[WIP] Mpi-Pingpong Benchmark

[michaelmckinsey1]

Description

• todo:
  • Figure out how to pass location of pingpong directory (since private repo) on command line, instead of hard coding
  • Set trials using ramble n_repeats
  • Set number of nodes instead of setting total ranks per machine
  • Set required variables for benchpark analyze

    benchpark/experiments/kripke/experiment.py

    Lines 141 to 145 in 95b4534

    	self.set_required_variables(
    	n_resources="{npx}*{npy}*{npz}",
    	process_problem_size="{nzx}*{nzy}*{nzz}/({npx}*{npy}*{npz})",
    	total_problem_size="{nzx}*{nzy}*{nzz}",
    	)

---

[michaelmckinsey1]

how to set this path dynamically? Can we do it like a ramble variable, e.g. {local_path}. I tried a relative path but it did not work ./mpi-pingpong/ or mpi-pingpong/

[pearce8]

you should use #620 for this.

[michaelmckinsey1]

Running trials in the same allocation parameter (ramble) https://ramble.readthedocs.io/en/latest/workspace_config.html#experiment-repeats

[michaelmckinsey1]

We need to set this in terms of nodes 8, so we don't have to specify this for each machine. However the affinity output will look different, so we need to figure this out.

[codecov-commenter]

Welcome to Codecov 🎉

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

Thanks for integrating Codecov - We've got you covered ☂️

[michaelmckinsey1]

This benchmark will need the builtin-caliper tag from #888 once that is merged