Taskflows OpenFoam and ParaView

[dealenx]

Hi,

When i start OpenFoam taskflow and then try to visualise results of the taskflow in Visualizer, usually HPCCloud doesn't export files from traditional cluster into Visualizer.

I guess, i need add new cumulus tasks to openfoam taskflow for exporting files.

So i will be grateful for any advice!

[jourdain]

Normally the idea it that you mount the same drive that you had during the simulation into the visualization step. In other words, we don't move the data. In general we start the viz process on the a cluster that has access to the simulation data. HTH

[dealenx]

On the access to files it is clear, but i don't understand how to change it for the access into Visualization. I mean, do i have to change only openfoam/tutorial.py, or also i have to change any other files?

[jourdain]

I think for the visualization step, we just pass which directory should be used to load data from assuming it is actually on the same machine and accessible to the user.

This get passed here

[dealenx]

The moment of mounting is not clear.

Do I understand correctly that when HPCCloud start with containers hpccloud:compute-pyfr and hpccloud:visualize-osmesa, then the containers mount among themselves here and here, Visualizer is getting a data from a folder /scratch.

[jourdain]

I believe so.