Re: unstructured mesh generated by vorocrust

59 views
Skip to first unread message

Mahzari, Pedram

unread,
Feb 26, 2024, 5:53:18 PMFeb 26
to pflotr...@googlegroups.com
Hi Glenn,

Enclosed please find boundary file as well as the input deck. I have attempted to send zip files but bounced back. 

Apologies if I changed the format to txt as other formats could not be delivered.  

Best regards,
Pedram Mahzari | Jacobs | Subsurface Flow Modelling Consultant

Pedram....@jacobs.com
HQ Building, Thomson Avenue, Harwell Campus | Didcot, OXON, OX11 0GD | United Kingdom

 



NOTICE - This communication may contain confidential and privileged information that is for the sole use of the intended recipient. Any viewing, copying or distribution of, or reliance on this message by unintended recipients is strictly prohibited. If you have received this message in error, please notify us immediately by replying to the message and deleting it from your computer.
bottom_bc.txt
model.txt
east_bc.txt
north_bc.txt
south_bc.txt
top_bc.txt
west_bc.txt

Hammond, Glenn E

unread,
Feb 26, 2024, 5:58:09 PMFeb 26
to pflotr...@googlegroups.com

Still need the mesh.h5 file.

 

From: 'Mahzari, Pedram' via pflotran-dev <pflotr...@googlegroups.com>
Sent: Monday, February 26, 2024 2:53 PM
To: pflotr...@googlegroups.com
Subject: [pflotran-dev: 6235] Re: unstructured mesh generated by vorocrust

 

Check twice before you click! This email originated from outside PNNL.

 

--
You received this message because you are subscribed to the Google Groups "pflotran-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pflotran-dev...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pflotran-dev/LO4P265MB66183D2B24C6037203E355E2E65A2%40LO4P265MB6618.GBRP265.PROD.OUTLOOK.COM.

Mahzari, Pedram

unread,
Feb 28, 2024, 11:26:04 AMFeb 28
to pflotr...@googlegroups.com
Hi PFLOTRAN team,

I have tried to send over my files. I'll be grateful if you could confirm that my files are delivered.

Best regards,
Pedram. 


From: Mahzari, Pedram <Pedram....@jacobs.com>
Sent: 26 February 2024 22:53
To: pflotr...@googlegroups.com <pflotr...@googlegroups.com>
Subject: Re: unstructured mesh generated by vorocrust
 

Bisht, Gautam

unread,
Feb 28, 2024, 11:30:49 AMFeb 28
to pflotr...@googlegroups.com, Pedram....@jacobs.com

Hi Pedram,

 

Maybe you were somehow left out of Glenn’s last email.

 

-Gautam

 

Mahzari, Pedram

unread,
Feb 28, 2024, 11:39:18 AMFeb 28
to Bisht, Gautam, pflotr...@googlegroups.com
Hi Gautam,

Thanks for letting me know. The mesh file was large and the mailbox could not accept zip files. I have uploaded the mesh file onto OneDrive. I hope this helps.


Best regards,
Pedram. 
From: Bisht, Gautam <gautam...@pnnl.gov>
Sent: 28 February 2024 16:30
To: pflotr...@googlegroups.com <pflotr...@googlegroups.com>; Mahzari, Pedram <Pedram....@jacobs.com>
Subject: [EXTERNAL] Re: unstructured mesh generated by vorocrust
 

Mahzari, Pedram

unread,
Feb 29, 2024, 6:00:55 AMFeb 29
to Bisht, Gautam, pflotr...@googlegroups.com

Hi Gautam and PFLOTRARN team,

 

Apologies if I have sent numerous repetitive emails. Can you please confirm you have received the required files?

 

Also, we have tried vorocrust meshes with TOUGH3 1.0V where PETSc version was very old and, it failed due to ParMetis routine failure. The Berkely team reran the model with a recent PETSc (version 3.2) and, it ran smoothly. Can it be a similar issue with PFLOTRAN?

 

Best regards,

Pedram.     

Glenn Hammond

unread,
Feb 29, 2024, 11:55:55 AMFeb 29
to pflotran-dev
Pedram,

The below is screen output from the simulation when PFLOTRAN is compiled with PETSc configured for debugging. Notice the statement , "Row 0 has diagonal entry; Parmetis forbids diagonal entry". When I inspect your mesh.h5 file, there is a connection between Cell 1 and Cell 1 (see connection 9 in 0-based indexing viewed with HDFView). Cell 4 is connected to itself at connection 35, and this goes on for 5, 6, etc. There is a bug in your mesh. Cells cannot be connected to themselves.

Glenn

...
pflotran card:: STRATA
pflotran card:: STRATA
pflotran card:: END_SUBSURFACE
[0]PETSC ERROR: --------------------- Error Message --------------------------------------------------------------
[0]PETSC ERROR: Invalid argument
[0]PETSC ERROR: Row 0 has diagonal entry; Parmetis forbids diagonal entry
[0]PETSC ERROR: WARNING! There are unused option(s) set! Could be the program crashed before usage or a spelling mistake, etc!
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_auto_scale_max value: 2.e8 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_auto_scale_multiphase value: TRUE source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_delta0 value: 0.1 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_deltaM value: 0.5 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta1 value: 1.e-3 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta2 value: 0.25 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta3 value: 0.75 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_t1 value: 0.25 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_t2 value: 2.0 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_tol value: 1.e-13 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_use_cauchy value: TRUE source: code
[0]PETSC ERROR: See https://petsc.org/release/faq/ for trouble shooting.
[0]PETSC ERROR: Petsc Release Version 3.20.2, unknown
[0]PETSC ERROR: /home/hamm495/software/pflotran/src/pflotran/pflotran on a gnu-c-debug named sugarbee by hamm495 Thu Feb 29 08:48:07 2024
[0]PETSC ERROR: Configure options --PETSC_ARCH=gnu-c-debug --with-cc=mpicc --with-fc=mpif90 --with-cxx=mpicxx --COPTFLAGS="-g -O0" --CXXOPTFLAGS="-g -O0" --FOPTFLAGS="-g -O0 -Wno-unused-function" --with-clanguage=c --with-shared-libraries=no --with-debugging=yes --download-hdf5=yes --with-hdf5-fortran-bindings=yes --with-valgrind=yes --download-parmetis=yes --download-metis=yes --download-hypre=yes --download-fblaslapack=yes --with-c2html=no
[0]PETSC ERROR: #1 MatPartitioningApply_Parmetis_Private() at /home/hamm495/software/lib/petsc-3.20/src/mat/partition/impls/pmetis/pmetis.c:78
[0]PETSC ERROR: #2 MatPartitioningApply_Parmetis() at /home/hamm495/software/lib/petsc-3.20/src/mat/partition/impls/pmetis/pmetis.c:228
[0]PETSC ERROR: #3 MatPartitioningApply() at /home/hamm495/software/lib/petsc-3.20/src/mat/partition/partition.c:312
[0]PETSC ERROR: #4 /home/hamm495/software/pflotran/src/pflotran/grid_unstructured_aux.F90:1072
[0]PETSC ERROR: --------------------- Error Message --------------------------------------------------------------
[0]PETSC ERROR:   It appears a new error in the code was triggered after a previous error, possibly because:
[0]PETSC ERROR:   -  The first error was not properly handled via (for example) the use of
[0]PETSC ERROR:      PetscCall(TheFunctionThatErrors()); or
[0]PETSC ERROR:   -  The second error was triggered while handling the first error.
[0]PETSC ERROR:   Above is the traceback for the previous unhandled error, below the traceback for the next error
[0]PETSC ERROR:   ALL ERRORS in the PETSc libraries are fatal, you should add the appropriate error checking to the code
[0]PETSC ERROR: Invalid pointer
[0]PETSC ERROR: Invalid Pointer to PetscObject: Argument 'part' (parameter # 1)
[0]PETSC ERROR: WARNING! There are unused option(s) set! Could be the program crashed before usage or a spelling mistake, etc!
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_auto_scale_max value: 2.e8 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_auto_scale_multiphase value: TRUE source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_delta0 value: 0.1 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_deltaM value: 0.5 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta1 value: 1.e-3 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta2 value: 0.25 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_eta3 value: 0.75 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_t1 value: 0.25 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_t2 value: 2.0 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_tol value: 1.e-13 source: code
[0]PETSC ERROR:   Option left: name:-tran_snes_trdc_use_cauchy value: TRUE source: code
[0]PETSC ERROR: See https://petsc.org/release/faq/ for trouble shooting.
[0]PETSC ERROR: Petsc Release Version 3.20.2, unknown
[0]PETSC ERROR: /home/hamm495/software/pflotran/src/pflotran/pflotran on a gnu-c-debug named sugarbee by hamm495 Thu Feb 29 08:48:07 2024
[0]PETSC ERROR: Configure options --PETSC_ARCH=gnu-c-debug --with-cc=mpicc --with-fc=mpif90 --with-cxx=mpicxx --COPTFLAGS="-g -O0" --CXXOPTFLAGS="-g -O0" --FOPTFLAGS="-g -O0 -Wno-unused-function" --with-clanguage=c --with-shared-libraries=no --with-debugging=yes --download-hdf5=yes --with-hdf5-fortran-bindings=yes --with-valgrind=yes --download-parmetis=yes --download-metis=yes --download-hypre=yes --download-fblaslapack=yes --with-c2html=no
[0]PETSC ERROR: #1 ISPartitioningToNumbering() at /home/hamm495/software/lib/petsc-3.20/src/vec/is/is/utils/iscoloring.c:523
[0]PETSC ERROR: #2 /home/hamm495/software/pflotran/src/pflotran/grid_unstructured_aux.F90:1138
Segmentation fault (core dumped)

Mahzari, Pedram

unread,
Mar 5, 2024, 9:44:37 PMMar 5
to pflotr...@googlegroups.com, Bisht, Gautam

Hi PFLOTRAN team,

 

I wondered if you had a chance to review the issue of using the unstructured mesh (generated by vorocrust) with multi-processors.

Bisht, Gautam

unread,
Mar 5, 2024, 10:50:02 PMMar 5
to Mahzari, Pedram, pflotr...@googlegroups.com

Hi Pedram,

 

Glenn found the issue with the mesh and replied a few days ago and maybe you were mistakenly left out of the email thread. Glenn’s response is available at https://groups.google.com/g/pflotran-dev/c/7msfgZsLgOM/m/Mgm6TX2jAQAJ.

 

-Gautam

 

 

From: Mahzari, Pedram <Pedram....@jacobs.com>
Date: Tuesday, March 5, 2024 at 6:44 PM
To: pflotr...@googlegroups.com <pflotr...@googlegroups.com>
Cc: Bisht, Gautam <gautam...@pnnl.gov>
Subject: RE: unstructured mesh generated by vorocrust

Hi PFLOTRAN team,

 

I wondered if you had a chance to review the issue of using the unstructured mesh (generated by vorocrust) with multi-processors.

 

Best regards,

Pedram.  

 

From: Mahzari, Pedram
Sent: Thursday, February 29, 2024 11:01 AM
To: Bisht, Gautam <gautam...@pnnl.gov>; pflotr...@googlegroups.com
Subject: RE: unstructured mesh generated by vorocrust

 

Hi Gautam and PFLOTRARN team,

 

Apologies if I have sent numerous repetitive emails. Can you please confirm you have received the required files?

 

Also, we have tried vorocrust meshes with TOUGH3 1.0V where PETSc version was very old and, it failed due to ParMetis routine failure. The Berkely team reran the model with a recent PETSc (version 3.2) and, it ran smoothly. Can it be a similar issue with PFLOTRAN?

 

Best regards,

Pedram.     

 

From: Mahzari, Pedram <Pedram....@jacobs.com>
Sent: Wednesday, February 28, 2024 4:39 PM
To: Bisht, Gautam <gautam...@pnnl.gov>; pflotr...@googlegroups.com
Subject: Re: unstructured mesh generated by vorocrust

 

Hi Gautam,

 

Thanks for letting me know. The mesh file was large and the mailbox could not accept zip files. I have uploaded the mesh file onto OneDrive. I hope this helps.

 

pedram mahzari

unread,
Mar 6, 2024, 8:34:15 AMMar 6
to pflotran-dev
Hi Glenn,

Apologies for not checking this group. 

I removed the dodgy connections and now, the model runs using multiple processors. Many thanks for reviewing the mesh file. 

Cheers,
Pedram. 

Reply all
Reply to author
Forward
0 new messages