0000387: scotch decomposition does not work with preservePatches in 2.1.x - OpenFOAM® Bugs Reporting - Mantis

View Issue Details [ Jump to Notes ]

[ Issue History ] [ Print ]

ID

Project

Category

View Status

Date Submitted

Last Update

0000387

OpenFOAM

[All Projects] Bug

public

2012-01-13 20:15

2012-03-02 22:56

Reporter

dcombest

Assigned To

Priority

normal

Severity

major

Reproducibility

always

Status

new

Resolution

open

Platform

Linux

OS

Other

OS Version

(please specify)

Product Version

2.1.x

Target Version

Fixed in Version

Summary

0000387: scotch decomposition does not work with preservePatches in 2.1.x

Description

Almost identical to the issue in 0000241, I get the error

/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 2.1.x |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Build : 2.1.x-988050d36a43
Exec : decomposePar -cellDist
Date : Jan 13 2012
Time : 13:57:48
Host : "cloud018.seas.wustl.edu"
PID : 9098
Case : /home/research/dcombest/OpenFOAM/dcombest-2.1.x/run/packedBed/mesh4/LamBremhorstKE/2ndOrderLB-KERep5000-inletOutlet
nProcs : 1
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Disallowing user-supplied system call operations

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

--> FOAM Warning :
    From function dlOpen(const fileName&, const bool)
    in file POSIX.C at line 1175
    dlopen error : libfixedMeanValueBC.so: cannot open shared object file: No such file or directory
--> FOAM Warning :
    From function dlLibraryTable::open(const fileName&, const bool)
    in file db/dynamicLibrary/dlLibraryTable/dlLibraryTable.C at line 96
    could not load "libfixedMeanValueBC.so"
Time = 0
Create mesh

Calculating distribution of cells

Keeping owner of faces in patches
2
(
inletOutlet_half0
inletOutlet_half1
)
on same processor. This only makes sense for cyclics.
Selecting decompositionMethod metis
Constrained decomposition:
    faces with same processor owner and neighbour : 90354
    faces all on same processor : 0

ERROR: graphCheck: arc data do not match

Steps To Reproduce

Using decomposePar -cellDist on a cyclic mesh with 12 processors, preserving the cyclic halves with

preservePatches (inletOutlet_half0 inletOutlet_half1);

Additional Information

This is a pure polyhedra mesh produced using Fluent and converted in fluent3DMeshToFoam. I would not have filed this, but the mesh works and is decomposed with scotch in 2.0.x just this morning. i switched to 2.1.x on our cluster and it fails, giving the error above.

Linux version 2.6.32-131.21.1.el6.x86_64 (mockbuild@sl6.fnal.gov) (gcc version 4.4.5 20110214 (Red Hat 4.4.5-6) (GCC) ) #1 SMP Tue Nov 22 14:15:09 CST 2011

Test cases, other parallel solvers, and foamInstallationTest all pass.

Tags

No tags attached.

log.decomposePar [^] (7,937 bytes) 2012-01-14 23:31

Relationships

Notes
(0000927) mattijs (manager) 2012-01-16 09:40	For decompositionMethods that use the mesh-connectivity (scotch, ptscotch) you don't need this option - they already take cyclic connections into account.

(0000931) dcombest (reporter) 2012-01-16 20:58	Removing the preservePatches (inletOutlet_half0 inletOutlet_half1); option produces the same error. Also, using mappedPatch instead of cyclic produces the same error (maybe different bug report should be filed).

(0000933) mattijs (manager) 2012-01-17 13:53	I ran the channel395 tutorial with only sides1 cyclic and that decomposes ok (scotch) without preservePatches. (it does indeed not run with preservePatches and I'm investigating that)

(0001106) dcombest (reporter) 2012-03-02 22:56	For others that are dealing with this, a temporary workaround is to use OF-2.0.x and roll back to a previous version that didn't have this bug. From my log file above, this is build 5ae28021cd90 # rolling back your current git repo git reset --hard 5ae28021cd90 or you can clone a new 2.0.x repo and then execute this command to roll back the new repo to the correct version.

Issue History
Date Modified	Username	Field	Change
2012-01-13 20:15	dcombest	New Issue
2012-01-14 23:31	dcombest	File Added: log.decomposePar
2012-01-16 09:40	mattijs	Note Added: 0000927
2012-01-16 20:58	dcombest	Note Added: 0000931
2012-01-17 13:53	mattijs	Note Added: 0000933
2012-03-02 22:56	dcombest	Note Added: 0001106