-
Notifications
You must be signed in to change notification settings - Fork 27
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
Use flux instead of srun on tioga #1011
Conversation
Per advice from LC, attempt to explicitly use `tioga10` runner since something's wrong with the `tioga11` runner.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Excellent. Nice work @bmhan12
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @bmhan12 -- hopefully this will get us past the current CI problem, and we can find a better solution in the meantime.
set(MPIEXEC_EXECUTABLE "/usr/bin/flux" CACHE PATH "") | ||
|
||
set(MPIEXEC_NUMPROC_FLAG "-n" CACHE STRING "") | ||
set(MPIEXEC_NUMPROC_FLAG "mini;run;-n" CACHE STRING "") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This will get lost the next time we run build_tpls.py
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This will get lost the next time we run build_tpls.py
I will do a follow-up PR with the spack fix for this host-config.
I think there are quite a few PR's that are ready to go, but just need this tioga fix for gitlab CI to pass.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @bmhan12 -- I agree that we should get something working as soon as possible.
Longer term, I think the MPIEXEC_NUMPROC_FLAG
should allow for spaces (i.e. flux mini run
) and we should make changes to blt/axom to support that.
@white238 pointed out a place yesterday where MPIEXEC_NUMPROC_FLAG
likely needs to be quoted.
e.g.
https://github.com/LLNL/blt/blob/8c229991e65e7a9603c621b47cb3ba158bb7468c/cmake/BLTMacros.cmake#L904-L914
This PR: