View Issue Details

IDProjectCategoryView StatusLast Update
0001386OpenFOAMBugpublic2015-10-01 09:20
Reporteruser979Assigned Towill  
PriorityhighSeveritycrashReproducibilityalways
Status resolvedResolutionno change required 
PlatformLinux 64OSCentOSOS Version6.3
Summary0001386: OF fails after "Create mesh for time = 0"
DescriptionThe error log is following for one of the tests:

----------------------------------------------------------------
The following have been reloaded with a version change:
  1) mvapich2/1.9a2 => mvapich2/2.0b

TACC: Starting up job 3890257
TACC: Setting up parallel environment for MVAPICH2+mpispawn.
TACC: Starting parallel tasks...
/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 2.3.0 |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Build : 2.3.0-f5222ca19ce6
Exec : /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/platforms/linux64IccDPDebug/bin/pisoFoam -parallel
Date : Aug 12 2014
Time : 16:23:09
Host : "c405-403.stampede.tacc.utexas.edu"
PID : 97078
Case : /work/02898/bharat/OpenFOAM/bharat-2.3.0/rod-airfoil-new-3d-pisoFoam-les-scotch-n256
nProcs : 256
Slaves :
255
(
"c405-403.stampede.tacc.utexas.edu.97079"
"c405-403.stampede.tacc.utexas.edu.97080"
"c405-403.stampede.tacc.utexas.edu.97081"
"c405-403.stampede.tacc.utexas.edu.97082"
"c405-403.stampede.tacc.utexas.edu.97083"
.
.
.
"c411-903.stampede.tacc.utexas.edu.127990"
"c411-903.stampede.tacc.utexas.edu.127991"
)

Pstream initialized with:
    floatTransfer : 0
    nProcsSimpleSum : 0
    commsType : nonBlocking
    polling iterations : 0
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Disallowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create mesh for time = 0

[83]
[83]
[83] --> FOAM FATAL ERROR:
[83] index 1 out of range 0 ... 0
[83]
[83] From function UList<T>::checkIndex(const label)
[83] in file /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H at line 109.
[83]
FOAM parallel run aborting
[83]
[27]
[27] [137]
[27]
--> FOAM FATAL ERROR:
[27] index 6 out of range 0 ... 5[137]

[27]
[137] --> FOAM FATAL ERROR:
[27] From function [137] index 3 out of range 0 ... 2UList<T>::checkIndex(const label)
[27]
[137]
    in file /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H[137] From function at line 109.UList<T>::checkIndex(const label)

[27] [137] in file /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H
FOAM parallel run aborting
[27]
 at line 109.
[137]
FOAM parallel run aborting
[137]
[83] #0 Foam::error::printStack(Foam::Ostream&)[80]
[80]
[80] --> FOAM FATAL ERROR:
[80] index 1 out of range 0 ... 0
[80]
[80] From function UList<T>::checkIndex(const label)
[80] in file /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H at line 109.
[80]
FOAM parallel run aborting
----------------------------------------------------------------

This was a debug mode compilation and attempted with 256 processors. The error from one particular process is below:
----------------------------------------------------------------
[83]
[83]
[83] --> FOAM FATAL ERROR:
[83] index 1 out of range 0 ... 0
[83]
[83] From function UList<T>::checkIndex(const label)
[83] in file /home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H at line 109.
[83]
[83]
[83] #0 Foam::error::printStack(Foam::Ostream&)[80]
[83] #1 Foam::error::abort() at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OSspecific/POSIX/printStack.C:221
[83] #2 Foam::Ostream& Foam::operator<< <Foam::error>(Foam::Ostream&, Foam::errorManip<Foam::error>) at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OpenFOAM/lnInclude/error.C:230
[83] #3 Foam::UList<Foam::face>::checkIndex(int) const at ~/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/errorManip.H:85
[83] #4 Foam::UList<Foam::face>::operator[](int) const at ~/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H:111
[83] #5 Foam::processorPolyPatch::updateMesh(Foam::PstreamBuffers&) at ~/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/UListI.H:200
[83] #6 Foam::processorCyclicPolyPatch::updateMesh(Foam::PstreamBuffers&) at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OpenFOAM/meshes/polyMesh/polyPatches/constraint/processor/processorPolyPatch.C:396
[83] #7 Foam::polyBoundaryMesh::updateMesh() at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OpenFOAM/meshes/polyMesh/polyPatches/constraint/processorCyclic/processorCyclicPolyPatch.C:282
[83] #8 Foam::polyMesh::polyMesh(Foam::IOobject const&) at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OpenFOAM/meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C:1095
[83] #9 Foam::polyMesh::polyMesh(Foam::IOobject const&) at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/OpenFOAM/meshes/polyMesh/polyMesh.C:302
[83] #10 Foam::fvMesh::fvMesh(Foam::IOobject const&) at /work/02898/bharat/OpenFOAM/OpenFOAM-2.3.0-Debug/src/finiteVolume/fvMesh/fvMesh.C:236
[83] #11
[83] at ~/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/createMesh.H:11
[83] #12 __libc_start_main[87] at ~/OpenFOAM/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/createMesh.H:11
[83] #13
[83] in "/home1/02898/bharat/OpenFOAM/OpenFOAM-2.3.0/platforms/linux64IccDPDebug/bin/pisoFoam"
----------------------------------------------------------------

I noticed the issue "0000879: OF crashes using cyclic BC with scotch decomposition". I see that it is marked resolved but I feel that my issue is similar.
Steps To ReproduceI have attached a set of input files excluding the mesh. My unpartitioned mesh is 7GB which I created in Pointwise.

I have put the Pointwise mesh file
on http://bharatr.public.iastate.edu/pub/periodic_3D.pw

This can be opened in Pointwise to save the full 7GB mesh for OF. Also, the boundary file written by Pointwise is incomplete and the set of input files I provide in tarball contain the complete boundary file inside constant folder.

After having the complete setup, I would do decomposePar and then pisoFoam -parallel with 256 processors.

Additional InformationThe tarball also contains full log files for the crashes when the case was ran in both Opt and Debug modes.
TagsNo tags attached.

Activities

user979

2014-08-29 20:30

 

bug-report.tar.bz2 (47,743 bytes)

user979

2014-09-25 15:05

  ~0003238

I think I figured this out, how do I mark it resolved?

will

2014-10-15 11:10

manager   ~0003257

We can mark this as resolved. However, before we do, it would be nice to know what the issue was, and how you fixed it. Could you please post that information?

user979

2014-11-01 00:38

  ~0003276

The issue was using periodic boundary conditions with scotch partitioning. I got rid of that error by specifying preservePatches keyword in decomposeParDict for periodic boundaries.

Issue History

Date Modified Username Field Change
2014-08-29 20:30 user979 New Issue
2014-08-29 20:30 user979 File Added: bug-report.tar.bz2
2014-09-25 15:05 user979 Note Added: 0003238
2014-10-15 11:10 will Note Added: 0003257
2014-11-01 00:38 user979 Note Added: 0003276
2014-11-03 09:05 will Status new => resolved
2014-11-03 09:05 will Resolution open => no change required
2014-11-03 09:05 will Assigned To => will