Accessing NCI resources

Welcome to the NCI workshop for Flinders researchers wanting to access NCI resources!

Download the slides

Getting started

You should first create an account at Then you can request an allocation for your own project or join and existing project.

Once you have an account on NCI, you can login to NCI using ssh as follows. You can also use PuTTY on Windows or the Australian Research Environment. You have to replace <username> with your username.

ssh <username>

Once you are logged in, you can check the path on the login node using pwd or echo $HOME.

You can check your default project using echo $PROJECT and your default shell using echo $SHELL. $PROJECT and $SHELL are set in the file $HOME/.config/gadi-login.conf.

You can move to your scratch directory using the following command. You have to replace <project_code> with your own project code and <username> with your username.

cd /scratch/<project_code>/<username>

Similarly, you can move to your gdata directory using the following command.

cd /d/data/<project_code>/<username>

Sample job script

You can create a job script (e.g., using any text editor such as nano or vi. A sample job script is provided below. The lines starting with #PBS are known as PBS directives. More details about PBS directives can be found here.

#PBS -P <project_code>
#PBS -q normal
#PBS -l ncpus=16
#PBS -l mem=1GB
#PBS -l jobfs=64GB
#PBS -l walltime=01:00:00
#PBS -l wd
#PBS -l other=hyperthread
#PBS -l storage=scratch/<project_code>+gdata/<project_code>

# Your code/commands
echo "Hello World!"

Gadi commands cheatsheet

Storage allocation and usage

Check home quota

quota –s

Check the grant/usage of the project on gdata

nci_account -P <project_code>

Check how much storage you have access to on scratch and gdata through which projects


Files created by user/project

nci-files-report --project <project_code>

Transferring files to/from Gadi

Upload data to Gadi from local PC

scp –r <file/folder> <username><destination>
rsync -ravPS <file/folder> <username><destination>

Download data from Gadi to local PC

scp –r <username><file/folder> <destination>
rsync -ravPS <username><file/folder> <destination>


List out software applications centrally available on Gadi

module avail <software>

Load the module into your current shell environment and check

module load <software>/<version>
which <software>

List currently loaded modules

module list 

Remove a module from your environment

module unload <software>


Submit a job script.


Delete a job from the queue

qdel <job_id>

Check status of your jobs

qstat -u <username>

Check status of your jobs

qstat -u <username>

Check status of a particular job in the queue

qstat -sw <job_id>

Check CPU and memory usage the job

nqstat_anu <job_id>

Useful NCI links

Gadi Quick Reference Guide -

PBS Directives Explained -

Compute Grant and Job Debiting -

Queue Limits -

NCI helpdesk -