[CIG-LONG] Wrong "Piece Extend" coordinates on field*.pvtu files

[Patrice Rey]

G'day all

I have noticed that, when running Gale-1.5 on a cluster, the field*.pvtu files are coming out with the wrong "Piece Extend" coordinates. The vtu files called in .pvtu files have the same coordinates.

...
   <Piece Extent="0 28 0 21 0 0"
             Source="fields.0.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.1.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.2.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.3.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.4.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.5.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.6.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.7.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.8.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.9.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.10.00000.vts"/>
    <Piece Extent="0 28 0 21 0 0"
             Source="fields.11.00000.vts"/> 
....

This makes loading in Paraview a bit strenuous as one has to upload all individual *.vtu files, instead of calling the *.pvtu.

The solution I am using at the moment is to edit the .pvts file, ok when running on a few CPUs, not so ok when running on many.

Interestingly, this does not happen when I run the same input script on the 8 cores of my MacPro.

I was wondering if other users have experienced the same problem?

Cheers

Patrice



Associate Professor Patrice F. Rey
EarthByte Research Group
School of Geosciences
The University of Sydney
+61 2 9351 2067
patrice.rey at mac.com
patrice.rey at sydney.edu.au