On 07/16/2017 11:11 AM, 'franck75' via deal.II User Group wrote:
I am having that error message when I run in parallel with 2 processor like this
mpirun -n 2 ./darcyparl
and when I run with one MPI process with mpirun -n 1 ./darcyparl I have this
error message
An error occurred in line <474> of file </home/franckm/darcyparl/darcyparl.cc>
in function
void Step43::LinearSolvers::InverseMatrix<Matrix,
Preconditioner>::vmult(VectorType&, const VectorType&) const [with VectorType
= dealii::TrilinosWrappers::MPI::Vector; Matrix =
dealii::TrilinosWrappers::SparseMatrix; Preconditioner =
dealii::TrilinosWrappers::PreconditionIC]
The violated condition was:
false
Additional information:
--------------------------------------------------------
An error occurred in line <590> of file
</home/franckm/deal.ii-candi/deal.II-v8.5.0/include/deal.II/lac/solver_cg.h>
in function
void dealii::SolverCG<VectorType>::solve(const MatrixType&, VectorType&,
const VectorType&, const PreconditionerType&) [with MatrixType =
dealii::TrilinosWrappers::SparseMatrix; PreconditionerType =
dealii::TrilinosWrappers::PreconditionIC; VectorType =
dealii::TrilinosWrappers::MPI::Vector]
The violated condition was:
false
Additional information:
Iterative method reported convergence failure in step 81. The residual in the
last step was 49.0941.
This error message can indicate that you have simply not allowed a
sufficiently large number of iterations for your iterative solver to converge.
how large should be the number of iterations for it to converge?
linear solver being the same as in step-43
That depends on the size of your problem. I suspect that the problem you're
trying to solve has only 81 unknowns. In that case, your solver should
converge in fewer than 81 iterations, and if it doesn't, then your matrix is
likely either nor symmetric or not positive definite.
It also depends on the tolerance you require. If, for example, the entries of
your matrix and rhs vector are very large, then you ought to use a tolerance
that is only small *relative to these entries*. If it is small in absolute
terms, then you may not reach convergence.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: [email protected]
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 [email protected].
For more options, visit https://groups.google.com/d/optout.
