I have several apps on jetson tx2 that are communicating with each other. They are using TensorRT, Qt and other libs.
I have configured autostart using autologin and startup application and its working, but I think its not the “right way” for deployment.
What I need:
- Autostart Qt app without login/desktop
- Customize nvidia icon on boot
- Make system safer from “hacking”
What I’ve tried:
- Make systemd scripts (doesn’t work for qt and cuda apps)
- This method https://devtalk.nvidia.com/default/topic/988275/tx1-disable-ubuntu-unity-and-start-qt-app-on-startup/ (doesn’t work too)
Has anyone had the same problems? Could you give me some advice for devOps technicians or tell your experience for tx2 project deployment?
Thanks and Best Regards.