• Hardware Platform: Nvidia RTX 3060
• DeepStream Version 6.1
• NVIDIA GPU Driver Version: 510.47.03
• Programming language: Python3
• Issue Type: questions
My project is about using multiple camera at the same time. So i use stream input as rtsp stream, mux the frame with nvstreammux, process frames through nvinfer and split them with nvstreamdemux, then take the output as rtsp as well. (Ex: for 3 inputs i have 3 outputs)
The problem is when inputs are larger than 3, deepstream throws an error at the output element when set rtsp at output type, though when i change output to monitor screen directly it works normally.
Thus, i would like to know are there any way to have more than 3 outputs in rtsp’s type? or are there any type of output can achieve this in other output’s type (by using deepstream or gstreamer or any thing else)?
Sorry for my wierd-words using, my pipelines looks like this:
rtsp output: source-bin > nvstreammux > nvinfer1 > nvtracker > nvinferserver 2> streamdemux > queue> nvvidconv1> capsfilter1 > nvvidconv2 > osd > nvvidconvert3 > capsfilter2 > encoder > rtppay > udpsink
direct stream output: source-bin > nvstreammux > nvinfer1 > nvtracker > nvinferserver 2> streamdemux > queue> nvvidconv1> capsfilter1 > nvvidconv2 > osd > nveglglessink
LOGS (with 4 inputs and 4 outputs expected):
0:00:06.619007178 23 0x7f1c641bb920 WARN nvinferserver gstnvinferserver.cpp:503:gst_nvinfer_server_push_buffer: error: Internal data stream error.
0:00:06.619026910 23 0x7f1c641bb920 WARN nvinferserver gstnvinferserver.cpp:503:gst_nvinfer_server_push_buffer: error: streaming stopped, reason not-negotiated (-4)