Docs: Add MPIEvaluator tutorial for multi-node HPC systems, including DelftBlue#308
Docs: Add MPIEvaluator tutorial for multi-node HPC systems, including DelftBlue#308
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
5364a18 to
b426fb0
Compare
|
I have no idea why the ReadtheDocs build is failing. I compared the notebook metadata to
|
e3bb224 to
7ba1ddc
Compare
|
One of those cases you spend more than an hour to find that one exact line code is a sea of possible configuration options. But it now works. Ready for review, merge (by fast-forward) after #299. |
- Introduced a detailed tutorial on how to use the new MPIEvaluator from EMAworkbench. - Covered environment setup, including OpenMPI and mpi4py installations. - Provided step-by-step instructions for defining, testing, and running a model on an MPI cluster. - Illustrated how to save and load results using pickle. This is the first part of the tutorial, the second part will follow in the next commit and cover how to run on HPC systems with a SLURM scheduler, like DelftBlue.
Added a detailed guide on how to run the model on the DelftBlue supercomputer, which utilizes the SLURM scheduler: - Steps for creating a SLURM script tailored for DelftBlue. - Overview of the DelftBlue supercomputer's capabilities. - Instructions on environment setup, logging in, file transfer, job scheduling, and result retrieval for DelftBlue.
|
@quaquel Thanks for reviewing, I implemented the feedback and cleaned up the commits. If you approve we can fast-forward merge. Edit: CI doesn't run on docs changes, and currently there is a required check which will prevent merging. Personally I don't think we need the required checks, we won't merge (blindly) when CI fails anyway. |
Can you remove the required check? Otherwise I cannot merge. |
This PR introduces a two-part tutorial for the MPIEvaluator introduced in #299.
The first part (d26f54e) focuses on the general setup and usage of MPIEvaluator in multi-node HPC systems:
The second part (9e04566) extends the tutorial with specific instructions for the DelftBlue supercomputer, which uses the SLURM scheduler:
Together, the tutorial provides an end-to-end guide for leveraging the MPIEvaluator on HPC systems.