No subject


Mon Aug 10 18:29:21 EDT 2009 



i tried running the same data using np1 until np10=2C and i get the followi=
ng reading for still render=2C the rest of the readings are quite consisten=
t=2C which makes me wonder why the time (still render) decrease initially b=
ut increase eventually? shouldn't it become much faster when more nodes are=
 used? all my server nodes are identical to each other (intel core 2 duo=2C=
 2.66 GHz)



ParaView doesn't use parallelism to scale in terms of processing speed=2C i=
t uses it to scale in terms of achievable data size. That means=2C if you h=
ave sufficient memory on one machine=2C processing the data on two machines=
 isn't likely to give you any speed up and will usually slow it down becaus=
e of the extra code and especially communication that has to happen in the =
parallel program (Ahmdahls law). If the problem is too big for one machine=
=2C then the only alternative is to split the data and run it on many machi=
nes (Gustafsen's law). That said=2C if the data is nearly to big for one ma=
chine=2C for instance when swapping=2C then splitting it and running it on =
several machines will make it faster for a while until the overhead starts =
to dominate the performance.

=20


----------------------------------------------







number of process used=20
still render (seconds)

np1
7.16043

np2
3.93390

np3
3.05784

np4
3.02900

np5
3.02851

np6
3.04962

np7
3.43479

np8
3.47883

np9
3.71554

np10
3.80835
----------------------------------------------

=20
appreciate your reply!

regards=2C
chewping




From: kmorel at sandia.gov
To: lcp81um at msn.com=3B paraview at paraview.org
CC: kitware at kitware.com
Date: Mon=2C 28 Sep 2009 08:37:08 -0600
Subject: Re: [Paraview] How to interpret timer log




Both scenarios are wrong.  ParaView will not push out data from process 0 t=
o processes 1-3 unless you explicitly run a filter that does that (or the r=
eader does that internally=2C but I know of no such reader).  What is actua=
lly happening is more along the lines of:



Processes 0-3 each read in a partition of data from the file.=20
Each process extracts polygonal geometry from their local data.=20
Per your settings=2C ParaView decides to send the geometry to the client.  =
The data is collected to process 0 and sent to the client.

The reason you are not seeing vtkFileSeriesReader on all of the servers is =
that there is a threshold in the timer log to not show anything that execut=
es under a certain amount of time (by default 0.01 seconds).  If you change=
 the Time Threshold to Show All=2C you should be able to see everything tha=
t executes=2C even if it completes immediately.

You should note that how readers read partitions is determined by the reade=
r itself.  Many of the readers do not really handle partitioned reading.  T=
hus=2C the reader will do something na=EFve like read everything on process=
 0.  Based on your timings=2C that is probably what is happening to you.  T=
hat is=2C processes 1-3 probably have empty data.  You never specified what=
 format of data you are reader=2C so I cannot answer the data completely.  =
However=2C if you want to know how your data is partitioned on the server (=
at least=2C before rendering)=2C you can run the Process Ids filter.

-Ken


On 9/24/09 9:09 PM=2C "chew ping" <lcp81um at msn.com> wrote:


Hi all=2C
=20
I'm doing parallel rendering using 1 client (dual core laptop) and 2 cluste=
r servers (dual core desktop)
below is the timer log result i collected when i run: mpirun -np 4 pvserver
=20
---------------------------------------------------------------------------=
--------------------
Local Process
Still Render=2C 3.029 seconds
Execute vtkMPIMoveData id: 457=2C 1.98248 seconds
=20
=20
Server=2C Process 0
Execute vtkFileSeriesReader id: 176=2C 0.637821 seconds
Execute vtkMPIMoveData id: 457=2C 1.49186 seconds
Dataserver gathering to 0=2C 0.829525 seconds
Dataserver sending to client=2C 0.661658 seconds

Server=2C Process 1
Execute vtkMPIMoveData id: 457=2C 0.141821 seconds
Dataserver gathering to 0=2C 0.141544 seconds

Server=2C Process 2
Execute vtkMPIMoveData id: 457=2C 0.243584 seconds
Dataserver gathering to 0=2C 0.243318 seconds

Server=2C Process 3
Execute vtkMPIMoveData id: 457=2C 0.191589 seconds
Dataserver gathering to 0=2C 0.191303 seconds

---------------------------------------------------------------------------=
--------------------------
=20
i have difficulty interpreting the timer log=2C my guess is:
=20
Scenario 1:=20
Process 0 reads the whole data=2C disseminate the dats into 4 pieces=2C the=
n distribute to itself and Process 1&2&3=2C
each node will process the data and send it back Process 0=2C
Process 0 gather all data and send it back to client=2C
client renders the data
=20
Scenario 2:
Process 0 reads the whole data=2C distribute the whole data to Process 0&1&=
2&3=2C
each node will 'take' their own piece of data to process=2C then send it ba=
ck Process 0=2C
Process 0 gather all data and send it back to client=2C
client renders the data
  =20
Which scenario is the correct one? or both are wrong?
=20
is there any resources i could refer to find what does it mean by: Execute =
vtkFileSeriesReader=2C Execute vtkMPIMoveData?
=20
any help / feedback is highly appreciated!
thanks!
=20
regards=2C
chewping
=20
=20
=20
=20
      =20


   ****      Kenneth Moreland
    ***      Sandia National Laboratories
*********** =20
*** *** ***  email: kmorel at sandia.gov
**  ***  **  phone: (505) 844-8919
    ***      web:   http://www.cs.unm.edu/~kmorel


_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensour=
ce/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paravi=
ew.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview


 		 	   		  =

--_06166292-e530-4a4b-a8f7-bdf39d0cb381_
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<style><!--
.hmmessage P
{
margin:0px=3B
padding:0px
}
body.hmmessage
{
font-size: 10pt=3B
font-family:Verdana
}
--></style>
</head>
<body class=3D'hmmessage'>
Hi all=2C&nbsp=3B<BR>
&nbsp=3B<BR>
thanks for the reply below.<BR>
&nbsp=3B<BR>
my friend and i each run parallel rendering using our own notebook as clien=
t=2C but we use the same data=2C same servers and same&nbsp=3Bversion of Pa=
raView=2C CMake=2C QT=2C MPI.&nbsp=3BWe use same settings and did the same =
thing=2C however our&nbsp=3Btimer log results are a little bit different.<B=
R>
&nbsp=3B<BR>
<STRONG>My timer log result</STRONG><BR>
----------------------------------------------------------------------<BR><=
SPAN lang=3DEN>
Local Process<BR>
--- Disable display lists.<BR>
--- Disable triangle strips.<BR>
Still Render=2C 0.043315 seconds<BR>
Still Render=2C 0.021299 seconds<BR>
Still Render=2C 0.024769 seconds<BR>
Still Render=2C 0.02782 seconds<BR>
Still Render=2C 10.7477 seconds<BR>
Execute vtkMPIMoveData id: 462=2C 3.73372 seconds<BR>
&nbsp=3B<BR>
&nbsp=3B<BR>
Server=2C Process 0<BR>
Execute vtkFileSeriesReader id: 182=2C 0.641624 seconds<BR>
Execute vtkPVGeometryFilter id: 248=2C 0.005231 seconds<BR>
Execute vtkPVCacheKeeper id: 459=2C 6.9e-05 seconds<BR>
Execute vtkMPIMoveData id: 462=2C 3.3546 seconds<BR>
Dataserver gathering to 0=2C 0.831371 seconds<BR>
Dataserver sending to client=2C 2.5226 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 0.000142 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 1<BR>
Execute vtkFileSeriesReader id: 182=2C 0.000345 seconds<BR>
Execute vtkPVGeometryFilter id: 248=2C 0.005231 seconds<BR>
Execute vtkPVCacheKeeper id: 459=2C 5.7e-05 seconds<BR>
Execute vtkMPIMoveData id: 462=2C 0.14266 seconds<BR>
Dataserver gathering to 0=2C 0.142351 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 9.8e-05 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 2<BR>
Execute vtkFileSeriesReader id: 182=2C 0.000325 seconds<BR>
Execute vtkPVGeometryFilter id: 248=2C 0.004975 seconds<BR>
Execute vtkPVCacheKeeper id: 459=2C 6.3e-05 seconds<BR>
Execute vtkMPIMoveData id: 462=2C 0.244045 seconds<BR>
Dataserver gathering to 0=2C 0.243799 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 9.7e-05 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 3<BR>
Execute vtkFileSeriesReader id: 182=2C 0.000408 seconds<BR>
Execute vtkPVGeometryFilter id: 248=2C 0.00521 seconds<BR>
Execute vtkPVCacheKeeper id: 459=2C 5.6e-05 seconds<BR>
Execute vtkMPIMoveData id: 462=2C 0.192467 seconds<BR>
Dataserver gathering to 0=2C 0.192195 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 0.000113 seconds<BR>
</SPAN>&nbsp=3B<BR>
-----------------------------------------------------------------------<BR>
&nbsp=3B<BR>
<STRONG>My friend's timer log result</STRONG><BR>
&nbsp=3B<BR>
-----------------------------------------------------------------------<BR>=
<SPAN lang=3DEN>
Local Process<BR>
Still Render=2C 3.98618 seconds<BR>
&nbsp=3B<BR>
&nbsp=3B<BR>
Server=2C Process 0<BR>
Execute vtkFileSeriesReader id: 2927=2C 1.18069 seconds<BR>
Execute vtkPVGeometryFilter id: 2992=2C 0.041116 seconds<BR>
Execute vtkPVCacheKeeper id: 3203=2C 6.7e-05 seconds<BR>
Execute vtkMPIMoveData id: 3206=2C 0.000195 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 0.000138 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 0.000101 seconds<BR>
Execute vtkPolyDataMapper id: 2853=2C 5.5e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 4.5e-05 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 1<BR>
Execute vtkFileSeriesReader id: 2927=2C 0.000275 seconds<BR>
Execute vtkPVGeometryFilter id: 2992=2C 0.019066 seconds<BR>
Execute vtkPVCacheKeeper id: 3203=2C 6.4e-05 seconds<BR>
Execute vtkMPIMoveData id: 3206=2C 0.000123 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 8.9e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 7.3e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2853=2C 5.5e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 4.8e-05 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 2<BR>
Execute vtkFileSeriesReader id: 2927=2C 0.000303 seconds<BR>
Execute vtkPVGeometryFilter id: 2992=2C 0.02033 seconds<BR>
Execute vtkPVCacheKeeper id: 3203=2C 5.3e-05 seconds<BR>
Execute vtkMPIMoveData id: 3206=2C 0.000158 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 9.6e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 8.9e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2853=2C 5.2e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 4.7e-05 seconds<BR>
&nbsp=3B<BR>
Server=2C Process 3<BR>
Execute vtkFileSeriesReader id: 2927=2C 0.000311 seconds<BR>
Execute vtkPVGeometryFilter id: 2992=2C 0.02033 seconds<BR>
Execute vtkPVCacheKeeper id: 3203=2C 6e-05 seconds<BR>
Execute vtkMPIMoveData id: 3206=2C 0.000117 seconds<BR>
Execute vtkOrderedCompositeDistributor =2C 9e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 8.3e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2853=2C 5.2e-05 seconds<BR>
Execute vtkPolyDataMapper id: 2998=2C 4.7e-05 seconds<BR></SPAN>
&nbsp=3B<BR>
--------------------------------------------------------------------------<=
BR>
<BR>My question are:<BR>
&nbsp=3B<BR>
my local process has <SPAN lang=3DEN>Execute vtkMPIMoveData=2C but my frien=
d's doesn't=2C why?</SPAN><BR>
<SPAN lang=3DEN>my process 0 has <SPAN lang=3DEN>Dataserver gathering to 0=
=2C Dataserver sending to client=2C but my friend's doesn't=2C why?</SPAN><=
/SPAN><BR>
<SPAN lang=3DEN><SPAN lang=3DEN>my process 1=2C2=2C3 has <SPAN lang=3DEN>Da=
taserver gathering to 0=2C but my friend's doesn't=2C why?<BR></SPAN></SPAN=
>
</SPAN>&nbsp=3B<BR>
Appreciate your reply=2C thanks!<BR>
&nbsp=3B<BR>
regards=2C<BR>
chew ping<BR>
&nbsp=3B<BR>
<HR id=3DstopSpelling>
From: dave.demarle at kitware.com<BR>Date: Wed=2C 30 Sep 2009 09:27:32 -0400<B=
R>Subject: Re: [Paraview] How to interpret timer log<BR>To: lcp81um at msn.com=
<BR>CC: kmorel at sandia.gov=3B paraview at paraview.org=3B kitware at kitware.com<B=
R><BR>I counted more than one question (six in fact) in that last email. :)=
<BR>
<DIV class=3Decxgmail_quote>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">does these mean t=
hat=2C at the client's end=2C it took 3.029 seconds to renders the geometry=
 send back from process 0?<BR></FONT></SPAN></FONT></DIV></BLOCKQUOTE>
<DIV><BR>I believe the 3.029 seconds includes the transfer time=2C but I ma=
y be incorrect on that. Someone please correct me if I am wrong.<BR>&nbsp=
=3B<BR></DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">what does still r=
ender means? </FONT></SPAN></FONT></DIV></BLOCKQUOTE>
<DIV><BR>While you drag the mouse and move the camera paraview does interac=
tive renders. Iteractive renders will (depending on data and the preference=
 settings) use subsampled geometry and subsampled images. Once you release =
the mouse=2C ParaView does a still render at full resolution.<BR><BR>&nbsp=
=3B</DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">how about the thi=
rd line? after still render=2C why does the MPI need to move data? </FONT><=
/SPAN></FONT></DIV></BLOCKQUOTE>
<DIV><BR>See answer to question 1=2C the MPI time (and processing) is being=
 counted as part of the rendering. Since it is "part of" the rendering=2C i=
t shows up after in the log (even if it happened first).<BR></DIV>
<DIV>&nbsp=3B</DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">which data? </FON=
T></SPAN></FONT></DIV></BLOCKQUOTE>
<DIV><BR>Geometry (the surface of whatever your data is).<BR>&nbsp=3B<BR></=
DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">to where?<BR></FO=
NT></SPAN></FONT></DIV></BLOCKQUOTE>
<DIV><BR>From the root node of the data server to the client.<BR><BR></DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial"><BR>i tried runni=
ng the same data using np1 until np10=2C and i get the following reading fo=
r still render</FONT></SPAN></FONT><FONT style=3D"FONT-SIZE: 10pt" size=3D2=
><SPAN style=3D"FONT-SIZE: 10pt"><FONT face=3D"Tahoma=2C Verdana=2C Helveti=
ca=2C Arial">=2C the rest of the readings are quite consistent=2C which mak=
es me wonder why the time (still render) decrease initially but increase ev=
entually? shouldn't it become much faster when more nodes are used? all my =
server nodes are identical to each other (intel core 2 duo=2C 2.66 GHz)<BR>=
</FONT></SPAN></FONT><BR></DIV></BLOCKQUOTE>
<DIV><BR>ParaView doesn't use parallelism to scale in terms of processing s=
peed=2C it uses it to scale in terms of achievable data size. That means=2C=
 if you have sufficient memory on one machine=2C processing the data on two=
 machines isn't likely to give you any speed up and will usually slow it do=
wn because of the extra code and especially communication that has to happe=
n in the parallel program (Ahmdahls law). If the problem is too big for one=
 machine=2C then the only alternative is to split the data and run it on ma=
ny machines (Gustafsen's law). That said=2C if the data is nearly to big fo=
r one machine=2C for instance when swapping=2C then splitting it and runnin=
g it on several machines will make it faster for a while until the overhead=
 starts to dominate the performance.<BR><BR>&nbsp=3B</DIV>
<BLOCKQUOTE class=3Decxgmail_quote style=3D"PADDING-LEFT: 1ex">
<DIV><FONT style=3D"FONT-SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10p=
t"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial"><BR>-------------=
---------------------------------<BR></FONT></SPAN></FONT>
<TABLE height=3D187 cellSpacing=3D0 rules=3Dnone width=3D268 border=3D0>
<COLGROUP>
<COL width=3D176>
<COL width=3D145></COLGROUP>
<TBODY>
<TR>
<TD align=3Dleft width=3D176 height=3D17>number of process used </TD>
<TD align=3Dleft width=3D145>still render (seconds)</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np1</TD>
<TD align=3Dright>7.16043</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np2</TD>
<TD align=3Dright>3.93390</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np3</TD>
<TD align=3Dright>3.05784</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np4</TD>
<TD align=3Dright>3.02900</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np5</TD>
<TD align=3Dright>3.02851</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np6</TD>
<TD align=3Dright>3.04962</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np7</TD>
<TD align=3Dright>3.43479</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np8</TD>
<TD align=3Dright>3.47883</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np9</TD>
<TD align=3Dright>3.71554</TD></TR>
<TR>
<TD align=3Dleft height=3D17>np10</TD>
<TD align=3Dright>3.80835</TD></TR></TBODY></TABLE><BR><FONT style=3D"FONT-=
SIZE: 10pt" size=3D2><SPAN style=3D"FONT-SIZE: 10pt"><FONT face=3D"Tahoma=
=2C Verdana=2C Helvetica=2C Arial">----------------------------------------=
------<BR><BR>&nbsp=3B<BR>appreciate your reply!<BR><BR>regards=2C<BR>chewp=
ing<BR><BR><BR></FONT></SPAN></FONT>
<HR>
From: <A href=3D"mailto:kmorel at sandia.gov">kmorel at sandia.gov</A><BR>To: <A =
href=3D"mailto:lcp81um at msn.com">lcp81um at msn.com</A>=3B <A href=3D"mailto:pa=
raview at paraview.org">paraview at paraview.org</A><BR>CC: <A href=3D"mailto:kit=
ware at kitware.com">kitware at kitware.com</A><BR>Date: Mon=2C 28 Sep 2009 08:37=
:08 -0600<BR>Subject: Re: [Paraview] How to interpret timer log
<DIV>
<DIV></DIV>
<DIV class=3Dh5><BR><BR><FONT face=3D"Calibri=2C Verdana=2C Helvetica=2C Ar=
ial"><SPAN style=3D"FONT-SIZE: 11pt">Both scenarios are wrong. &nbsp=3BPara=
View will not push out data from process 0 to processes 1-3 unless you expl=
icitly run a filter that does that (or the reader does that internally=2C b=
ut I know of no such reader). &nbsp=3BWhat is actually happening is more al=
ong the lines of:<BR><BR></SPAN></FONT>
<OL>
<LI><FONT face=3D"Calibri=2C Verdana=2C Helvetica=2C Arial"><SPAN style=3D"=
FONT-SIZE: 11pt">Processes 0-3 each read in a partition of data from the fi=
le. </SPAN></FONT></LI>
<LI><FONT face=3D"Calibri=2C Verdana=2C Helvetica=2C Arial"><SPAN style=3D"=
FONT-SIZE: 11pt">Each process extracts polygonal geometry from their local =
data. </SPAN></FONT></LI>
<LI><FONT face=3D"Calibri=2C Verdana=2C Helvetica=2C Arial"><SPAN style=3D"=
FONT-SIZE: 11pt">Per your settings=2C ParaView decides to send the geometry=
 to the client. &nbsp=3BThe data is collected to process 0 and sent to the =
client.<BR></SPAN></FONT></LI></OL><FONT face=3D"Calibri=2C Verdana=2C Helv=
etica=2C Arial"><SPAN style=3D"FONT-SIZE: 11pt"><BR>The reason you are not =
seeing vtkFileSeriesReader on all of the servers is that there is a thresho=
ld in the timer log to not show anything that executes under a certain amou=
nt of time (by default 0.01 seconds). &nbsp=3BIf you change the Time Thresh=
old to Show All=2C you should be able to see everything that executes=2C ev=
en if it completes immediately.<BR><BR>You should note that how readers rea=
d partitions is determined by the reader itself. &nbsp=3BMany of the reader=
s do not really handle partitioned reading. &nbsp=3BThus=2C the reader will=
 do something na=EFve like read everything on process 0. &nbsp=3BBased on y=
our timings=2C that is probably what is happening to you. &nbsp=3BThat is=
=2C processes 1-3 probably have empty data. &nbsp=3BYou never specified wha=
t format of data you are reader=2C so I cannot answer the data completely. =
&nbsp=3BHowever=2C if you want to know how your data is partitioned on the =
server (at least=2C before rendering)=2C you can run the Process Ids filter=
.<BR><BR>-Ken<BR><BR><BR>On 9/24/09 9:09 PM=2C "chew ping" &lt=3B<A href=3D=
"http://msn.com/">lcp81um at msn.com</A>&gt=3B wrote:<BR><BR></SPAN></FONT>
<BLOCKQUOTE><FONT size=3D2><FONT face=3D"Verdana=2C Helvetica=2C Arial"><SP=
AN style=3D"FONT-SIZE: 10pt">Hi all=2C<BR>&nbsp=3B<BR>I'm doing parallel re=
ndering using 1 client (dual core laptop) and 2 cluster servers (dual core =
desktop)<BR>below is the timer log result i collected when i run: mpirun -n=
p 4 pvserver<BR>&nbsp=3B<BR>-----------------------------------------------=
------------------------------------------------<BR></SPAN></FONT><SPAN sty=
le=3D"FONT-SIZE: 10pt"><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Aria=
l">Local Process<BR>Still Render=2C 3.029 seconds<BR>Execute vtkMPIMoveData=
 id: 457=2C 1.98248 seconds<BR>&nbsp=3B<BR>&nbsp=3B<BR>Server=2C Process 0<=
BR>Execute vtkFileSeriesReader id: 176=2C 0.637821 seconds<BR>Execute vtkMP=
IMoveData id: 457=2C 1.49186 seconds<BR>Dataserver gathering to 0=2C 0.8295=
25 seconds<BR>Dataserver sending to client=2C 0.661658 seconds<BR></FONT><F=
ONT face=3D"Verdana=2C Helvetica=2C Arial"><BR></FONT><FONT face=3D"Tahoma=
=2C Verdana=2C Helvetica=2C Arial">Server=2C Process 1<BR>Execute vtkMPIMov=
eData id: 457=2C 0.141821 seconds<BR>Dataserver gathering to 0=2C 0.141544 =
seconds<BR></FONT><FONT face=3D"Verdana=2C Helvetica=2C Arial"><BR></FONT><=
FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">Server=2C Process 2<B=
R>Execute vtkMPIMoveData id: 457=2C 0.243584 seconds<BR>Dataserver gatherin=
g to 0=2C 0.243318 seconds<BR></FONT><FONT face=3D"Verdana=2C Helvetica=2C =
Arial"><BR></FONT><FONT face=3D"Tahoma=2C Verdana=2C Helvetica=2C Arial">Se=
rver=2C Process 3<BR>Execute vtkMPIMoveData id: 457=2C 0.191589 seconds<BR>=
Dataserver gathering to 0=2C 0.191303 seconds<BR></FONT><FONT face=3D"Verda=
na=2C Helvetica=2C Arial"><BR>---------------------------------------------=
--------------------------------------------------------<BR>&nbsp=3B<BR>i h=
ave difficulty interpreting the timer log=2C my guess is:<BR>&nbsp=3B<BR>Sc=
enario 1: <BR>Process 0 reads the whole data=2C disseminate the dats into 4=
 pieces=2C then distribute to itself and Process 1&amp=3B2&amp=3B3=2C<BR>ea=
ch node will process the data and send it back Process 0=2C<BR>Process 0 ga=
ther all data and send it back to client=2C<BR>client renders the data<BR>&=
nbsp=3B<BR>Scenario 2:<BR>Process 0 reads the whole data=2C distribute the =
whole data to Process 0&amp=3B1&amp=3B2&amp=3B3=2C<BR>each node will 'take'=
 their own piece of data to process=2C then send it back Process 0=2C<BR>Pr=
ocess 0 gather all data and send it back to client=2C<BR>client renders the=
 data<BR>&nbsp=3B&nbsp=3B&nbsp=3B<BR>Which scenario is the correct one? or =
both are wrong?<BR>&nbsp=3B<BR>is there any resources i could refer to find=
 what does it mean by: Execute vtkFileSeriesReader=2C Execute vtkMPIMoveDat=
a?<BR>&nbsp=3B<BR>any help / feedback is highly appreciated!<BR>thanks!<BR>=
&nbsp=3B<BR>regards=2C<BR>chewping<BR>&nbsp=3B<BR>&nbsp=3B<BR>&nbsp=3B<BR>&=
nbsp=3B<BR>&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B<BR></FO=
NT></SPAN></FONT></BLOCKQUOTE><FONT size=3D2><SPAN style=3D"FONT-SIZE: 10pt=
"><FONT face=3D"Verdana=2C Helvetica=2C Arial"><BR></FONT><FONT face=3D"Con=
solas=2C Courier New=2C Courier"><BR>&nbsp=3B&nbsp=3B&nbsp=3B**** &nbsp=3B&=
nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3BKenneth Moreland<BR>&nbsp=3B&nbsp=3B&nbsp=3B=
&nbsp=3B*** &nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3BSandia National Laborat=
ories<BR>*********** &nbsp=3B<BR>*** *** *** &nbsp=3Bemail: <A href=3D"http=
://sandia.gov/">kmorel at sandia.gov</A><BR>** &nbsp=3B*** &nbsp=3B** &nbsp=3B=
phone: (505) 844-8919<BR>&nbsp=3B&nbsp=3B&nbsp=3B&nbsp=3B*** &nbsp=3B&nbsp=
=3B&nbsp=3B&nbsp=3B&nbsp=3Bweb: &nbsp=3B&nbsp=3B<A href=3D"http://www.cs.un=
m.edu/~kmorel">http://www.cs.unm.edu/~kmorel</A><BR></FONT></SPAN></FONT><F=
ONT face=3D"Calibri=2C Verdana=2C Helvetica=2C Arial"><SPAN style=3D"FONT-S=
IZE: 11pt"><BR></SPAN></FONT></DIV></DIV></DIV><BR>________________________=
_______________________<BR>Powered by <A href=3D"http://www.kitware.com/">w=
ww.kitware.com</A><BR><BR>Visit other Kitware open-source projects at <A hr=
ef=3D"http://www.kitware.com/opensource/opensource.html">http://www.kitware=
.com/opensource/opensource.html</A><BR><BR>Please keep messages on-topic an=
d check the ParaView Wiki at: <A href=3D"http://paraview.org/Wiki/ParaView"=
>http://paraview.org/Wiki/ParaView</A><BR><BR>Follow this link to subscribe=
/unsubscribe:<BR><A href=3D"http://www.paraview.org/mailman/listinfo/paravi=
ew">http://www.paraview.org/mailman/listinfo/paraview</A><BR><BR></BLOCKQUO=
TE></DIV><BR> 		 	   		  </body>
</html>=

--_06166292-e530-4a4b-a8f7-bdf39d0cb381_--

More information about the ParaView mailing list