Did you manage to solve it in the end? I have a similar problem... Thanks!
On Monday, February 13, 2017 at 8:04:44 AM UTC+1, RAJAT ARORA wrote: > > Hello all, > > I am using deal.ii to solve a 3D solid mechanics problem along with Petsc > and P4est. The problem has 9 Dofs per node. > I run my code on stampede. > > This error may or may not be directly related to an issue with deal.ii > (may be sparsity pattern ?) or its usage. But, given a diverse community, I > was wondering if anyone has encountered an error of this type > and might be able to help me solving the issue. If the deal.ii > developers/authors feel that the post should be removed, feel free to let > me know. :) > > The issue is that I get an error (mentioned below, Info(1) = -9, Info(2) = > 188672) when I run the code on greater than 2 nodes ( ~40 processors). > There is no error on running the code on 2 nodes or less. I looked in the > manual to see what the error Info(1) = -9 means. Also, the time at which > the error happens is random. > > For Info(1) = -9, the documentation says "The main internal real/complex > workarray S too small. If INFO(2) is positive, then the number of entries > that are missing in S at the moment when the error is raised is available > in INFO(2). If INFO(2) is negative, then its absolute value should be > multiplied by 1 million. If an error > –9 occurs, the user should increase the value of ICNTL(14) before calling > the factorization > (JOB=2) again, except if ICNTL(23) or LWK USER are provided, in which case > ICNTL(23) > or LWK USER should be increased." > > I haven't changed any internal variables like ICNTL(23) or ICNTL(14). I am > using the default values by calling MUMPS from Petsc. (Honestly, I dont > understand what these variables mean as well.) > > I wanted to confirm if such an error can occur just because I increase the > number of processors to be used as it gives no error when using upto 32 > processors. > > If it indeed is happening just because of increasing the processors, is > increasing the value of ICNTL(14) still the solution ? > > Also, can someone please tell me where can I learn how to change these > internal variables ? > > Thanks a lot for the help. > > ##################### ERROR ############################## > [61]PETSC ERROR: Error in external library > [61]PETSC ERROR: Error reported by MUMPS in numerical factorization phase: > INFO(1)=-9, INFO(2)=188672 > > [61]PETSC ERROR: See http://www.mcs.anl.gov/petsc/documentation/faq.html for > trouble shooting. > [61]PETSC ERROR: Petsc Release Version 3.6.4, Apr, 12, 2016 > [61]PETSC ERROR: ./fdm on a arch-linux2-c-debug named > r542.pvt.bridges.psc.edu on Sat Feb 11 13:35:23 2017 > [61]PETSC ERROR: Configure options --with-x=0 --download-hypre=1 > --download-parmetis --download-metis --download-scalapack --download-mumps= > http://ftp.mcs.anl.gov/pub/petsc/externalpackages/MUMPS_5.0.1-p1.tar.gz > [61]PETSC ERROR: #1 MatFactorNumeric_MUMPS() line 1172 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/mat/impls/aij/mpi/mumps/mumps.c > [61]PETSC ERROR: #2 MatLUFactorNumeric() line 2958 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/mat/interface/matrix.c > [61]PETSC ERROR: #3 PCSetUp_LU() line 152 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/ksp/pc/impls/factor/lu/lu.c > [61]PETSC ERROR: #4 PCSetUp() line 983 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/ksp/pc/interface/precon.c > [61]PETSC ERROR: #5 KSPSetUp() line 332 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/ksp/ksp/interface/itfunc.c > [61]PETSC ERROR: #6 KSPSolve() line 547 in > /Code-Libraries/petsc3.6.4/petsc-3.6.4_gcc_openmpi/src/ksp/ksp/interface/itfunc.c > -- 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+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
