yade-mpi team mailing list archive

[Janek Kozicki <janek_listy@xxxxx>]

You could add a new flag ``mpiRequired`` to attribute flags [1] and
by default assume that the attribute does not need to be send by mpi.

Or maybe this flag is simply ``noSave`` and ``readonly`` and
``hidden`` flags together? This can be possible. Why send over mpi
something that does not need to be saved? Why send something
readonly? Why send something hidden from GUI inspect?


[1] https://gitlab.com/yade-dev/trunk/blob/master/lib/serialization/Serializable.hpp#L31

Janek



Deepak Kn said:     (by the date of Sun, 28 Apr 2019 19:57:54 +0200)

> Hello All,
> 
> I have finished implementing the functions to serialize and deserialize the
> bodies and the necessary communications to exchange the serialized bodies.
> The serialized messages (binary) are several times larger than the usual
> (state) messages in mpy/mpf.py (vel,pos, ori, id, bounds) etc. And as a
> result the simulations are far slower ! Here's how it scales : for
> exchanging 220 bodies, the serialized messages were around 215456 elements
> (chars), and in the previous versions, the message sizes were 4180
> (doubles). (testMPI2D no merge, on 3 procs, 20000 bodies.)
> 
> Most of the functions are implemented in Subdomain.cpp Subdomain.hpp and
> MPIBodyContainer.hpp, (branch : mpi_dpk on gitlab) I followed the same
> logic of the previous version (mpy/mpf), non blocking sends and blocking
> recvs (with probes) are used for the communications between the workers,
> and blocking communication for sending the forces to the master (the same
> as in mpy/mpf).
> 
> I think this can be improved, we can serialize the states,bounds and send
> these instead of the bodies? what do you think? (all the functions do
> serialize, deserialize and communicate are ready. We just need a container
> to hold only the required info.)