Justin,
I've attached a small test case that simply initializes a parallel vector and
then attempts to write it to file via DataOutFaces. The finite element used
only lives on the faces of the cells so DataOutFaces in appropriate.
This test case provided me with some insight into my problem. First,
DataOutFaces can support parallel triangulations. I've also attached a figure
showing the subdomains for the test (note that this is from grouping the
individual file outputs in Paraview not writing generating a subdomain field
as part of the output). The other thing this test case showed me is to be
cautious in the problems I am running when running on multiple processors. I
am still new to parallel computing and this problem arose because I was not
careful. My triangulation did not have enough cells to partition the domain
properly when using a higher number of processors (6 or 12). Increasing the
number of cells removed the error and the code ran without an issue.
This is shown in the attached code as well. I ran the code of 4, 6 and 12
processors for two different levels of refinement. The much finer mesh
executed the code without problem. The coarser mesh showed the errors I had
shown on the first post of this thread for 6 and 12 processors.
I tried to run your code on 6 and 12 processors, but it finishes without an
error. With the code you sent, do you just have to do
mpirun -np 12 ./test
to trigger the error on your machine?
What version of deal.II are you using? I (re-)discovered this patch
https://github.com/dealii/dealii/pull/2054
that is part of deal.II 8.4 I believe.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: bange...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/
--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
