[Paraview-developers] FW: [EXTERNAL] Re: ParaView | Memory consumption on Paraview (#18035)

[Moreland, Kenneth]

I believe the cell to point/point to cell filters already request an added layer of ghost cells. It's just that if no ghost cells are not available in the data (which is common), they are not automatically computed.

I believe the request is that ParaView automatically run the generate ghost cells filter under the covers if the layer of ghost cells is not available upstream. I think this is a dangerous idea. The cell to point filter is fast and takes little memory. Generating ghost cells, in contrast, is slow and will take lots of memory (at least temporarily until the operation completes). Doing an automatic ghost cell generation can create lots of surprising problems, especially when run at large scales.

-Ken


On 3/20/18, 3:08 PM, "Paraview-developers on behalf of Utkarsh Ayachit" <paraview-developers-bounces at public.kitware.com on behalf of utkarsh.ayachit at kitware.com> wrote:

    >
    >Why can’t we have the cell data to point data filter be smart enough to add
    >ghost cells if necessary for calculations, then throw them away?
    
    Unless I am missing something, no reason. Seems to me that's just the
    way it was written. I can't see why the filter can't request 1 extra
    ghost-level from its input. (cc, Berk).