Which plug-in outputs an error called "batch size Not 1"?

Hi, I’m working on a Deepstream-app based,
and I created a UI plug-in based on the dsexample plug-in and configured the pipeline.

My sauce is CSI camera and it works at 30 fps.

The UI plug-in uses Opencv to draw the UI.

When drawing takes a long time or overlaps with nvosd, an error called “batch size Not 1” suddenly occurs.

I think this comes from nvoverlaysink, but this is my guess and it’s probably printed from g_print in the development environment.

When an error occurs, the fps are fixed to zero shortly and the app stops

Unlike other error logs, the plug-in’s information is not included.
Only “ERROR: Batch size not 1” is printed.

I wonder if you know anything about this error and if an error occurs if 30 fps is not met due to the delay of certain plug-ins in the pipeline.

log

    Nov 16 18:42:26 localhost user_kw_timer.sh[3273]: **PERF: 29.98 (29.99)        29.98 (30.00)`
    Nov 16 18:42:31 localhost user_kw_timer.sh[3273]: **PERF: 29.99 (29.99)        29.99 (30.00)
    Nov 16 18:42:31 localhost user_kw_timer.sh[3273]: ERROR: Batch size not 1
    Nov 16 18:42:36 localhost user_kw_timer.sh[3273]: **PERF: 29.98 (29.99)        29.98 (30.00)
    Nov 16 18:42:38 localhost user_kw_timer.sh[3273]: /dev/mmcblk1p1
    Nov 16 18:42:38 localhost user_kw_timer.sh[3273]: Detected SD Card
    Nov 16 18:42:38 localhost user_kw_timer.sh[3273]: remain SD Storage 59106M Pass Delete Process
    Nov 16 18:42:38 localhost user_kw_timer.sh[3273]: filename : /media/000000007_2020_11_16/000000007_000107_2020_11_16_18_42_38.mp4
    Nov 16 18:42:41 localhost user_kw_timer.sh[3273]: **PERF: 29.98 (29.99)        29.98 (30.00)
    Nov 16 18:42:46 localhost user_kw_timer.sh[3273]: **PERF: 0.00 (29.99)        0.00 (30.00)

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

  1. pls share your setup with us
  2. For your issue, I don’t see any clue, could you narrow down your issue firstly?