[petsc-users] How to force the code run with a given number of iterations?

[Fande Kong]

Thanks,

I got this message  *Linear solve converged due to CONVERGED_RTOL
iterations 2*.

What is the definition of RTOL in petsc?

I doubt the RTOL is small.*


*


On Wed, Oct 30, 2013 at 6:38 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> Fande Kong <fd.kong at siat.ac.cn> writes:
>
> > Hi all,
> >
> > How to force the code run with a given number of iterations?
> >
> > I use the following script to run code, but the code stop after doing
> > *two*iterations:
> >
> > * mpirun -n 2 ./linearElasticity  -ksp_type gmres -pc_type asm
> > -pc_asm_overlap 1 -sub_pc_type ilu -sub_pc_factor_levels 2 -sub_ksp_type
> > preonly -sub_pc_factor_mat_ordering_type rcm -ksp_view - -ksp_rtol 1e-5
> > -ksp_max_it 500 -ksp_monitor -ksp_monitor_singular_value*
> >
> > The residual history is listed as follows:
> >
> > *  0 KSP Residual norm 9.732477596672e+31
> >   0 KSP Residual norm 9.732477596672e+31 % max 1.000000000000e+00 min
> > 1.000000000000e+00 max/min 1.000000000000e+00
> >   1 KSP Residual norm 8.532845325800e+31
> >   1 KSP Residual norm 8.532845325800e+31 % max 4.821161757157e+55 min
> > 4.821161757157e+55 max/min 1.000000000000e+00
> >   2 KSP Residual norm 3.648279603061e+23
> >   2 KSP Residual norm 3.648279603061e+23 % max 1.234581012401e+79 min
> > 2.318814661396e+55 max/min 5.324190125905e+23
> > *
> >  The residual is not small enough to stop.
>
> -ksp_converged_reason will tell you why it exited early.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20131030/1aff9bbf/attachment.html>