[CIG-SHORT] Question about message "fixing the coarse-level matrix dead dofs"

[Matthew Knepley]

On Sun, Mar 5, 2017 at 11:52 AM, Demian Gomez <demiang at gmail.com> wrote:

> Dear all,
>
> I'm working on refining/improving the mesh of one of my models and every
> now and then, depending on the mesh I use, I get a message just before the
> iterations of the quasi-static problem start: "[II] fixing the coarse-level
> matrix dead dofs."
>
> I've searched the message's string in Pylith's code and binaries and the
> only place where it appears is in a PETSC library binary. I also searched
> the PETSC's code but I was unable find the message at all. The problem runs
> fine, converges and the outputs are OK, but still I'd like to understand
> what this message means and if I need to take any action. Here is a more
> "in context" (but still brief) output:
>

That print is in the ML source. Very bad behavior from a library package,
but not all packages are housebroken. I
don't think there is anything we can do about this, but it is not a fatal
error.

  Thanks,

    Matt


>  >> /home/demian/pylith-2.1.0/lib/python2.7/site-packages/
> pylith/problems/Formulation.py:601:_reformResidual
>  -- implicit(debug)
>  -- [6] CPU time: 00:04:09, Memory usage: 1529.86 MB
> [II] fixing the coarse-level matrix dead dofs
>   0 KSP Residual norm 1.667629230249e-01
>   0 KSP preconditioned resid norm 1.667629230249e-01 true resid norm
> 2.680786441247e-01 ||r(i)||/||b|| 1.000000000000e+00
>   1 KSP Residual norm 1.280993223073e-01
>   1 KSP preconditioned resid norm 1.280993223073e-01 true resid norm
> 2.688535249989e-01 ||r(i)||/||b|| 1.002890498334e+00
>
> [... iterations continue]
>
> Here's the PETSC configuration I'm using:
>
> [pylithapp.petsc]
> ksp_rtol = 1.0e-8
> fs_pc_type = fieldsplit
> fs_pc_use_amat = true
> fs_pc_fieldsplit_type = multiplicative
> fs_fieldsplit_displacement_pc_type = ml
> fs_fieldsplit_lagrange_multiplier_pc_type = jacobi
> fs_fieldsplit_displacement_ksp_type = preonly
> fs_fieldsplit_lagrange_multiplier_ksp_type = preonly
>
> ksp_monitor = true
> ksp_view = true
> log_summary = true
> ksp_max_it = 800
> ksp_gmres_restart = 100
> # start_in_debugger = true
>
> snes_monitor = true
> snes_view = true
> ksp_converged_reason = true
> snes_converged_reason = true
> ksp_monitor_true_residual = true
>
> Do you guys know the meaning of this message and if I should take any
> action?
>
> Thank you,
> Demián
>
> --
> *Dr. Demián D. Gómez*
> Postdoctoral Researcher
> The Ohio State University - School of Earth Sciences
> 275 Mendenhall Laboratory
> 125 South Oval Mall
> Columbus, Ohio 43210
> Cell: +1 (901) 900-7324 <(901)%20900-7324>
> email: gomez.124 at osu.edu
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-short/attachments/20170305/ff773233/attachment.html>