Frame13/ayon-core
1
0
Fork 1
mirror of https://github.com/ynput/ayon-core.git synced 2026-01-01 08:24:53 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions

Documentation

Browse source
This commit is contained in:
Toke Stuart Jepsen 2023-02-03 12:16:47 +00:00
parent 97ff180d03
commit 9d8eb73835
2 changed files with 32 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

16
website/docs/artist_hosts_maya.md
View file

@ -504,6 +504,22 @@ In the scene from where you want to publish your model create *Render subset*. P
model subset (Maya set node) under corresponding `LAYER_` set under *Render instance*. During publish, it will submit this render to farm and
after it is rendered, it will be attached to your model subset.
### Tile Rendering
:::note Deadline
This feature is only supported when using Deadline. See [here](module_deadline#openpypetileassembler-plugin) for setup.
:::
On the render instance objectset you'll find:
* `Tile Rendering` - for enabling tile rendering.
* `Tile X` - number of tiles in the X axis.
* `Tile Y` - number of tiles in the Y axis.
When submittig to Deadline, you'll get:
- for each frame a tile rendering job, to render each from Maya.
- for each frame a tile assembly job, to assemble the rendered tiles.
- job to publish the assembled frames.
## Render Setups
### Publishing Render Setups

28
website/docs/module_deadline.md
View file

@ -28,16 +28,16 @@ For [AWS Thinkbox Deadline](https://www.awsthinkbox.com/deadline) support you ne
OpenPype integration for Deadline consists of two parts:
- The `OpenPype` Deadline Plug-in
- A `GlobalJobPreLoad` Deadline Script (this gets triggered for each deadline job)
- A `GlobalJobPreLoad` Deadline Script (this gets triggered for each deadline job)
The `GlobalJobPreLoad` handles populating render and publish jobs with proper environment variables using settings from the `OpenPype` Deadline Plug-in.
The `OpenPype` Deadline Plug-in must be configured to point to a valid OpenPype executable location. The executable need to be installed to
The `OpenPype` Deadline Plug-in must be configured to point to a valid OpenPype executable location. The executable need to be installed to
destinations accessible by DL process. Check permissions (must be executable and accessible by Deadline process)
- Enable `Tools > Super User Mode` in Deadline Monitor
- Go to `Tools > Configure Plugins...`, find `OpenPype` in the list on the left side, find location of OpenPype
- Go to `Tools > Configure Plugins...`, find `OpenPype` in the list on the left side, find location of OpenPype
executable. It is recommended to use the `openpype_console` executable as it provides a bit more logging.
- In case of multi OS farms, provide multiple locations, each Deadline Worker goes through the list and tries to find the first accessible
@ -45,12 +45,16 @@ executable. It is recommended to use the `openpype_console` executable as it pro
![Configure plugin](assets/deadline_configure_plugin.png)
### OpenPypeTileAssembler Plugin
To setup tile rendering copy the `OpenPypeTileAssembler` plugin to the repository;
`[OpenPype]\openpype\modules\deadline\repository\custom\plugins\OpenPypeTileAssembler` > `[DeadlineRepository]\custom\plugins\OpenPypeTileAssembler`
## Troubleshooting
#### Publishing jobs fail directly in DCCs
- Double check that all previously described steps were finished
- Check that `deadlinewebservice` is running on DL server
- Check that `deadlinewebservice` is running on DL server
- Check that user's machine has access to deadline server on configured port
#### Jobs are failing on DL side
@ -61,40 +65,40 @@ Each publishing from OpenPype consists of 2 jobs, first one is rendering, second
- Jobs are failing with `OpenPype executable was not found` error
Check if OpenPype is installed on the Worker handling this job and ensure `OpenPype` Deadline Plug-in is properly [configured](#configuration)
Check if OpenPype is installed on the Worker handling this job and ensure `OpenPype` Deadline Plug-in is properly [configured](#configuration)
- Publishing job is failing with `ffmpeg not installed` error
OpenPype executable has to have access to `ffmpeg` executable, check OpenPype `Setting > General`
![FFmpeg setting](assets/ffmpeg_path.png)
- Both jobs finished successfully, but there is no review on Ftrack
Make sure that you correctly set published family to be send to Ftrack.
Make sure that you correctly set published family to be send to Ftrack.
![Ftrack Family](assets/ftrack/ftrack-collect-main.png)
Example: I want send to Ftrack review of rendered images from Harmony :
- `Host names`: "harmony"
- `Families`: "render"
- `Families`: "render"
- `Add Ftrack Family` to "Enabled"
Make sure that you actually configured to create review for published subset in `project_settings/ftrack/publish/CollectFtrackFamily`
![Ftrack Family](assets/deadline_review.png)
Example: I want to create review for all reviewable subsets in Harmony :
Example: I want to create review for all reviewable subsets in Harmony :
- Add "harmony" as a new key an ".*" as a value.
- Rendering jobs are stuck in 'Queued' state or failing
Make sure that your Deadline is not limiting specific jobs to be run only on specific machines. (Eg. only some machines have installed particular application.)
Check `project_settings/deadline`
![Deadline group](assets/deadline_group.png)
Example: I have separated machines with "Harmony" installed into "harmony" group on Deadline. I want rendering jobs published from Harmony to run only on those machines.