[Wrf-users] 3 nested domain stops without an error

Κώστας Μυρωνάκης kmironakis at gmail.com
Thu Feb 18 02:26:37 MST 2016


In my case, I use a single domain 500x500x30 in a Intel(R) Xeon(R) CPU
E3-1245 V2 @ 3.40GHz (Cores 8)
   Memory:  15960 MB.
It takes the 50% of memory ~= 8GB RAM and its complete in 4 hours (for 36
hours forecast).
In your case you need 16GB RAM minimum and more!


Efstathios Konstantinos Mironakis
http://www.forecastmaps.eu


2016-02-17 19:41 GMT+02:00 <mmkamal at uwaterloo.ca>:

> Hi Dimitris,
>
> By looking at your rsl.error and namelist.input files it seems to me
> that your allocated computing resource may not be adequate to
> integrate such a big domain (domain 1: 142x66x30; domain 2:
> 211x154x30; domain 3: 448x391x30 grid points). Your domain is really
> very big although I do not know how many core you have allocated. FYI,
> you double your domain resolution (decrease grid spacing half) will
> require 4 times more computing resource.
>
> In addition, I noticed another problem, your log file is showing that
> while "module_io.F" failed to read al 2D input data generated during
> real.exe steps and metgrid.exe stepts. So, please have a look at first
> whether metgrid.exe step finished message "successfully finished
> metgrid.exe" and real.exe step with "successfully finished real.exe".
> If the above two steps are done successfully then increase your
> allocated computing resources and give a try.
>
> I got the following error message in your log file:
> ----------------------------------
> input_wrf.F reading 2d real xlat
>   date 2010-06-01_00:00:00
>   ds            1           1           1
>   de          141          65           1
>   ps          107          50           1
>   pe          141          65           1
>   ms          100          43           1
>   me          147          71           1
>    module_io.F: in wrf_read_field
>   input_wrf.F reading 2d real xlat Status =            0
> input_wrf.F reading 2d real xlong
>   date 2010-06-01_00:00:00
>   ds            1           1           1
>   de          141          65           1
>   ps          107          50           1
>   pe          141          65           1
>   ms          100          43           1
>   me          147          71           1
>    module_io.F: in wrf_read_field
>   input_wrf.F reading 2d real xlong Status =            0
> -------------------------------------------
>
>
> That's all I can say right now.
>
>
>
> Thanks
> Mostofa Kamal
> PhD Candidate
> Earth and Environmental Sciences
> University of Waterloo
> Ontario, Canada
>
>
>
>
>
>
>
>
> Quoting Dimitris Aman <dimaman25 at gmail.com>:
>
> > My greetings to all,
> > everything works just fine with the 2 domains but when I add a third
> > one the simulation stops without any error when it reaches domain 3.
> > Unfortunately, I cannot figure out what is the problem because as
> > soon as I change the max_dom option from 3 to 2 everything works
> > without any problem. I could real use some guidance. I am including
> > the namelist.input and the rsl.error file.
> >
> >
> >
> > Thank u all for your time
> >
> >
> >
> > Dimitrios Georgios Amanatidis, MSc,
> > PhD Candidate University Of Crete, Department of Chemistry,
> > Environmental Chemical Processes Laboratory
> >
> >
> >
> >
> >
> >
> >
> >
> >
>
>
>
>
>
> _______________________________________________
> Wrf-users mailing list
> Wrf-users at ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20160218/c1c10eb5/attachment.html 


More information about the Wrf-users mailing list