I attempted to use MPI_ALLReduce to update the total amount of particles.
Don't know if I did it correctly, but it didn't work. The error I get is:
p5_32211: p4_error: : 8641
rm_l_5_32240: (48.375608) net_send: could not write to fd=5, errno = 32
p0_6538: p4_error: net_recv read: probable EOF on socket: 1
p0_6538: (55.415350) net_send: could not write to fd=4, errno = 32
It shows up after I delete a particle. I think the issue is that the
particle is deleted only on one core, because the particles are split up
between the cores. Is there a way to sync together all the cores easily?
-Dimitry
On Wed, December 2, 2009 02:34, Sander Valcke wrote:
>
>
>
> Dimitry Ayzenberg wrote:
>
>> I know some have set up gadget to delete particles. I was able to do so
>> when running gadget on just 1 processor. I'm not sure what needs to be
>> done so the same will work on multiple processors. I'm having an issue
>> where only one processor registers the deletion and that causes gadget
>> to crash. I know I have to somehow update all the processors with the
>> new information but I don't know how to do that. Any help would be
>> appreciated.
>>
>> -Dimitry Ayzenberg
>>
>>
> It depends on what you are trying to do. You may be needing
> MPI_Allreduce?
>
>
> --
> --------------------------------------------------
> | Sander Valcke |
> | Sterrenkundig Observatorium, Universiteit Gent |
> | Krijgslaan 281 S9, B-9000 Gent, Belgium |
> | phone +32-9-264-4796 | fax +32-9-264-4989 |
> | https://sourceforge.net/projects/hyplot/ |
> --------------------------------------------------
> Adore, v.:
> To venerate expectantly.
> -- Ambrose Bierce, "The Devil's Dictionary"
> --------------------------------------------------
>
>
>
>
>
> -----------------------------------------------------------
>
>
> If you wish to unsubscribe from this mailing, send mail to
> minimalist_at_MPA-Garching.MPG.de with a subject of: unsubscribe gadget-list
> A web-archive of this mailing list is available here:
> http://www.mpa-garching.mpg.de/gadget/gadget-list
>
>
>
Received on 2009-12-02 16:09:50