Version 3.06.2 - obsolete stable Tweet

[eyeonestudio]

XPT is not rendered.
Something is wrong. There is no problem with 3.05.x

xpt.zip

(90.8 KiB) Downloaded 163 times

[brasco]

XPT is not rendered.
Something is wrong. There is no problem with 3.05.x

2017-06-19_171015.jpg

xpt.zip

That seems to work ok here:



cheers
brasc

[VVG]

XPT is not rendered.
Something is wrong. There is no problem with 3.05.x

all OK

[eyeonestudio]

XPT is not rendered.
Something is wrong. There is no problem with 3.05.x

all OK

It is not rendered in the picture viewer. Is it just me?

Video_2017-06-20_104101.wmv [ 9.22 MiB | Viewed 3577 times ]

[artdude12]

It seems to render OK from the PV if I start with frame 1.

[fuelnewmedia]

Hi aoktar,

we're running into a serious problem with 3.06. Is there anything known about such a bug?

Thanks in advance for your help, much appreciated!
Paul


Setup

Multiple Render Clients with
OS: Win10
RAM: 32 GB
GPUs: 970, 980,1070 with 4 to 8 GB RAM
C4D with Octane Plugin, Rendering via Commandline


Problem

Multiple clients crash randomly.


Octane commandline Log

CRITICAL: Stop [shader_bitmap.cpp(694)]
Scene created in plugin version -1

Rendering failed: Out of memory


Unsuccessful tries so far

Resend Only Updated Objects
Auto Detect & All Objects Movable
Check Materials Every Frame ON & OFF
Ignore Unused Material ON & OFF

Resend All Scene Data (Deprecated)

[aoktar]

Hi aoktar,

we're running into a serious problem with 3.06. Is there anything known about such a bug?

Thanks in advance for your help, much appreciated!
Paul

It says out-of-memory. That's not bug i think. What are you rendering in your scene? Volume objects?

[fuelnewmedia]

Hi aoktar!

Just static & skinned & pla-animated meshes with shaders and lots of textures plus a scatter object.
Octanes GPU stats say the scene consumes about 3.2 GB of GPU RAM, which should be perfectly ok for our render slaves - and worked in the past.

The really weird thing that the crash occurs randomly on all the clients, no matter if the GPU has 4, 6 or 8 GB RAM.

What about the logged "shader_bitmap.cpp"? Does that mean the problem occurs during the shader preparation?

May I send you the scene via pm? But be warned, its several gigs of data

Thanks a lot in advance,
Paul

[aoktar]

Hi aoktar!

Just static & skinned & pla-animated meshes with shaders and lots of textures plus a scatter object.
Octanes GPU stats say the scene consumes about 3.2 GB of GPU RAM, which should be perfectly ok for our render slaves - and worked in the past.

The really weird thing that the crash occurs randomly on all the clients, no matter if the GPU has 4, 6 or 8 GB RAM.

What about the logged "shader_bitmap.cpp"? Does that mean the problem occurs during the shader preparation?

May I send you the scene via pm? But be warned, its several gigs of data

Thanks a lot in advance,
Paul

Before trying to send it please delete some parts of objects/materials and render until error gone away. It can help to understand what's causing that. Then send me what causes a crash please.

[fuelnewmedia]

Hi aoktar,


Will try to narrow that down, but I dont think this will lead anywhere,
because the renders fail randomly.
It happens that the exact same client finishes the exact same job when restarted.

I will get back to you with a hopefully smaller but still buggy scene


Thanks for time & best regards,
Paul