Push notifications in your browser are not yet configured.
You are not logged in, you may not see all content and functionalities. If you have an account, please login .
Problem with writing restart file
Dear colleagues, I have a problem with the nesting simulation (cclm-to-cclm) at the second cclm stage ( COSMO - CLM 5 version). The model has been run on 4 months, from 05/2010 to 08/2010 for Moscow region. But the model was interrupted on 240 hours step. 240 hours is the nhour_restart value, therefore, I have suggested, the problem is associated with restart parameters. Moreover, the restart file wasn’t written, and I can’t resume my simulation from this step. Could you suggest, please, where the problem is. Maybe, it associated with many Namelist parameters. I have attached the log-file (slurm-721469.out), launching script (cclm5_nest.sh). Furthermore, the OUTPUT file wasn’t written.
Thank you very much for any suggestions!
Please check, if the file
IF (my_id 0) THEN IF (ymode(1:1) ‘r’) THEN OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ OLD ’, & ACTION =‘ READ ’, IOSTAT =ierror) ELSEIF ((ymode(1:1) 'w') .OR. (ymode(1:1) ‘a’)) THEN ! there is no append for restart-files. They are NEW ! OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ NEW ’, & OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ UNKNOWN ’, & ACTION =‘ WRITE ’, IOSTAT =ierror)7780OUT_DIR/restarts/lrfd2010051100oalready exists.If so, either delete the file or better:
In
io_utilities.f90inopen_filearound about line 770 change ‘ NEW ’ to ‘ UNKNOWN ’ for the file openingand compile a new binary.
This generally changed in the climate version, but not yet done for the “fresh” release of the version cosmo5.0
The file lrfd2010051100o doesn’t exist. But, it seems to me, the problem could be somewhere else, because we have run this compiled binary file previously without any problems. Maybe, I didn’t understand something, but I have run the COSMO 5 climate version, not the NWP one. I have looked into io_utilities.f90 file, the following states there:
IF (my_id 0) THEN IF (ymode(1:1) ‘r’) THEN OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ OLD ’, & ACTION =‘ READ ’, IOSTAT =ierror) ELSEIF ((ymode(1:1) 'w') .OR. (ymode(1:1) ‘a’)) THEN ! there is no append for restart-files. They are NEW OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ NEW ’, & ACTION =‘ WRITE ’, IOSTAT =ierror)As I understand, this is correctly…
I have set nhour_restart parameter to hstop value. Just all simulation was OK, except the last step before writing the restart file. I have remarked in log-file (attached as slurm-722843.out), the lbfd….nc file is opening and closing at every step, but there are no any lbfd…nc files in any directories. After closing these files the string
was written. Also, the OUTPUT log file wasn’t written (it is empty)… Maybe, it all could be associated with some NAMELIST parameters, which I don’t know well…
Thank you for any suggestions and help!
From the log file the file
IF (my_id 0) THEN IF (ymode(1:1) ‘r’) THEN OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ OLD ’, & ACTION =‘ READ ’, IOSTAT =ierror) ELSEIF ((ymode(1:1) 'w') .OR. (ymode(1:1) ‘a’)) THEN ! there is no append for restart-files. They are NEW ! OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ NEW ’, & OPEN (nudat, FILE = TRIM , FORM =‘ UNFORMATTED ’, STATUS =‘ UNKNOWN ’, & ACTION =‘ WRITE ’, IOSTAT =ierror)/mnt/msu/users/vplatonov/COSMO_CLM/CLM/EXPERIMENTS/ETR/summer2010test_downscaling/LM/lbfd2010083118.ncdoes either not exist or has wrong permissions at the time when COSMO was executed.
Regarding restarts please change your source code according to
otherwise you run into trouble in future.
Make always sure that I/O directories and necessary input files exist and have read/write permission.