Describe your problem or question:
Good afternoon! I am having difficulty running CTSM v5.2.005--while the run starts out well, as soon as CTSM attempts to read the surface dataset, the run crashes. As can be seen in the attached rsl.out.0000.txt file, CTSM initiates appropriately until it begins to read the surface dataset (in this case, surfdata_2018_d01_360.nc). Then, as can be seen in rsl.error.0000.txt, the program crashes. It cites "PIOc_openfile+0x15" and libesmf.so.
While this error is associated with WRF-CTSM, I can vouch for the integrity of WRF (it worked with a different LSM), the integrity of ESMF (it has been reinstalled multiple times), and the integrity of LILAC (as you can read in rsl.out.0000, all land-atmosphere coupling was successful). Moreover, the wonderful Dr. Olesen helped create the fsurfdat file (Segmentation fault when running mksurfdata_esmf), so I have difficulty imagining there is a problem with the file. This issue simply must lie with how CTSM is reading this file.
I appreciate any direction that you all can provide. Reading earlier message board posts, there was one similar (some question in cesm.exe). Though there was no resolution, and my problem is different in nature. There was another post that suggested ParallelIO that is locally installed within CTSM is too old compared to the version utilized in ESMF. To the best of my understanding, both are running PIO v2.6.2. Perhaps, however, CTSM's internal PIO infrastructure is a bit different than the internal ESMF PIO.
This machine is running with mpich, no batch software is being utilized, and config_compilers.xml and config_machines.xml are attached. My machine is, indeed, ported to cime.
Good afternoon! I am having difficulty running CTSM v5.2.005--while the run starts out well, as soon as CTSM attempts to read the surface dataset, the run crashes. As can be seen in the attached rsl.out.0000.txt file, CTSM initiates appropriately until it begins to read the surface dataset (in this case, surfdata_2018_d01_360.nc). Then, as can be seen in rsl.error.0000.txt, the program crashes. It cites "PIOc_openfile+0x15" and libesmf.so.
While this error is associated with WRF-CTSM, I can vouch for the integrity of WRF (it worked with a different LSM), the integrity of ESMF (it has been reinstalled multiple times), and the integrity of LILAC (as you can read in rsl.out.0000, all land-atmosphere coupling was successful). Moreover, the wonderful Dr. Olesen helped create the fsurfdat file (Segmentation fault when running mksurfdata_esmf), so I have difficulty imagining there is a problem with the file. This issue simply must lie with how CTSM is reading this file.
I appreciate any direction that you all can provide. Reading earlier message board posts, there was one similar (some question in cesm.exe). Though there was no resolution, and my problem is different in nature. There was another post that suggested ParallelIO that is locally installed within CTSM is too old compared to the version utilized in ESMF. To the best of my understanding, both are running PIO v2.6.2. Perhaps, however, CTSM's internal PIO infrastructure is a bit different than the internal ESMF PIO.
This machine is running with mpich, no batch software is being utilized, and config_compilers.xml and config_machines.xml are attached. My machine is, indeed, ported to cime.
