Hi folks! This is my first time posting in the CESM forums so I apologize in advance if I forget to include important information. I have looked around the existing threads and haven't found anything similar to this issue. If there is an existing thread on something like this that I just missed, I apologize as well!
What version of the code are you using?
I am trying to port CESM2.2.2 to Narval, an HPC associated with the Digital Research Alliance of Canada. The result of ./describe_version is:
release-cesm2.2.2-0-g779b0a3
Have you made any changes to files in the source tree?
I have only changed the wallclock time and stop options. I am running a quick port test for 2 months.
Describe every step you took leading up to the problem:
I created the case with: ./create_newcase --case port_test_new1 --res f09_f09_mg17 --compset FHIST
Then set it up and built it with:
./case.setup
./case.build --clean
./case.build
I had already downloaded data during an earlier port test with: ./check_input_data --download
I did not repeat this step since I figured the input data needed to only be downloaded once
Then submitted the case with ./case.submit
I have attached my config files for this port to Narval. They are txt files as I wasn't able to upload them as xml files, hope this is alright! I was able to successfully run CESM2.1.5 with the same commands as above, but I want out of the box nudging support and so I am now attempting to port CESM2.2.2
Describe your problem or question:
My issue is basically described in the title of the thread. The model builds, submits, and runs successfully without crashing. It produces two output files (port_test_new1.cam.h0.1979-01.nc and port_test_new1.cam.h0.1979-02.nc) under ~/out/initial_port_test/archive/case/atm/hist that have all the correct output variables stored, but the values of every variable are bogus. I have attached a screenshot of the surface pressure variable, which seems to be in the range of -3.37*10^38 to 3.37*10^38.
I'm not even sure how to start debugging something like this. My supervisor has a hunch that there is an issue with a netCDF library such that the model is simulating all values correctly, but they essentially get totally garbled when being written to the output files. Is there a way for me to check this without modifying the source code to print out intermediate steps somehow? I am at a total loss right now and would love to know if anyone has seen anything like this before.
Thanks in advance for your help! I appreciate any guidance.
What version of the code are you using?
I am trying to port CESM2.2.2 to Narval, an HPC associated with the Digital Research Alliance of Canada. The result of ./describe_version is:
release-cesm2.2.2-0-g779b0a3
Have you made any changes to files in the source tree?
I have only changed the wallclock time and stop options. I am running a quick port test for 2 months.
Describe every step you took leading up to the problem:
I created the case with: ./create_newcase --case port_test_new1 --res f09_f09_mg17 --compset FHIST
Then set it up and built it with:
./case.setup
./case.build --clean
./case.build
I had already downloaded data during an earlier port test with: ./check_input_data --download
I did not repeat this step since I figured the input data needed to only be downloaded once
Then submitted the case with ./case.submit
I have attached my config files for this port to Narval. They are txt files as I wasn't able to upload them as xml files, hope this is alright! I was able to successfully run CESM2.1.5 with the same commands as above, but I want out of the box nudging support and so I am now attempting to port CESM2.2.2
Describe your problem or question:
My issue is basically described in the title of the thread. The model builds, submits, and runs successfully without crashing. It produces two output files (port_test_new1.cam.h0.1979-01.nc and port_test_new1.cam.h0.1979-02.nc) under ~/out/initial_port_test/archive/case/atm/hist that have all the correct output variables stored, but the values of every variable are bogus. I have attached a screenshot of the surface pressure variable, which seems to be in the range of -3.37*10^38 to 3.37*10^38.
I'm not even sure how to start debugging something like this. My supervisor has a hunch that there is an issue with a netCDF library such that the model is simulating all values correctly, but they essentially get totally garbled when being written to the output files. Is there a way for me to check this without modifying the source code to print out intermediate steps somehow? I am at a total loss right now and would love to know if anyone has seen anything like this before.
Thanks in advance for your help! I appreciate any guidance.
