Step-77: line search fails

[Andrew McBride]

Hello deal.ii,

I’m trying to replicate the results reported in step-77. However when I run step-77 the line-search algorithm fails on the first refinement step with the error message below. I’ve tested different versions of deal and on different platforms and it’s not obvious to me what needs to be changed to get the documented results.

Any suggestions would be very helpful.



Mesh refinement step 0
Target_tolerance: 0.001

Computing residual vector... norm=0.231202
Computing Jacobian matrix
Factorizing Jacobian matrix
Solving linear system
Computing residual vector... norm=0.231202
Computing residual vector... norm=0.402211
Computing residual vector... norm=0.627191
Computing residual vector... norm=0.857696
Computing residual vector... norm=1.08878
Computing residual vector... norm=1.31995
Computing residual vector... norm=1.55112
Computing residual vector... norm=1.78231
Computing residual vector... norm=2.0135
Computing residual vector... norm=2.24469
Computing residual vector... norm=2.47588
Computing residual vector... norm=2.70708

[KINSOL ERROR] KINSol
The line search algorithm was unable to find an iterate sufficiently distinct from the current iterate.


--------------------------------------------------------
An error occurred in line <518> of file </var/folders/8z/hlb6vc015qjggytkxn84m6_c0000gn/T/heltai/spack-stage/spack-stage-dealii-9.3.0-zy7k3uwnakcqjvrajvacy5l4jrl7eaex/spack-src/source/sundials/kinsol.cc> in function
unsigned int dealii::SUNDIALS::KINSOL<dealii::Vector<double>>::solve(VectorType &) [VectorType = dealii::Vector<double>]
The violated condition was:
status >= 0
Additional information:
One of the SUNDIALS KINSOL internal functions returned a negative
error code: -5. Please consult SUNDIALS manual.

Best,
Andrew

[Wolfgang Bangerth]

On 6/14/22 09:59, Andrew McBride wrote:
> I’m trying to replicate the results reported in step-77. However when I run step-77 the line-search algorithm fails on the first refinement step with the error message below. I’ve tested different versions of deal and on different platforms and it’s not obvious to me what needs to be changed to get the documented results.

Hi Andrew,
we managed to merge a patch that breaks KINSOL in the 4-week window between
when I wrote step-77 and when 9.3.0 came out. I believe that in the latest
9.3.3 of the release branch, the offending patch was reverted, and it was also
reverted a couple of weeks ago on the current development version that will
become 9.4.0. So the question is which exact version you are using?

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/

[Andrew McBride]

Hi Wolfgang,

I was using version 9.2.0 and 9.3.0. Let me try the current version.

Thanks,
Andrew

> --
> 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 dealii+un...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/68731a3e-3b6e-ad8f-968a-0aaf26eb19f4%40colostate.edu.