Can't run a test pipeline inside a container on a Jetson using SSH

Accelerated Computing Intelligent Video Analytics DeepStream SDK
,
1

• Hardware Platform (Jetson / GPU) Jetson Orin Nano
• DeepStream Version 7.0
• JetPack Version (valid for Jetson only) 6.1
• TensorRT Version 8.6.2

I am trying to run a Deepstream application inside this container nvcr.io/nvidia/deepstream:7.0-triton-multiarch but I am getting these errors.

image
image1457×609 92.8 KB

The test pipeline runs just fine outside the container on another Jetson with same Deepstream version.
The pipeline consists of reading an mp4 file, making inferences using YoloV8 and saving another mp4 file with the inferences.
These are the plugin initializations for reference:

    source = Gst.ElementFactory.make("filesrc", "file-source")
    demuxer = Gst.ElementFactory.make("qtdemux", "qtmux-0")
    h264parser = Gst.ElementFactory.make("h264parse", "h264-parser")
    decoder = Gst.ElementFactory.make("nvv4l2decoder", "nvv4l2-decoder")
    streammux = Gst.ElementFactory.make("nvstreammux", "Stream-muxer")
    pgie = Gst.ElementFactory.make("nvinfer", "primary-inference")
    nvvidconv = Gst.ElementFactory.make("nvvideoconvert", "nvvideo-converter")
    nvosd = Gst.ElementFactory.make("nvdsosd", "onscreendisplay")
    nvvidconv2 = Gst.ElementFactory.make("nvvideoconvert", "nvvideo-converter2")
    encoder = Gst.ElementFactory.make("x264enc", "encoder")
    codeparser = Gst.ElementFactory.make("h264parse", "h264-parser2")
    container = Gst.ElementFactory.make("qtmux", "qtmux-1")
    sink = Gst.ElementFactory.make("filesink", "file-sink")

I am accessing both Jetsons by SSH. The first one was for test purposes that’s why I ran the test pipeline outside the container. Now I need to replicate it for other machines and we will need to use a container to make replication easier.

3

Could you try to run unset DISPLAY first?

4

I don’t get that error anymore if I unset DISPLAY but I am now getting this

image
image1553×709 134 KB

Not sure if it is related to the past error. I was following this GitHub - marcoslucianops/DeepStream-Yolo: NVIDIA DeepStream SDK 7.0 / 6.4 / 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 / 5.1 implementation for YOLO models to implement YoloV8 in Deepstream. With that I generate both the onnx file and custom-bbox-parser function.

Edit1: This jetson is running Jetpack 6.1, not sure if it matters though. The other one was running Jetpack 6.0

5

If you are using Jetpack 6.1, please use our DeepStream 7.1 version.

1 Like
6

It seems to have been a mismatch between Jetpack and Deepstream versions. I switched Jetpack to 6.0 and used Deepstream 7.0 container. It works now.

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.