Based on the gstreamer documentation, it sounds like if we put too many buffers into a gstreamer queue and those buffers aren't being pulled out of the queue fast enough, the push_buffer thread will be stalled. This has the potential of blocking a whole mmids workflow if this occurs.
One way this can theoretically be done is if a video or audio stream is passed into a basic_transcode that expects audio and video to come through. It sounds like this may be blocked due to one pipeline staying in paused mode until data comes through.
Based on the gstreamer documentation, it sounds like if we put too many buffers into a gstreamer queue and those buffers aren't being pulled out of the queue fast enough, the
push_bufferthread will be stalled. This has the potential of blocking a whole mmids workflow if this occurs.One way this can theoretically be done is if a video or audio stream is passed into a
basic_transcodethat expects audio and video to come through. It sounds like this may be blocked due to one pipeline staying in paused mode until data comes through.