Am I required to use port 9001 for my Jetson edge deployment?

Community Help
1

The documentation recommends that I utilize port 9001 to access my Jetson device after pulling the image.

base64 YOUR_IMAGE.jpg | curl -d @- \
"http://localhost:9001/your-model/42?api_key=YOUR_KEY"

Will modifying the access port using the -p flag lead to any unexpected behavior down the line?

Curious to know if there was a specific reasoning for port 9001.

Asking as my work environment may have firewall restrictions on certain ports that I’ll need to navigate around.

2

We picked 9001 as a little inside joke.

There is not currently a way to change it, but we would welcome a PR that passes a port in as an optional environment variable and passes it to app.listen in server/index.js here: GitHub - roboflow-ai/inference-server: Object detection inference with Roboflow Train models on NVIDIA Jetson devices.