ufs-weather-model
ufs-weather-model copied to clipboard
ufs-community
Error running rrfs_v1nssl_intel with -DDEBUG=ON using ufs-community:develop
Description
Experience failure when attempting to run rrfs_v1nssl_intel on Hera, with -DDEBUG=ON, against ufs-community:develop branch.
To Reproduce:
- log into Hera.
- Clone ufs-community:dev branch
- edit rt.conf to add -DDEBUG=ON to compile line of case rrfs_v1nssl_intel
- Run test
Additional context
Example from Denise's attempt
/scratch1/NCEPDEV/stmp2/Denise.Worthen/FV3_RT/rt_301756/rrfs_v1nssl_intel
That line is for bdsnow below
Output
Clarification...I think your output above showing the failure in module_mp_nssl_2mom.F90 is the failure that @BrianCurtis-NOAA got w/ the PR #1915 branch on Acorn in non-debug mode. The failure I posted was for debug mode at develop 68050e5
Running debug with gnu on Hercules. Some failures, but mostly due to timeouts. rrfs_v1beta_gnu seemed to be the only one to fail from another cause. I'll re-test the others that failed with an extended wall clock. working dir = /work2/noaa/stmp/zshrader/stmp/zshrader/FV3_RT/rt_2728437/
Logs: /work/noaa/nems/zshrader/hercules/rt-1915/tests/logs/RegressionTests_hercules.log
Clarification...I think your output above showing the failure in module_mp_nssl_2mom.F90 is the failure that @BrianCurtis-NOAA got w/ the PR #1915 branch on Acorn in non-debug mode. The failure I posted was for debug mode at develop 68050e5
Yes, that's true. I'll remove it.
@zach1221 I don't see the
rrfs_v1nssltest in your run directory.
@DeniseWorthen it doesn't look like rrfs_v1nssl is part of the gnu tests in rt.conf actually. I can try to compile it with gnu.
It fails with gnu. Screenshot of the err log but I'm not sure it's helpful.
/work2/noaa/stmp/zshrader/stmp/zshrader/FV3_RT/rt_2824213/rrfs_v1nssl_gnu
It fails with gnu. Screenshot of the err log but I'm not sure it's helpful.
/work2/noaa/stmp/zshrader/stmp/zshrader/FV3_RT/rt_2824213/rrfs_v1nssl_gnu
Can't see it well in the SS, but it looks like the glacier file where we see the issue with the intel compiler?
It fails with gnu. Screenshot of the err log but I'm not sure it's helpful.
Can't see it well in the SS, but it looks like the glacier file where we see the issue with the intel compiler?
Yes, sorry I cut it off just above the glacier file portion I think.
pointing to the same bdsnow line in module_sf_noahmp_glacier.F90 , that Denise posted previously.