[petsc-users] Convergence trouble with SNES and CG

[Matthew Knepley]

On Mon, Feb 22, 2016 at 6:02 PM, David Knezevic <david.knezevic at akselos.com>
wrote:

> On Mon, Feb 22, 2016 at 6:57 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
>
>>
>>   Add -ksp_converged_reason
>>
>
>
> That gives "DIVERGED_INDEFINITE_PC". The full output is below:
>

Just for testing, replace CG with GMRES.

  Thanks,

     Matt


>
>   105 KSP Residual norm 7.099534889984e-05
>   106 KSP Residual norm 6.615351528474e-05
>   107 KSP Residual norm 6.303646443688e-05
>   Linear solve converged due to CONVERGED_RTOL iterations 107
>   NL step  5, |residual|_2 = 1.996487e+01
>     0 KSP Residual norm 5.844456247091e+00
>     1 KSP Residual norm 1.308756943973e+00
>     2 KSP Residual norm 7.545133187463e-01
>     3 KSP Residual norm 1.346068942607e+00
>   Linear solve did not converge due to DIVERGED_INDEFINITE_PC iterations 4
> SNES Object: 1 MPI processes
>   type: newtonls
>   maximum iterations=50, maximum function evaluations=10000
>   tolerances: relative=1e-08, absolute=1e-50, solution=1e-08
>   total number of linear solver iterations=535
>   total number of function evaluations=11
>   norm schedule ALWAYS
>   SNESLineSearch Object:   1 MPI processes
>     type: bt
>       interpolation: cubic
>       alpha=1.000000e-04
>     maxstep=1.000000e+08, minlambda=1.000000e-12
>     tolerances: relative=1.000000e-08, absolute=1.000000e-15,
> lambda=1.000000e-08
>     maximum iterations=40
>   KSP Object:   1 MPI processes
>     type: cg
>     maximum iterations=10000, initial guess is zero
>     tolerances:  relative=1e-05, absolute=1e-50, divergence=10000
>     left preconditioning
>     using NATURAL norm type for convergence test
>   PC Object:   1 MPI processes
>     type: ilu
>       ILU: out-of-place factorization
>       0 levels of fill
>       tolerance for zero pivot 2.22045e-14
>       matrix ordering: natural
>       factor fill ratio given 1, needed 1
>         Factored matrix follows:
>           Mat Object:           1 MPI processes
>             type: seqaij
>             rows=967608, cols=967608
>             package used to perform factorization: petsc
>             total: nonzeros=7.55539e+07, allocated nonzeros=7.55539e+07
>             total number of mallocs used during MatSetValues calls =0
>               using I-node routines: found 322536 nodes, limit used is 5
>     linear system matrix = precond matrix:
>     Mat Object:    ()     1 MPI processes
>       type: seqaij
>       rows=967608, cols=967608
>       total: nonzeros=7.55539e+07, allocated nonzeros=7.55539e+07
>       total number of mallocs used during MatSetValues calls =0
>         using I-node routines: found 322536 nodes, limit used is 5
> Number of nonlinear iterations: 5
> Nonlinear solver convergence/divergence reason: DIVERGED_LINEAR_SOLVE
>
>
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20160222/419858e1/attachment.html>