Skip to content

Commit

Permalink
Remove outdated HPC instructions
Browse files Browse the repository at this point in the history
  • Loading branch information
unkaktus committed Jan 10, 2025
1 parent 1bb65ad commit 3d016bf
Showing 1 changed file with 1 addition and 51 deletions.
52 changes: 1 addition & 51 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -65,54 +65,4 @@ Running locally on your machine might be too slow or not fitting into RAM for hi

The idea here is to create the scene in ParaView locally, save it to a state file, and then render it on a remote cluster. As the paths to the data files might be different locally and on the cluster, one has to readjust them by editing the state file. As I didn't write yet the script to automatically swap the path, one has has to do it manually.

The below instructions are currently outdated.

1. Copy `render/render.begin` and your state file (`.pvsm`) to the cluster.

2. Adjust the number of nodes, path to `rose.sif` container, and other parameters in `render.begin`.

3. Begin the rendering job on your state using `spanner`:
```shell
spanner begin -f /path/to/render.begin state.pvsm
```
4. Check the status of the job:
```shell
spanner list
```

6. Track the process:
```shell
spanner logs -f rose_job_name
```

7. The job will create a directory with the same name as the state file, and output the frames there.

8. Create a video named `output.mp4` from the frames in the run directory, adding dark background:
```shell
apptainer exec --bind /scratch:/scratch --bind /work:/work ~/apptainers/rose-v2.0.0.sif \
create_video.py --frames-dir rendered --background-color 0x161616 --output output.mp4
```

### Running parallel frame postprocessing on an HPC cluster

1. Copy `render/overplot/overplot.begin` to the cluster.

2. Adjust the number of nodes, path to `rose.sif` container, and other parameters in `overplot.begin`.

3. Begin the rendering job on your state:
```shell
spanner begin -f /path/to/overplot.begin state.pvsm
```

4. The job will plot legends and a colorbar on top of the frames in the frame directory, and save
postprocessed frames into `overplotted` subdirectory.

5. Create a video from these frames:
```shell
apptainer exec --bind /scratch:/scratch --bind /work:/work ~/apptainers/rose-v2.0.0.sif \
create_video.py --frames-dir overplotted --output output-overplotted.mp4
```

This is your final movie, enjoy! 💫

_Aquí y ahora puede comenzar tu viaje._
_To be written_

0 comments on commit 3d016bf

Please sign in to comment.