Deepstream with RTSP errors

AI & Data Science Computer Vision & Image Processing
, , ,
1

Hello,
I have implemented a deepstream pipeline based on python code examples.
In a lot of cases, when I am connected to the RTSP stream, G-stream pipeline receives an error which makes error in my pipeline that eventually stops working. As I learned in Nvidia deepstream python examples, I am able to recognize that and restart those sources, and indeed system starts to work once again. but this takes some time (recognize error, wait, restart pipeline ) and therefore makes a lot of missed data.
This happens even if a stream has a small error I have tested the same stream with FFmpeg and VLC player, and both of them continue to play, and only my pipeline is stuck until it resets itself.
Is there any option to set G-streamer to ignore errors or set some kind of buffer to prevent small errors from breaking the pipeline

Thanks in advance

2

I’m having the same issue with some cameras with slow bandwidth, when using the rtsp plugin, it didn’t happened with the uridecodebin plugin, I don’t know why this happens.
As @sabi is explaining, this makes that with every error there is a delay while the pipeline is not working, and with the uridecodebin we have other kind of issues in the pipeline, so it is not the solution.