[Paraview] ParaView 4.3.1 client runs X out of memory

[Jeff Becker]

On 05/22/2015 01:38 PM, Burlen Loring wrote:
> Hi Jeff,
>
> It really sounds like you are unintentionally using X forwarding. It 
> could be baked into your ssh config or a bash alias.

After some digging, it appears X forwarding is configured on in our ssh 
config. I got it to work by explicitly using ssh -x. Thanks for your help.

-jeff
>> I see the window with the gears
> if you are really doing remote rendering, you won't see the window.
>
> you may be able to confirm accidental X forward using ssh with -vv 
> option. there would be a bunch of X connection messages if it was in use.
>
> also output of glxinfo may have some useful info. When I have remote 
> rendering setup I see
>
>     name of display: :0.0
>     display: :0  screen: 0
>     direct rendering: Yes
>
> with ssh -X
>
>     name of display: localhost:10.0
>     display: localhost:10  screen: 0
>     direct rendering: No (If you want to find out why, try setting
>     LIBGL_DEBUG=verbose)
>
> Burlen
>
> On 05/22/2015 12:10 PM, Jeff Becker wrote:
>> On 05/22/2015 07:28 AM, Utkarsh Ayachit wrote:
>>> Jeff,
>>>
>>> Can you make sure the "Remote Render Threshold" is set to 0?
>> It is
>>>
>>> Also, do you have an X server running (and accessible) on the backend
>>> nodes? You're not using "ssh -X" to login to the backend nodes, right?
>>
>> I do start an X server on the backend (and I don't use ssh -X). 
>> However, when I test the configuration using glxgears,
>> I see the window with the gears but they don't turn even though the 
>> app tells me how many frames
>> it rendered every 5 seconds.
>>
>> My setup goes from my desktop (with GPU) through a front-end node 
>> (without GPU) to the backend (with GPU). Maybe some additional 
>> configuration is needed.
>>
>> It seems like I need to get glxgears working properly before I try 
>> Paraview. What should I try? Thanks.
>>
>> -jeff
>>>
>>> On Thu, May 21, 2015 at 4:54 PM, Jeff Becker 
>>> <jeffrey.c.becker at nasa.gov> wrote:
>>>> Hi. I created a VTK unstructured grid (.vtu) file for a 500 by 500 
>>>> by 500
>>>> point data set. In order to visualize it, I fire up 8 pvservers on 
>>>> my GPU
>>>> multicore backend node using:
>>>>
>>>>   mpirun -np 8 pvserver --client-host=XXX --use-offscreen-rendering
>>>>
>>>> and connect to the root server from my ParaView client. I then open 
>>>> my vtu
>>>> file and hit apply, I can see the file gets read in, and then 
>>>> nothing much
>>>> else happens except X on my client grows to fill all of memory, and my
>>>> client system (my desktop) becomes unusable. I'm guessing that the 
>>>> client is
>>>> getting overwhelmed with X requests.
>>>>
>>>> What am I doing wrong, and how to fix?
>>>>
>>>> Thanks.
>>>>
>>>> -jeff
>>>> _______________________________________________
>>>> 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
>>
>> _______________________________________________
>> 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/20150522/48aadde9/attachment.html>