OTOY Forums

Thanks Lewis, i will try that.

Does this will impact negatively the performance and loading time for the GPU?

Best

https://martinpatoine.myportfolio.com/work

Yes loading time is slower but render speed is same, but if stability is issue it's worth try.

Thanks Lewis!

I did some new renders since turning off multithreaded scene extraction and no more crash for the moment.
I will do more testing next week after my actual 1200 frames to renders.....

Best

Hi Lewis!

Now i can confirm that turning off multithreaded scene extraction has solved my crashing issues!

Thanks Lewis!!!

Goid to hear

Hi again Lewis,

I am really confused my situation is special i think.
I had many issues with Lightwave and Octane Render.
Today i think i have resolved almost all my problems.
The source of my problems was Comodo Antivirus!!!
Now i did uninstalled Comodo and Lightwave, Modeler, and Octane all works perfectly!
First day since many weeks without hangs or crash on Lw, modeler or Octane even running Octane with Multithreaded Extraction all the day with no crash!!!

So my conclusion is something wrong with Comodo...

But now i did uninstalled Comodo i need another firewall so i did installed Tiny Firewall and seems good till now..

But a new thing happen.
I have 3 gpu on another computer that now have rendering error:

OptiX error for device 2: [ERROR] "sbt->raygenRecord" is null
OptiX API error for device 2: Invalid value
device 2: path tracing kernel failed
detected that all GPUs of the render node have failed -> restarting render node

Any idea?
In the project many new textures edit, some model edits, no plugins used except Octane.
The only new addition to the project is for the fist time in this project i have created a Compound node for a surface!


I have reduced my project size significantly since last few days but this problem now...

Hi,

I use regular Windows firewall/Defender so not really needed any 3rd party ones so far.

As for optic crashes it could be your GPU clock/overheating so check with MSI Afterburner what are clocks and try to underclock GPU to see.

Oh! got my first crash today M-Extraction on.
So i re-uncheck M-Extraction!

I will try MSI.

Could it be possible this is about TDR delays?

Thanks again

I found MT extraction tricky and xause of most crashes but also behave differently on different CPU and chipsets. I have 3 different nachines and one is having problems wit MTE like 50% of all tries while other maybe 10-15% and yet my workstation crashes with MTE maybe 5% of times.

But MTE also can sometimes not load all layers so it's buggy feature in general .

Hope LWD can fix it so if you have crash scenes would be good to send them that..

Well i did tried MSI but not success..

I think now my problem is related to the recent big changes in my windows system:
- I have uninstalled Comodo and set windows defender to replace the virus detections.
- I have installed and configured TinyWall and since my gpu on the network failed...
- So now i did uninstalled TinyWall and set back Windows Firewall

The problem persist even after rebooting multiple times...

Maybe my scene is now taking more memory...
My workstation have RTX3090 24 gig
My renderfarm have 3 RTX2080 TI 11 gig

I set Out-of-core on the workstation at 16 gig and the renderfarm at 48 gig but the renderfarm still saying not enough memory and a new error message!

CUDA error 201 on device 0: invalid device context -> could not get memory info..