PETSC ERROR: Not for unassembled matrix
Hello,
In one of the fvSolution files for the lid-driven cavity on the HPC repo, an option is used to cache the preconditioner and coefficient matrix. I have tried to replicate this fvSolution setup on a different case, but so far I am met with PETSC errors:
PIMPLE: iteration 1
DILUPBiCGStab: Solving for Ux, Initial residual = 2.462362568e-07, Final residual = 5.123302044e-13, No Iterations 1
DILUPBiCGStab: Solving for Uy, Initial residual = 2.39719708e-05, Final residual = 4.514665586e-13, No Iterations 2
DILUPBiCGStab: Solving for Uz, Initial residual = 1.706417187e-05, Final residual = 3.769441156e-13, No Iterations 2
Initializing PETSc
[0]PETSC ERROR: --------------------- Error Message --------------------------------------------------------------
[0]PETSC ERROR: Object is in wrong state
[0]PETSC ERROR: Not for unassembled matrix
[0]PETSC ERROR: See https://petsc.org/release/faq/ for trouble shooting.
[0]PETSC ERROR: Petsc Release Version 3.16.4, unknown
[0]PETSC ERROR: Unknown Name on a linux64GccDPInt64Opt named sdumont6165 by juan.salazar Thu Feb 10 13:32:52 2022
[0]PETSC ERROR: Configure options --force --with-64-bit-indices=0 --with-precision=double --with-debugging=0 --COPTFLAGS=-O3 --CXXOPTFLAGS=-O3 --FOPTFLAGS=-O3 PETSC_ARCH=linux64GccDPInt64Opt --with-blaslapack-dir=/scratch/app_sequana/intel-oneapi/2021.1.0-2659/mkl/2021.1.1 --with-mkl_sparse --with-mkl_sparse_optimize --with-mpi-dir=/scratch/app_sequana/openmpi/2.1.1 --download-hypre
[0]PETSC ERROR: #1 MatGetRowSum() at /scratch/song/juan.salazar/sequana/openfoam/ThirdParty-v2012/petsc/src/mat/interface/matrix.c:5159
[0]PETSC ERROR: #2 MatMult() at /scratch/song/juan.salazar/sequana/openfoam/ThirdParty-v2012/petsc/src/mat/interface/matrix.c:2425
Any guidance is very much appreciated.