On Mon, 23 Jun 2008, Derek Gaston wrote:

> Spoke too soon.  It turns out that ex10 is failing with parallel mesh
> and refinement.  To reproduce edit ex10.C and comment out the
> coarsen_fraction() and max_h_level() calls.  Now compile and run with:
>
> mpirun -np 3 ./ex10-opt -n_timesteps 25 -init_timestep 0
>
> It fails for me during the _second_ time it tries to refine.

I get the same error, also during the second adaptive refinement.

You said you got one system to fail earlier during the uniform
refinements?  I'd still like to know how that happened; whatever's
broken should at least be broken deterministically.

In case my sparse terse comments in the code aren't clear enough, let
me explain what the error is: libMesh has found, after refinement and
during prepare_for_use, that processor X thinks element N is owned by
processor Y, but processor Y disagrees.  Unfortunately there's any
number of reasons that could happen...

But as a temporary workaround, would you go to partitioner.C and
uncomment the "don't repartition in parallel" code on lines 47-48?
I'm not yet sure whether there's a bug in Ben's redistribution code or
whether that's just triggering a bug in my core or refinement code,
but I at least can't replicate the problem with repartitioning turned
off.
---
Roy

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
Libmesh-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-devel

Reply via email to