Re: [ice-dev] Questions about ICE-ParaView integration

[Jordan Deyton <deytonjh@xxxxxxxx>]

For what it's worth, I just tried the same thing with the same server from a client machine. With the different client, I can at least see the cone (although using the mouse throws lots of errors).

The server is running RHEL 6.6. The client that shows the cone is a Fedora 19 laptop. The client that doesn't show the cone is a Windows 7 machine.

Tangential question: How difficult would it be to get a Windows build of ParaView with all of the web components? I tried manually copying all of the python modules over, but obviously some of the VTK python scripts rely on natively compiled code. The reason I ask is because we will inevitably need to be able to integrate ParaView into ICE for Windows users, including the ability to launch/connect to a local instance of ParaView.

Thanks!
Jordan

Jordan Deyton
Oak Ridge National Laboratory
Telephone: (865) 574-1091
Email: deytonjh@xxxxxxxx

On 2/5/2015 12:08 PM, Sebastien Jourdain wrote:

The server needs to have access to X in order to do its rendering. If you don't have a X running or a graphics card, you will need to build ParaView with OSMesa (off screen mesa). But that will be much slower.

I think the issue is related to the fact that the server is not able to create a window and generate an image from that window.

I let Scott follow up with you with more details.

Seb

On Thu, Feb 5, 2015 at 9:56 AM, Jordan Deyton <deytonjh@xxxxxxxx> wrote:

Sebastien/Scott,

I can't seem to reproduce the previous error any more. However, I have encountered an entirely different problem.

First, I start off the http_pvw_server.py on a remote server (I'll call it someserver).

jordan@someserver ~/dev/x86_64/paraview/4.3.1 > bin/pvpython /path/to/http_pvw_server.py --host someserver --port 61000

I get the following message and an X window with the 6-sided cone:

Xlib: extension "NV-GLX" missing on display "localhost:12.0"

On my workstation, I kick off SwingExample.java from Eclipse. The X window gets approximately twice as big, while the Java client shows a black screen (image attached).

If I click on the window, I get all kinds of fun output from both the Eclipse console and the server. Those are also both attached.

My question is: am I launching the server correctly?
If so, any idea what could be the problem here?

Thanks,
Jordan

Jordan Deyton
Oak Ridge National Laboratory
Telephone: (865) 574-1091
Email: deytonjh@xxxxxxxx

On 2/4/2015 7:50 PM, Sebastien Jourdain wrote:

If we get more context, Scott might be able to help (in cc now).

Scott Wittenburg <scott.wittenburg@xxxxxxxxxxx>

Seb

On Wed, Feb 4, 2015 at 4:20 PM, Jay Jay Billings <jayjaybillings@xxxxxxxxx> wrote:

Jordan,

Patrick and Sebastien are our colleagues at Kitware. I have CCed them and the dev list.

Jay

On Feb 4, 2015 6:03 PM, "Jordan Deyton" <deytonjh@xxxxxxxx> wrote:

Jay/Alex,

I am currently experiencing a problem running the ParaViewWeb demo. Who is the primary POC for ParaView integration? (Also, feel free to CC this to the ice-dev list if you think that would be best.)

After much trial and error navigating available ports on our main dev server, I have figured out how to run the simple server Sebastian (?) wrote. But at the moment I get a FileNotFoundException on the Java client side.

java.io.FileNotFoundException: http://remotehost:61000/rpc-http/viewport.image.render
    at sun.net.www.protocol.http.HttpURLConnection.getInputStream0(HttpURLConnection.java:1834)
    at sun.net.www.protocol.http.HttpURLConnection.getInputStream(HttpURLConnection.java:1439)
    at com.kitware.vtk.web.VtkWebClientHttpImpl.makeRequest(VtkWebClientHttpImpl.java:74)
    at com.kitware.vtk.web.VtkWebClientHttpImpl.access$0(VtkWebClientHttpImpl.java:50)
    at com.kitware.vtk.web.VtkWebClientHttpImpl$3.call(VtkWebClientHttpImpl.java:133)
    at com.kitware.vtk.web.VtkWebClientHttpImpl$3.call(VtkWebClientHttpImpl.java:1)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
    at java.lang.Thread.run(Thread.java:745)

Jordan

--
Jordan Deyton
Oak Ridge National Laboratory
Telephone: (865) 574-1091
Email: deytonjh@xxxxxxxx