[mpas-developers] restart problems with ocean grid

[Todd Ringler]

Hi Mat,

Just to tease this apart a bit more, I suggest the following:

1. Run the shallow-water model with a global grid.nc and with test  
case 2
2. Have it write restarts
3. Try to restart from the restart.nc

This will tells us if the problem is specific to ocean domain restarts  
or not.

Cheers,
Todd

On May 27, 2010, at 12:48 PM, Mathew Maltrud wrote:

> Hi All--
>
> as Todd has experienced in the past, i'm having trouble getting the
> model to restart when using an ocean grid with the sw core.  here's
> what i get after trying to restart from a previous 10 step run:
>
> mm at cy-c1.lanl.gov {32}% tail -f log.0000.err
>  block_decomp_add_halo: Somehow we don't have the right number of
> total cells.
> [n252:13491] *** Process received signal ***
>
> my understanding is that the atm restarts fine, so i'm guessing it has
> something to do with referencing boundary cells, etc.  is some of the
> initialization done differently on restart compared to first time run?
>
> i diff'ed the results from ncdump -h on the grid.nc compared to the
> restart.nc file and everything looked to be consistent.
>
> i'm prepared to go ahead and debug this, but thought i would check
> with the experts if it is something obvious...
>
> thanks...
> -mat
> _______________________________________________
> mpas-developers mailing list
> mpas-developers at mailman.ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/mpas-developers