Using pnetcdf in WRF-Chem model: "DRYRUN" errors while creating wrf restart file

[Wei-keng Liao]

Hi, Zheng

Please cc to parallel netcdf mailing list, so someone in the community can also help.

The warning messages came from the file status, DH%FileStatus, of being
WRF_FILE_OPENED_NOT_COMMITTED in file external/io_pnetcdf/wrf_io.F90.
There are 2 places that set DH%FileStatus to WRF_FILE_OPENED_NOT_COMMITTED:
ext_pnc_open_for_write_begin() and ext_pnc_open_for_read_begin().

So, please check if your program is opening files for write/update (not read-only)
and it calls ext_pnc_open_for_write_commit() to change DH%FileStatus to
WRF_FILE_OPENED_FOR_WRITE or ext_pnc_open_for_update() to WRF_FILE_OPENED_FOR_UPDATE


Wei-keng

On Mar 12, 2015, at 5:25 PM, Zheng Lu wrote:

> I am using wrf-chem 3.6.1. I still got this error..
> 
> Zheng
> 
>> On Mar 12, 2015, at 4:17 PM, Wei-keng Liao <wkliao at eecs.northwestern.edu> wrote:
>> 
>> Hi, Zheng
>> 
>> At the end of that discussion thread, it was concluded that the latest version of WRF 3.6
>> solves the problem. Are you using 3.6?
>> 
>> Wei-keng
>> 
>> On Mar 12, 2015, at 3:49 PM, Zheng Lu wrote:
>> 
>>> Dear Pnetcdf group,
>>> 
>>> I am running wrf model with pnetcdf option on NCAR supercomputer Yellowstone. Because my model simulation is very long, I have to use the “restart” option - which means that the model has to be shut down every 12 hours and be resubmitted. The model did successfully generate output files (15G) and restart files (106G). But after several resubmission (5 times), the model failed to generate restart files and reports many “DRYRUN” errors. 
>>> The errors look like: 
>>> 
>>> d01 2014-09-07_06:00:00  Warning DRYRUN READ in wrf_io.F90, line        2996
>>> d01 2014-09-08_00:00:00  Warning 2 DRYRUNS 1 VARIABLE (LANDMASK) in wrf_io.F90, line        2403
>>> d01 2014-09-08_00:00:00  Warning 2 DRYRUNS 1 VARIABLE (XLAT) in wrf_io.F90, line        2403
>>> d01 2014-09-08_00:00:00  Warning 2 DRYRUNS 1 VARIABLE (XLONG) in wrf_io.F90, line        2403
>>> d01 2014-09-08_00:00:00  Warning 2 DRYRUNS 1 VARIABLE (LU_INDEX) in wrf_io.F90, line        2403
>>> 
>>> After browsing this email list, I found one thread talking about the exactly same problem as I am having:
>>> 
>>> http://lists.mcs.anl.gov/pipermail/parallel-netcdf/2014-June/001575.html
>>> 
>>> I am wondering if anyone solve this problem?
>>> 
>>> 
>>> Thanks in advance!
>>> 
>>> Zheng Lu
>> 
>