Getting started

Introduction

Centre for Computational Modelling and Simulation (CCMS) aims to promote and support computational modelling as a mainstream research activity amongst the faculty and researchers of NIT Calicut. A High Performance Computing (HPC) Cluster and NVIDIA DGX Station purchased under HEFA are the major computing facility under this centre. The facility consists of 31 CPU nodes and 2 GPU nodes. The NVIDIA DGX Station has four V100 GPU accelerators. The facility is mainly for developing and running parallel codes for research purposes. The HPC system installed has been ranked as the 38th in the list of top 100 computing machines in India, in January 2021, a list maintained by CDAC, Bangalore click here.

System Configuration

Wanted to know about system configuration 😀? Check the it here

Get an Account for You

Centre for computational Modelling and Simulation (CCMS) comprises of a High Performance Computing cluster and a DGX Station that are available to both faculty and students. Inorder to get an account you may download the application and submit filled application to the mail id : ccmsadmin@nitc.ac.in or through hardcopy

You will be notified once your account is approved. For NITC students, recommendation from the guide/ faculty is mandatory. A short proposal describing the computing to be carried out, justification for the use of HPC facility and expected outcome (1 page only) is to be attached with the submitted form.

System Access

The cluster can be accessed through Master node, which allows users to login, to submit jobs, transfer data and to compile source code. (If your compilation takes more than a few minutes, you should submit the compilation job into the queue to be run on the cluster.)

By default, a user will have access to home directory (/gpfs-home/) This directory is available on the login node as well as the other nodes on the cluster. And the /gpfs-scratch/ directory may be used for temporary data storage, generally used to store data required for running jobs. Any data stored in /gpfs-scratch will be deleted after 30 days.

Remote Access

Using ssh in Windows

Windows systems do not have any built-in support for using SSH, so you will have to download a software package to use SSH. PuTTY is the most popular open source (ie free) SSH client for Windows. To install it, visit the download site, and download the Installer package. Once installed, find the PuTTy application shortcut in your Start Menu, desktop. On clicking the PuTTy icon The PuTTy Configuration dialog should appear: Locate the Host Name input box in the PuTTy Configuration screen. Enter the server name you wish to connect to (e.g. [username]@[hpcipaddress]), and click Open. Enter your password when prompted, and press Enter. You are now connected!

Using ssh in Mac or Linux

Both Mac and Linux systems provide a built-in SSH client, so there is no need to install an additional package. Simply locate and run the Terminal app. Once in the terminal, you can connect to an SSH server by typing the following command:

```console // to access from your pc user@my-pc:~$ ssh username@hostid // if a remote GUI section is required user@my-pc:~$ ssh username@hostid -X ```

Note

Your user name and the hostid will be recieved through mail. Please check that. and replace username & hostid with appropriate value

If you want to access the clusters from outside NITC's network, you must use the NITC VPN.

How to change the user password?

Use the "passwd" command to change the password for the user from the login node.

Transferring files between local machine and HPC cluster

Users need to have the data and application related to their project/research work on Madhava HPC Cluster

To store the data special directories have been made available to the users with name “scratch and home” the path to this directory is “/gpfs-scratch” and “/gpfs-home”. Whereas these directories are common to all the users, a user will get his own directory with their username in /gpfs-scratch/ as well as /home-home/ directories where they can store their data. However, there is limit to the storage provided to the users, the limits have been defined according to quota over these directories, all users will be allotted same quota by default. When a user wishes to transfer data from their local system (laptop/desktop) to HPC system, they can use various methods and tools

A user using the ‘Windows’ operating system will get methods and tools that are native to Microsoft Windows and tools that could be installed on your Microsoft Windows machine.Linux operating system users do not require any tool. They can just use “scp” command on their terminal, as mentioned below.

Users are advised to keep a copy of their data with themselves, once the project/research work is completed by transferring the data in from Madhava HPC Cluster to their local system (laptop/desktop). The command shown below can be used for effecting file transfers (In all the tools):

    scp –r [path to the local data directory] [your username]@[IP of Madhava HPC:[path to directory on HPC where to save the data]

for example:

    scp –r /dir/dir/file sajil@:/gpfs-home/sajil

Same Command could be used to transfer data from HPC system to your local system (laptop/desktop).

    scp –r   [your username]@[IP of Madhava HPC:[path todirectory on HPC where to save the data] [path to the local data directory]

for example:

    scp –r  sajil@[cluster IP/Name]:/gpfs-home/sajil/file /dir/dir/file

Addressing Basic Security Concerns

Your account on Madhava HPC Cluster is ‘private to you’. You are responsible for any actions emanating from your account. It is suggested that you should never share the password to anyone.

Please note that, by default, a new account created on Madhava HPC Cluster is readable by everyone on the system. The following simple commands will make your account adequately safe.

Command	Discription
chmod 700 [directory name]	will ensure that only yourself can read, write and execute files in your directory
chmod 750 [directory name]	will enable yourself and the members of your group to read and execute files in your directory
chmod 775 [directory name]	will enable yourself, your group members and everyone else to read and execute files in your directory
chmod 777 [directory name]	will enable EVERY ONE on the system to read, write and execute files in your directory. This is a sort of ‘free for all’ situation. This should be used very judiciously