[Paraview] Clarification on Paraview pvbatch run

Fri Jun 6 12:53:37 EDT 2014

The FOAM reader should read the data in parallel. Try changing the CaseType
property from Reconstructed (meaning OpenFOAM did the gather) to Decomposed
(meaning OpenFOAM left the data as the parallel nodes in the simulation
wrote them) case and see if that helps.

David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4909

On Thu, Jun 5, 2014 at 2:07 PM, Sebastien Jourdain <
sebastien.jourdain at kitware.com> wrote:

> Ok, so you are correctly using everything.
> The question is now on the *OpenFOAMReader and if it properly use MPI to
> read and distribute the data.*
> *Which unfortunately I don't know.*
>
> *If you want to force a redistribution of your data across your nodes, you
> can use the D3 filter just after your reader.*
>
> *Seb*
>
>
> On Thu, Jun 5, 2014 at 11:44 AM, Daniele Obiso <
> daniele.obiso at phitecingegneria.it> wrote:
>
>>  Hi Sebastien,
>>
>> here you can find the main parts of the script.
>>
>>
>>
>> --------------------------------------------------------------------------------------------------------------------------------
>> *try: paraview.simple*
>> *except: from paraview.simple import **
>> *paraview.simple._DisableFirstRenderCameraReset()*
>>
>> *RenderView1 = CreateRenderView()*
>>
>> *##### Lights*
>> * options *
>> *##### Screen** options*
>>
>> *##### Set scale parameters*
>> *a1_Cp_PVLookupTable = GetLookupTableForArray( "Cp", 3, RGBPoints=[-1,
>> 0.0, 0.0, 1.0, 0.75, 1.0, 0.0, 0.0], VectorMode='Magnitude',
>> NanColor=[0.498039, 0.498039, 0.498039], NumberOfTableValues=100,
>> ColorSpace='HSV', LockScalarRange=1 )*
>> *a1_Cp_PiecewiseFunction = CreatePiecewiseFunction()*
>> *ScalarBarWidgetRepresentation1 = CreateScalarBar(Position=[0.33, 0.95],
>> Title='Cp [-]', Orientation = 'Horizontal', Position2=[0.4, 0.04],
>> Enabled=1, LabelFontSize=12, LabelColor=[0.0, 0.0, 0.0],
>> LookupTable=a1_Cp_PVLookupTable, TitleFontSize=12, TitleColor=[0.0, 0.0,
>> 0.0] )*
>>
>> *##### Legend position*
>> *GetRenderView().Representations.append(ScalarBarWidgetRepresentation1)*
>>
>> *##### Time*
>> *RenderView1 = GetRenderView()*
>> *AnimationScene1 = GetAnimationScene()*
>> *AnimationScene1.EndTime = 2500*
>> *AnimationScene1.AnimationTime = 2500*
>> *RenderView1.ViewTime = 2500*
>>
>> *##### Open the model: volume mesh*
>> *foam_foam = OpenFOAMReader( FileName='../case.foam' )*
>> *foam_foam.CellArrays = ['Cp']*
>> *foam_foam.CaseType = 'Reconstructed Case'*
>> *foam_foam.MeshRegions = ['internalMesh']*
>>
>> *DataRepresentation2 = Show()*
>> *DataRepresentation2.ScalarOpacityUnitDistance = 0.01*
>> *DataRepresentation2.Representation = 'Outline'*
>> *DataRepresentation2.EdgeColor = [0.0, 0.0, 0.5]*
>> *DataRepresentation2.Visibility = 0*
>>
>> *##### Slice*
>>
>> *RenderView1.CameraViewUp = [0.0, 1.0, 0.0]*
>> *RenderView1.CameraPosition = [1.5, 0.3, 18]*
>> *RenderView1.CameraFocalPoint = [**1.5, 0.3, 1]*
>> *RenderView1.CameraClippingRange = [18, 18]*
>> *RenderView1.CameraParallelScale = 0.9*
>>
>> *SetActiveSource(foam_foam)*
>>
>> *Slice1 = Slice( SliceType="Plane" )*
>> *Slice1.SliceOffsetValues = [0.0]*
>> *Slice1.SliceType.Origin = [0, 0, 0]*
>> *Slice1.SliceType.Normal = [0, 0, 1]*
>> *Slice1.SliceType = "Plane"*
>> *DataRepresentation3 = Show()*
>> *DataRepresentation3.Visibility = 1*
>> *DataRepresentation3.EdgeColor = [0.0, 0.0, 0.5]*
>> *DataRepresentation3.ColorArrayName = 'Cp'*
>> *DataRepresentation3.LookupTable = a1_Cp_PVLookupTable*
>>
>> *##### Animation Time*
>> *RenderView1 = GetRenderView()*
>> *AnimationScene1 = GetAnimationScene()*
>> *AnimationScene1.EndTime = 2500*
>> *AnimationScene1.AnimationTime = 2500*
>> *RenderView1.ViewTime = 2500*
>> *AnimationScene1.PlayMode = 'Sequence'*
>> *AnimationScene1.NumberOfFrames =10*
>>
>> *KeyFrameAnimationCue3 = GetAnimationTrack( 'Origin', 2,
>> proxy=Slice1.SliceType)*
>> *KeyFrame1111 = CompositeKeyFrame( KeyValues=[0] )*
>> *KeyFrame2222 = CompositeKeyFrame( KeyTime=1.0, KeyValues=[1.4] )*
>> *TimeAnimationCue1 = GetTimeTrack()*
>> *TimeAnimationCue1.UseAnimationTime = 0*
>> *KeyFrame3333 = CompositeKeyFrame( KeyTime=1.0, KeyValues=[2500] )*
>> *KeyFrame4444 = CompositeKeyFrame( KeyValues=[2500] )*
>> *KeyFrameAnimationCue3.KeyFrames = [ KeyFrame1111, KeyFrame2222 ] *
>> *TimeAnimationCue1.KeyFrames = [ KeyFrame3333, KeyFrame4444 ]*
>>
>>
>> *##### Write*
>>
>> *WriteAnimation('Cp.png', Magnification=1, Quality=2, FrameRate=1.000000)*
>>
>> *##### End*
>> *Render()*
>>
>>
>> --------------------------------------------------------------------------------------------------------------------------------
>>
>> Hope this can help you to understand the problem!
>>
>> Thanks for the availability!
>> Daniele
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On 06/05/2014 07:04 PM, Sebastien Jourdain wrote:
>>
>> Is it using the VTK classes or the paraview Proxy?
>>
>> Moreover, the reader/source is responsible to distribute the data to
>> properly handle parallelism.
>> If it does not, then you right, everything will happen on the root node.
>>
>>  Hense a quick look at the script can help us determine which part need
>> to change in order to truly use parallel processing.
>>
>>  Seb
>>
>>
>> On Thu, Jun 5, 2014 at 8:32 AM, Daniele Obiso <
>> daniele.obiso at phitecingegneria.it> wrote:
>>
>>>  Hi Sebastien,
>>>
>>> thanks for the reply!
>>>
>>> The script is about creating slice and sweep animation. But it's working
>>> fine, i think there is no problem in it.
>>> Do you need the script to understand something about the parallelization?
>>>
>>> Let me know!
>>> Daniele
>>>
>>>
>>>
>>>
>>>
>>> On 06/05/2014 04:17 PM, Sebastien Jourdain wrote:
>>>
>>> Hi Daniele,
>>>
>>>  can you share what your script look like?
>>>
>>>  Seb
>>>
>>>
>>>
>>> On Thu, Jun 5, 2014 at 1:56 AM, Daniele Obiso <
>>> daniele.obiso at phitecingegneria.it> wrote:
>>>
>>>>  Hi All,
>>>>
>>>> i'm writing here because i would understand better the functionality of
>>>> pvbatch in Paraview.
>>>>
>>>> At moment i'm able to run Paraview in batch launching a python script,
>>>> and it works good.
>>>>
>>>> Here the line command i launch:
>>>> */software/Paraview/ParaView-3.12.0-Linux-x86_64/lib/paraview-3.12/mpirun
>>>> -np 6 /software/Paraview/ParaView-3.12.0-Linux-x86_64/bin/pvbatch
>>>> python_script.py *
>>>>
>>>> The problem is that monitoring at processors loads and memory
>>>> consumption, it seems that it's actually working on one single processor;
>>>> in fact if i switch to *-np 2* or even *-np1* nothing changes in terms
>>>> of computing time.
>>>> This is happening with various Paraview versions (3.10 -3.12-4.1); is
>>>> there something missing in the line command? Some other options? Are some
>>>> Paraview filters just working as serial?
>>>>
>>>>
>>>>
>>>>
>>>> Another question would be about running pvbatch on clusters without
>>>> graphic cards: i know it's possible, compiling Paraview with OSMesa, and
>>>> adding in the command line  the option
>>>> --*use-offscreen-rendering*. Could someone clarify also this point?
>>>>
>>>> Thank you all in advance!
>>>> Daniele
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Powered by www.kitware.com
>>>>
>>>> Visit other Kitware open-source projects at
>>>> http://www.kitware.com/opensource/opensource.html
>>>>
>>>> Please keep messages on-topic and check the ParaView Wiki at:
>>>> http://paraview.org/Wiki/ParaView
>>>>
>>>> Follow this link to subscribe/unsubscribe:
>>>> http://www.paraview.org/mailman/listinfo/paraview
>>>>
>>>>
>>>
>>>
>>
>>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ParaView Wiki at:
> http://paraview.org/Wiki/ParaView
>
> Follow this link to subscribe/unsubscribe:
> http://www.paraview.org/mailman/listinfo/paraview
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20140606/c250a39c/attachment-0001.html>