORC. When you need to run a compute job, you will tell slurm what resources you need (read this if you need help determining what resources you need). Slurm will dispatch your work to one or more computers that make up SPORC. When your work is done, slurm frees up those resources so other researchers can use them.

Before you write and submit your first job, you should review the terminology and examples below. We highly recommend that you take a look at our Slurm Quick Start Tutorial, which covers the basics of submitting and monitoring jobs. We also a quick reference for basic slurm commands.

Slurm Terminology

Before we move on, let’s clarify some terminology. Slurm has Accounts and Users. When you fill out the Questionnaire, we will create your Account and User.

A slurm Account is a tool for controlling the resources allocated to a user (or a group of users). A User must be added to an Account before they can use it. So, if multiple researchers want to use the same Account, each of their slurm Users must be added to the Account.

A User is just a simple way to connect your RIT credentials to the slurm environment. Important: Having a User and being able to log in to sporcsubmit does not mean you have been allocated resources. Your User must be added to an Account before resources are allocated.

Using a Slurm Batch File (sbatch)

This is the preferred method for creating and running compute jobs on SPORC. An sbatch file consists of some commands that tell slurm what respirces you need, followed by the Linux commands necessary to perform the requested work. Here you can tell slurm how many cores (CPUs), how much memory (RAM), and how much time you need for your compute job. You can submit your compute job using the sbatch command, which places your job in a queue until the requested resources are available.

For example sbatch scripts, please start with our Slurm Quick Start Tutorial, which also points to more examples.

Using Interactive Mode (sinteractive)

If you need to interact with your job, then you should run with sinteractive. This will ask you for the resources you need and then connect you to the scheduled node. If you don’t know what that entails, just try it. Be sure to exit from your sinteractive session by running exit when you are done. If you don’t exit your session, then you are holding onto resources that you aren’t using, which is unfair to other researchers.

For the full process of using sinteractive, please check out our Submitting a Job with sinteractive documentation.

• Try to avoid requesting more resources than you need on the cluster. Any resources that you tell Slurm you’re going to use cannot be offered to other users regardless of whether you actually use those resources. If you fudge your numbers too much, you’re eligible to have your allocation terminated by admins.
• sinteractive is not the best way to run jobs by any means. For any sort of repeatable workflow, you will need to write SBATCH jobs

Your research more than likely needs software, whether it’s for creating, analyzing, or visualizing data. So before you start researching, you need to learn how to use modules and Spack. If there is specific software you need, we might already have it. If we don’t have it, send us an email at rc-help@rit.edu.

You can read more about modules and Spack here:

• Using Modules
• Module load vs Spack load

If you need help using any Research Computing resources, don’t hesitate to contact us. Our infrastructure is changing all the time and our documentation may not always be up-to-date. We’re more than happy to work with you to accommodate your research needs

If you use Research Computing resources, we ask that you cite us in your publications.