ParaView

Description

ParaView website  

ParaView is an open-source, multi-platform data analysis and visualization application. ParaView users can quickly build visualizations to analyze their data using qualitative and quantitative techniques. The data exploration can be done interactively in 3D or programmatically using ParaView’s batch processing capabilities.

Environment Modules

Run module spider ParaView to find out what environment modules are available for this application. See modules documentation for details.

System Variables

• HPC_PARAVIEW_DIR - installation directory
• HPC_PARAVIEW_BIN - executable program directory
• HPC_PARAVIEW_LIB - library directory

How To Run

Using GUI on HiPerGator

To run ParaView with GUI, please follow GUI application instructions.

Using Client/Server Mode

In client/server mode, you install ParaView on your local machine as a client and use HiPerGator as the back end server to do the processing. To accomplish this, you will need to start the server on HiPerGator and then use SSH port forwarding to be able to connect to that server from your local machine:

You will need the same version of ParaView installed on your local machine as the server you load on HPG. For example, if you start the server using "module load paraview/5.9.1, you will need the 5.9.1 version of the client installed on your local machine in order to connect to it. You can get current and previous versions of ParaView from https://www.paraview.org/download/

1. Start the ParaView server on HiPerGator

• Create a job script with the appropriate resources to run the server
• The job script should load the ParaView module and launch the "pvserver" component:

module load gcc/8.2.0 openmpi/4.0.3 paraview/5.9.1
srun --mpi=pmix_v3 pvserver --force-offscreen-rendering --server-port=11111

For previous versions of Paraview, such as paraview/5.4.1, the option --force-offscreen-rendering is not available. Instead, invoke --use-offscreen-rendering.

(see ParaView_Job_Scripts page for an example server job script.)

• Start the server job using "sbatch myjobscript.sh" (replacing "myjobscript.sh" with the name of the job script you created in the previous step) and look for the node that it is running on using "squeumine" (getting the node name from the "NODELIST" column):

$ sbatch myjobscript.sh
Submitted batch job 16305129
$ squeuemine
    JOBID PARTITION                 NAME         USER ST       TIME  NODES NODELIST(REASON)
 16305129       gpu paraview_server_test albert.gator  R       0:08      1 c1101a-s17

2. Forward the port to your local machine

• In a second terminal on your local machine, use the following command to forward the server port to you. Replace "NODENAME" with the name of the node that you got from the "NODELIST" column previously and replace HPGUSERNAME with your own:

ssh -L 11111:NODENAME:11111 HPGUSERNAME@hpg.rc.ufl.edu

Once you have completed the multi-factor authentication, leave that terminal open as it is silently forwarding the HPG traffic to you.

3. Connect ParaView to your server

• In ParaView (the client...running on your local machine), choose File | Connect and then click [Add Server]
• Edit the server configuration:
  • Name: HiPerGator
  • Server Type: Client/Server
  • Host: localhost
  • Port: 11111
• Click [Configure]
• Set the "Startup Type" drop-down list to "Manual" and click [Save]
• Highlight (click on) the HiPerGator entry and click [Connect]

You might receive a warning to the effect of "Display is not accessible on the server side. Remote rendering will be disabled.". This is expected (since we are rendering on the client side) and can safely be ignored.

Now, when you choose File | Open..., you should see your HPG home directory and your processing will take place there under the resources you specified in your server job.

If you disconnect from the server (or close ParaView on your local machine), the server job will be cancelled and you will need to resubmit it to connect again.