Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Scheduler-based testing does not allow for changing module envirorment #614

Open
rcabell opened this issue Apr 21, 2022 · 0 comments
Open
Assignees
Labels
testing Issues related to model testing

Comments

@rcabell
Copy link
Collaborator

rcabell commented Apr 21, 2022

When using the WRF-Hydro CI testing framework with the PBSCheyenne scheduler, there is no way to set a custom module environment for jobs, which can make testing different compiler/MPI combinations difficult and/or cause false-negative test failures.

Expected Behavior

Either the existing module environment set up before calling the testing framework should be passed through, or a provision added to support a custom module set

Current Behavior

Jobs submitted to the PBS scheduler pick up the user's default module environment, instead of any potential custom module environment used to compile the model

Possible Solution

Using the -V flag to qsub would pass through the existing environment from the testing framework to the batch job. This is probably the easiest solution, but an additional command-line argument to specify a module set could work too.

Steps to Reproduce (for bugs)

  1. Load a compiler and MPI library module other than the user's default (in my case, gnu/mpt with a default of intel/impi)
  2. Run the testing framework
  3. ???
  4. Crash!
@rcabell rcabell self-assigned this Apr 21, 2022
@rcabell rcabell added the testing Issues related to model testing label Apr 21, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
testing Issues related to model testing
Projects
None yet
Development

When branches are created from issues, their pull requests are automatically linked.

1 participant