Deploy video workflows on the edge using Azure IoT and AVA

Hi , thanks for the detailed blog.

I'm stuck at step 1. Create the model server (bring my own model) and test it with an image.

I tried to create a Yolov3 server and call the server in the pipeline but it's not working.

I follow your instruction and deploy the yolov3.onnx to the Container Registry "$CONTAINER_REGISTRY_USERNAME_myacr.azurecr.io/avaextension".

Then I use the Yolov3 pipeline provided in https://docs.microsoft.com/en-us/azure/azure-video-analyzer/video-analyzer-docs/edge/analyze-live-video-use-your-model-http?pivots=programming-language-python

In short, the topology includes: 

"yolov3": 
{ 
 "version": "1.0", "type": "docker", "status": "running", "restartPolicy": "always",  "settings": { 
   "image": "$CONTAINER_REGISTRY_USERNAME_myacr.azurecr.io/avaextension:latest",    "createOptions": {} 
 } 
} 

Then, the live pipeline calls the server: 

{ "name": "httpAIServerAddress", "value": "http://yolov3/score" } 

When activating the live pipeline, it couldn't find http://yolov3/score:

"body": { "code": "nameResolutionError", "target": "http://yolov3/score" } 

Could you please help me to identify the issues? Thank you

KaushikRoy