After successfully running usd_viewer 0.2.0, I was postcurl -X POST https://api.example.com/stream \ -H "Content-Type: application/json" \ -d '{ "id": "usd-viewer", "profile": "default", "version": "latest" }'
it not work.
here is error log
error.log (2.2 KB)
Can you give me a little more context. This is not a lot to work with.
I’m also quite Confused. I have completed the deployment of the application, application profile, and application version. The container is usdviewer0.2.0. I followed the documentation step by step and verified on the application API doc that usdviewer has been successfully deployed.
curl -X 'GET' \
'http://IP:30975/cfg/apps' \
-H 'accept: application/json'
{
"offset": 0,
"limit": 50,
"count": 1,
"items": [
{
"id": "usd-viewer",
"name": "Omniverse USD Viewer",
"description": "View Universal Scene Description files.",
"tags": [
"USD",
"visualisation"
]
}
]
}
And it only output a message “failed to start a new stream due to a unknown error”
At very minimum your streaming API endpoint is incorrect. It should be /streaming/stream
, secondly why don’t you use GitHub - NVIDIA-Omniverse/web-viewer-sample: This sample demonstrates how a front-end client can present a streamed Omniverse Kit application and how to send messages between the two apps. and copy exact API calls as curl
as soon as you see it working. You can also use embedded docs at /streaming/docs
to involve SwaggerUI.
I have not yet configured WebRTC, so I haven’t used the webviewersample for testing. I am concerned that the issue might be due to the usd_viewer container not enabling omni.kit.streaming.webrtc, which could prevent the application pod from obtaining a session ID. I will proceed to test with the webviewsample.
nvcr.io/nvidia/omniverse/usd-viewer:0.2.0 is already configured with all required extensions to run WebRTC streaming, you can use it as is without any modification to the app itself.