Hi WRF forum folks,
I've been adapting a cycling DART setup to run a new case, and WRF is crashing with an MPICH error when it's time to advance the model to the next cycle. This setup worked before on Derecho (and I've tried rerunning one of the ensemble members from the original case to check if the issue was the new input data), but won't work now with the same submit script and environment. Have there been any module updates on Derecho that might have broken this? An example of the error can be found in /glade/derecho/scratch/mawilson/OSSE_WRF/rundir/advance_temp1/rsl.out.integration. The WRF version I'm using is 4.6.0 and the submit script is tmp_run.csh (in the same directory). I've also tried running WRF with the input files from the OSE I adapted the cycling scripts for using this version of WRF and the precompiled 4.6.0 and 4.7.0 versions on derecho and got the same error each time, although with v4.7.1 WRF runs for 24 seconds before the error appears (the latest test with this input data is in /glade/derecho/scratch/mawilson/wrf_test_OSE). Let me know if there's anything I should try to fix this!
Thanks,
Matt
I've been adapting a cycling DART setup to run a new case, and WRF is crashing with an MPICH error when it's time to advance the model to the next cycle. This setup worked before on Derecho (and I've tried rerunning one of the ensemble members from the original case to check if the issue was the new input data), but won't work now with the same submit script and environment. Have there been any module updates on Derecho that might have broken this? An example of the error can be found in /glade/derecho/scratch/mawilson/OSSE_WRF/rundir/advance_temp1/rsl.out.integration. The WRF version I'm using is 4.6.0 and the submit script is tmp_run.csh (in the same directory). I've also tried running WRF with the input files from the OSE I adapted the cycling scripts for using this version of WRF and the precompiled 4.6.0 and 4.7.0 versions on derecho and got the same error each time, although with v4.7.1 WRF runs for 24 seconds before the error appears (the latest test with this input data is in /glade/derecho/scratch/mawilson/wrf_test_OSE). Let me know if there's anything I should try to fix this!
Thanks,
Matt