On Fri, 8 Jan 2010, Kevin.Buckley at ecs.vuw.ac.nz wrote: > Hello again, > > this follows on somewhat from my earliar posting regarding > initial experiences building and running PETSc for the > NetBSD platform. > > Having built both PETSc and the application that uses it > for our NetBSD hosted OpenMPI/SGE grid, I have observed > that a large number of file descriptors are seemingly > remaining open on the compute nodes within the grid when > running the application. > > I have determined that the file to which these descriptors > point is the one that is being passed to PETSc's "-save_to" > option.
There is no -save_to option in PETSc. Perhaps its an option defined/used by the appliation code. > > I appreciate this may be something that's only being seen > on my platform but I'd like to remove the issue, if only to > be tidy. > > In case there's some knowledge that might speed up my search > for a solution out there: > > 1) are there/have there been/ any known issues with file descriptors > not being closed within PETSc, for that file or otherwise, that > would inform this issue ? Don't know how to check for unclosed file descriptors. But perhaps you can check this by running an example code - say src/snes/examples/tutorials/ex19.c [one of the runex19 targets in the makefile] Satish > > 2) which /routines/source code files/ would be the places to start > looking for clues ? > > We are running PETSc-3.0.0-p7. > > Kevin > >
