One has two options: the recommended one is use ondemand.triton.alto.fi, though alternatively, one can run a job.
Using Open OnDemand (OOD) is essential and easy, login to https://ondemand.triton.aalto.fi. There on you will have several options, if the application you want to run is part of the Interactve Apps menu, then proceed from there. For instance Matlab, Paraview, R studio are there. In other case, launch a Triton Desktop and you recieve a normal Linux GUI environment of your choice (GNOME or alike).
It will be like you run a Linux Desktop on one of the Triton's compute node with native access to /scratch and module .... Start a terminal within the session and proceed.
OOD is the recommended way.
Not recommended, but still an option.
Prerequisites
Before submitting a job: Optimally, through tests, have a rough idea, how long your job takes, how much memory it needs and how much CPU(s)/GPU(s) it needs.
Required Reading:
Submitting jobs on triton <jobs>
Required Setup:
- Setting up your system to connect to Triton according to the
connection guide <connecting> - Your script and any data need to be on triton (follow e.g. the
data transfer quick-start guide <data>) - Specific to Windows: Install an XServer
First off, in general, using graphical user interfaces to programming languages (e.g. graphical Matlab, or RStudio) is not recommended, since there is no real advantage to submitting a job to the cluster.
However, there are instances where you might need large amount of resources e.g. to visualize data which is indeed intended use. There are two things you need to do to run a graphical program on the cluster:
- Start X-forwarding (when login to Triton, use
ssh -XY ...) - request an interactive job on the cluster (
sinteractive)
Once you are on a node, you can load and run your program.