| Table of Contents |
|---|
Getting Help
Support staff may be contacted via:
...
Quick Classification: HPC
Service: eResearch
Category: Research Computation, Analysis and Visualisation
Subcategory: High Performance Computing
Team: RSD - eResearch and Specialised Platforms - HPC
eResearch and Specialised Platforms, Griffith University
Email: eresearch-support@griffith.edu.au
Web: https://www.griffith.edu.au/eresearch-services
Getting Started
Logging in.
To log in to the cluster, ssh to
| No Format |
|---|
10.250.250.7
clogin1.rcs.griffith.edu.au |
What is ssh
SSH Secure Shell provides users with a secure, encrypted mechanism to log into systems and transfer files
ssh clients
windows platform
To install a ssh client e.g. putty, please follow this instruction
Linux platform and mac platform
ssh -Y snumber@clogin1.rcs.griffith.edu.au
Multi-Factor Authentication (MFA) has been implemented on Griffith HPC as an additional layer of cybersecurity. You will receive this message
Authenticating using PingID...
Please use your pingid authenticator app to approve. After that you will receive this message:
PingID authentication completed successfully
PS: Further information about pingID can be found here.
Transferring files between your desktop and gowonda cluster
Transferring files between your desktop and Griffith HPC cluster
Once you are on the system, have a look around. Your home directory is stored in:
/exports/home/<SNumber>
where you have 50GB of allocated space.
Work space is available at:
/scratch/<snumber>
which is a "scratch" area for short lived data. All data older than one week is deleted from this folder.
You should not read or write directly into your home directory with submitted jobs - they should always use the "/scratch/<snumber>" filesystem
Running Jobs
We provide modules that set up the environments you need to use for specific packages. You can see what modules are available with the command:
module available
and you can load them with the command:
module load MODULENAME
See sample pbs scripts here.
If you don't specify a maximum runtime, you will end up with a maximum runtime as short as 30 minutes, depending on which node your job runs on.
Job Stats
when you run qstat from the command line, you will see your job's state within the queue. The following form lists the possible states and their descriptions
...
state code
...
State description
...
q
...
queued
...
r
...
running
Cluster Status
(When done, you can press "Q" to quit from the below text-based browsers
Use the commands to check the cluster status: pbsnodes, pbsjobs, and pbsqueues
elinks http://localhost:3000/nodes
elinks http://localhost:3000/jobs
elinks http://localhost:3000/queuesTo find the current usage, run this command:
du -kh /export/home/snumber
du -kh /scratch/snumber
You may run the following command to check free resources on cluster at any moment:
qstatr
qhost
Reference:
1. ECDF. 20 July 2011. U of Edinburgh. 7 Oct. 2008 <www.ecdf.ed.ac.uk>