Nuke not copying frames to server

55 views
Skip to first unread message

Francois Kruger

unread,
Jun 27, 2022, 2:49:42 AM6/27/22
to Royal Render Knights Tavern
Hi,

Thanks for any help in advance as always :)

I'm having issues with Nuke renders. I've taken over work that a previous dev has done, so there is a big chance that a lot of things from your original Nuke script has been tweaked. But essentially I'm trying to submit a Nuke 12 job. I noticed that Nuke 12 doesn't have a cfg file, but there is a Nuke 10 that seems to get picked up and seems to work fine-ish. As in, it picks up the job and starts the work correctly, but I'm having one issue: The renders are being saved locally on the node pc, which sounds correct according to your online docs. But the frames never get "copied to the server once done". So it seems like the job just keeps looping over the same frames over and over since it keeps thinking they haven't been finished yet.

I looked at the local outputted files, and they seem perfect. So I think the render part is going well. It's just the "copy to sever" that isn't happening.

That said, a good place to start might be for me to replace the ..\render_apps\scripts\nuke_localrenderout.py with your latest version. Where could I get a copy of all the scripts that RR uses? Would I have to reinstall RR to get it or something?

Thanks for the help
Francois Kruger

RR, Schoenberger

unread,
Jun 27, 2022, 5:28:30 AM6/27/22
to rrKn...@googlegroups.com

Hi

 

 

> But the frames never get "copied to the server once done".

There should be a list of frames at the end of the render log (rrControl, Job Log).

It states if the file was „C“ (=Copied).

 

 

 

> Nuke 12 doesn't have a cfg file, but there is a Nuke 10

 

The Nuke config file that we provide is C13__Nuke8.cfg

(As Nuke has not changed their commandline since version 8. They just added new features that we implement into our submission and render scripts).

 

If you have a Nuke 10 config file, then it might not work if you update nuke_localrenderout.py.

As we have modified the script at some time.

 

You should compare your Nuke 10 config file with the „original“ file  in subfolder _config/rr_original/C13__Nuke8.cfg first.

Then you can see what you have changed.

I would say that your own Nuke render config is not required in 90% of pipelines.

 

As you can modify submitter default settings without creating a new file
(you just have to add __inhouse.inc files http://www.royalrender.de/help/ConfigurationRenderConfigs.html).

And so far I do not know why you have a Nuke10 config, the Nuke8 config should work fine.

 

 

> Where could I get a copy of all the scripts that RR uses? Would I have to reinstall RR to get it or something?

You should update RR with all files.

There is no guarantee that files are compatible if you mix RR versions.

http://www.royalrender.de/index.php/download.html

 

 

 

regards,
Holger Schönberger

 

Please use the rrKnights Tavern
or our support system for new questions.

Francois Kruger

unread,
Jun 27, 2022, 6:30:57 AM6/27/22
to Royal Render Knights Tavern
Hi,

Okay thanks for all the info. So what I have done is to remove our Nuke 10 etc scripts. I also replace the Nuke 8 cfg and render py files with the rr_original ones. So we "should" be back to default now I think. That said, same thing is still happening. I did check, and the Nuke8 file is getting picked up now, so my updates definitely worked.

But at the end of the jobs I see this (0 render files found?)
R163| ************************* LOCAL RENDER INFORMATION *************************
R164| Total render files found:     0
R165| Files successful copied:      0
R166| Files broken/not copied:      0
R167| Total render files found (main pass only):     0
R168|  FrNr mCXRFV:   FSize  CopySpeed      Filename    ('mCXRFV' legend:  m=main  C=copied   R=successfully read   X=Exr cropped   F=image verify fail   V=rrViewer cache created)
R169| ************************* CLIENT REPORTS DURING THIS RENDER ************************* 

But just above that, I see a lot of frames getting made:
R127| .4
R128| Writing C:/RR_localdata/cachedscenes/.../scene/render_scene_0129.exr took 4.96 seconds
R129| Frame 129 (29 of 31)
R130|  
R131| .9.1.4.7
R132| Writing C:/RR_localdata/cachedscenes/.../scene/render_scene_0130.exr took 4.64 seconds
R133| Frame 130 (30 of 31)
R134|  
R135| .1.5
R136| Writing C:/RR_localdata/cachedscenes/.../render_scene_0131.exr took 4.60 seconds
R137| Frame 131 (31 of 31) 

And when I go to these folders, there are indeed frames rendered out.

Any idea what else it could be? It still doesn't seem to copy the files to the server, which is my big problem. Let me know what else I could maybe supply you with?

Thanks
Francois Kruger

RR, Schoenberger

unread,
Jun 27, 2022, 9:37:16 AM6/27/22
to rrKn...@googlegroups.com

Hi

 

The issue is that you have specified a scripted path in your write node in your nuke script.

If RR would set the local render out path, then RR renders into C:\RR_LocalData\renderout.

But your job renders to cachedscenes.

 

So either there is an issue with our localrenderout script  (not able to change your script)

OR it is not active/executed (requires the rrJob option Local Render Out and Local Scene Copy).

 


Solution:

Please disable „Local Scene Copy“ for your job.

 

 

 

Long term solution: Disable it for all new jobs.

Create a file named

C13__Nuke____global__inhouse.inc

in your config folder.

Add the line

AllowLocalSceneCopy= 1~0

Francois Kruger

unread,
Jun 28, 2022, 2:43:33 AM6/28/22
to Royal Render Knights Tavern
Hi,

Ok so we got closer. After changing the AllowLocalSceneCopy, it now outputs to the location that was send in the Nuke write node instead of on the render pc's C: drive. How do I get it to output to the supplied XML/RR output path?

Just to note, we are using our own custom write node, guessing that could have an effect? Maybe RR is trying to set the output path but it can't because it can't find the right parameter or something?

RR, Schoenberger

unread,
Jun 28, 2022, 4:17:27 AM6/28/22
to rrKn...@googlegroups.com

>How do I get it to output to the supplied XML/RR output path?

This feature was implemented in some version, but not yet enabled.

 

 

Please edit your Nuke render config and remove the # in front of

#                    <Layer -rrImageName  ….

 

Note that it will only work if

Local Scene Copy   is enabled

Local Render Out   is enabled

You submit a specific Write node (in the rrJob „Layer“)

 

 

 

 

>Just to note, we are using our own custom write node, guessing that could have an effect?

Yes, what might be an issue.


Is it a Gizmo or a real new node type?

 

If it is a new node type, there is a function:
def getAllWriteNodes():

    return getAllNodes('Write') + getAllNodes('DeepWrite') + getAllNodes('WriteGeo') + getAllNodes('WriteGeo2')

 

Then the scripts sets the parameter

writeNode["file"]

 


You might need to edit the nuke_localrenderout.py file.

 

 

regards,
Holger Schönberger

 

Please use the rrKnights Tavern
or our support system for new questions.

--
If you reply, the message is send to the user group which is sufficient and desired.
("Reply All" might send the message twice to the last author which is not required.
Replying "in private" prevents other users to see the answers and might not be seen by the single receiver if he has email rules in place.
If you want to talk to us in private, please use support@RoyalRender instead)
---
You received this message because you are subscribed to the Google Groups "Royal Render Knights Tavern" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rrKnights+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rrKnights/8e085db9-d8a7-4e8e-9d8b-282e2bcb6bc6n%40googlegroups.com.

Reply all
Reply to author
Forward
0 new messages