[petsc-dev] Scaling test with ex13 (snes)

Lawrence Mitchell wence at gmx.li
Sat Oct 3 14:18:28 CDT 2020

> On 3 Oct 2020, at 19:12, Matthew Knepley <knepley at gmail.com> wrote:
> Just to make sure I understand completely, You partition a serial mesh (SELF) onto one process per node (1PROC), and then refine, and repartition the new mesh onto the whole machine (WORLD). Thus I
> need three communicators, right? And also a method, for moving a Plex on a subcomm onto the larger comm, using 0 parts on the new ranks.

I had a go at this (migration of a plex from commA to commB) some time last year, but ran out of steam and time. I pushed my WIP as wence/feature/dmplex-distribute-onto-comm in case its helpful.

I think implementing this should be part of a bigger discussion about what the interface to dmplexdistribute and similar look like.

See also the thread starting here: https://lists.mcs.anl.gov/pipermail/petsc-dev/2019-August/024843.html


More information about the petsc-dev mailing list