WRF doesn't produce wrfouts

[fis_clima]

I'm running WRF on a cluster, with 320 tasks, distributed in 4 nodes
It only produced three wrfouts for the 00:00. Then it calculated d03 and d02 values for the 00:10, in 20 and 70 seconds. Then it has been running for three days and didn't even calculated the 00:10 values for d01
I'm sharing the cluster running code and the namelist. Also, just in case, the domains maps, with dx 9, 3 and 1 km
What am I doing wrong?

 

[Ming Chen]

I am suspicious that your case failed immediately after wrf.exe started. Did you look at your rsl files? I expect that you can find some error messages that are helpful for us to figure out what is wrong,

By the way, which version of WRF did you run? What is the input data you used to drive this case?

 

[fis_clima]

I am suspicious that your case failed immediately after wrf.exe started. Did you look at your rsl files? I expect that you can find some error messages that are helpful for us to figure out what is wrong,

By the way, which version of WRF did you run? What is the input data you used to drive this case?

The rsl file seems to show no errors
But its last lines are

Timing for main: time 2025-10-07_00:00:18 on domain 3: 39.75020 elapsed seconds
Timing for main: time 2025-10-07_00:00:36 on domain 3: 1.07021 elapsed seconds
Timing for main: time 2025-10-07_00:00:54 on domain 3: 1.07446 elapsed seconds
Timing for main: time 2025-10-07_00:00:54 on domain 2: 80.31788 elapsed seconds

and doesn't print the next time step


The WRF version is 4.7.1. The input is from GFS

 

[Ming Chen]

Error messages don't necessarily show in rsl.error.0000. They can be in any rsl files. Have you checked all the rsl files?

Your namelist looks fine to me except that time step is way too small. For 9km grid intervals, maximum time_step can be 54. Just stay a little conservative, I would suggest that you set time_step =45.

Please also turn on w_damping and damp_opt.

If this case still fails right after wrf.exe starts, please recompile WRF in debug mode, then rerun the case. With debug mode, the log file will tell exactly when and where something goes wrong first. Such information will be helpful for us to debug possible issues.