[Paraview] ParaviewWeb with remote, distributed data

[Sven Kramer]

I could improve the lag by either distributing the MPI processes to another
node or using just a single process. I think the framerate is fine for
larger datasets. Only for my tiny (~20 triangles) dataset I was slightly
disappointed, because I expected more fluent user interaction.

Back to my initial, second question on live rendering of incoming data: I
found that this is probably best handled by Catalyst, isn't it? Then I will
try study how to include the Catalyst Adaptors and Pipelines into my code,
and I will probably come back with a new post asking about some Catalyst
details.

Thank you
Sven

2016-05-10 0:11 GMT+02:00 Sebastien Jourdain <sebastien.jourdain at kitware.com
>:

> Hi Scen,
>
> that's a correct way, but I'm not sure to know why it is that much slower.
> Wondering if all the pvserver are fighting for the same GPU and hence
> waiting more than anything... ?
> You may be able to change the remote rendering threshold to make sure only
> pvpython is doing the rendering.
>
> Seb
>
> On Mon, May 9, 2016 at 3:34 PM, Sven Kramer <svenkramer40 at gmail.com>
> wrote:
>
>> Hello Seb,
>> I got a working connection from pv_web_visualizer.py to a couple of
>> pvservers (started with mpirun on <hostname>) by just giving "--ds-host
>> <hostname> --ds-port 11111" from the link you mentioned. Is this correct,
>> without --rs-host? It is working in principle, but the renderview update
>> rate is not more than 2 updates/second for a tiny datasource even when
>> everything is running on the same machine.
>>
>> Thank you
>> Sven
>>
>> 2016-05-06 22:26 GMT+02:00 Sebastien Jourdain <
>> sebastien.jourdain at kitware.com>:
>>
>>> Hi Sven,
>>>
>>> 1) Yes, assuming you are using pv_web_visualizer.py you can look here:
>>>
>>>
>>> https://github.com/Kitware/ParaView/blob/master/Web/Applications/Visualizer/server/pv_web_visualizer.py#L18-L28
>>>
>>> 2) The server can push notification to the client. Need to slightly edit
>>> the client and server to match your needs.
>>>
>>> Seb
>>>
>>>
>>> On Fri, May 6, 2016 at 2:17 PM, Sven Kramer <svenkramer40 at gmail.com>
>>> wrote:
>>>
>>>> Dear all,
>>>> I am currently running a VTK application, which is visualizing large
>>>> grids distributed on a MPI cluster.
>>>> My task is to migrate this application to the Web. But I cannot manage
>>>> more than what is described at
>>>> http://www.paraview.org/ParaView3/Doc/Nightly/www/js-doc/index.html#!/guide/quick_start,
>>>> that is I can serve a local data directory through the provided webapps.
>>>>
>>>> I have two essential questions that stop me from getting started:
>>>> 1) How do I connect the pvpython server to remote data/render servers
>>>> just like I can do in Paraview?
>>>> 2) All ParaviewWeb examples seem to react on webclient requests. Our
>>>> application is processing data as they come in, and also the visualization
>>>> pipeline is frequently updated depending on an automatic analysis of the
>>>> arriving data. Can the newly rendered datasets be pushed to the webclient,
>>>> or does the webclient have to poll for available datasets?
>>>>
>>>> Thank you
>>>> Sven
>>>>
>>>> _______________________________________________
>>>> 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
>>>>
>>>> Search the list archives at: http://markmail.org/search/?q=ParaView
>>>>
>>>> Follow this link to subscribe/unsubscribe:
>>>> http://public.kitware.com/mailman/listinfo/paraview
>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20160511/b700751d/attachment.html>