Gstreamer pipeline freezes with nvmsgbroker

[PrabakaranR]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) - DGPU
• DeepStream Version - 6.2
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only) 535
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Note: Using Nvidia docker container

Hi,

I am trying to execute below gst pipeline which comprises of nvmsgbroker as well as nvmsgconv.
But the pipeline freezes (with blank screen in the display window).
In my initial level of analysis, I suspect nvmsgbroker might be the cause.

Also, note that Kafka broker is set up and running properly alongside and I am able to validate the connection thro standalone application “deepstream-occupancy-analytics”.

Kindly let me know what am I missing here.

gst-launch-1.0 filesrc location=/opt/nvidia/deepstream/deepstream/samples/streams/sample_1080p_h264.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! m.sink_0 nvstreammux name=m batch-size=1 width=1920 height=1080 ! nvinfer config-file-path=/opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/config_infer_primary.txt batch-size=1 unique-id=1 ! nvtracker ll-lib-file=/opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so ! nvinfer config-file-path=/opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/config_infer_secondary_carcolor.txt batch-size=16 unique-id=2 infer-on-gie-id=1 infer-on-class-ids=0 ! tee name=t1 t1. ! nvmsgconv config=/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-occupancy-analytics/config/dstest_occupancy_analytics.txt msg2p-lib=/opt/nvidia/deepstream/deepstream/lib/libnvds_msgconv.so ! nvmsgbroker conn-str=“localhost;9092;quickstart-events” proto-lib=/opt/nvidia/deepstream/deepstream/lib/libnvds_kafka_proto.so sync=FALSE async=FALSE topic=quickstart-events t1. ! nvmultistreamtiler rows=1 columns=1 width=1280 height=720 ! nvvideoconvert ! nvdsosd ! nveglglessink

Best Regards,
Prabakaran R

[yuweiw]

Could you try running two pipelines separately without using the tee plugin first? One is msgconv+msgbroker, the other is nveglglessink.

[PrabakaranR]

When I run the pipeline seperately,

nveglglessink (without msgconv+msgbroker), it works
msgconv+msgbroker (without nveglglessink), execution closes immediately

BTW, I tried and explored different options, and I found that when I introduce ‘queue’ plugin (as like below), then the pipeline is working. Hope the approach is correct

gst-launch-1.0 filesrc location=/opt/nvidia/deepstream/deepstream/samples/streams/sample_1080p_h264.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! m.sink_0 nvstreammux name=m batch-size=1 width=1920 height=1080 ! nvinfer config-file-path=/opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/config_infer_primary.txt batch-size=1 unique-id=1 ! nvtracker ll-lib-file=/opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so ! nvinfer config-file-path=/opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/config_infer_secondary_carcolor.txt batch-size=16 unique-id=2 infer-on-gie-id=1 infer-on-class-ids=0 ! tee name=t1 t1. ! queue ! nvmsgconv config=/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-occupancy-analytics/config/dstest_occupancy_analytics.txt msg2p-lib=/opt/nvidia/deepstream/deepstream/lib/libnvds_msgconv.so ! nvmsgbroker conn-str=“localhost;9092;quickstart-events” proto-lib=/opt/nvidia/deepstream/deepstream/lib/libnvds_kafka_proto.so sync=FALSE async=FALSE topic=quickstart-events t1. ! nvmultistreamtiler rows=1 columns=1 width=1280 height=720 ! queue ! nvvideoconvert ! queue ! nvdsosd ! queue ! nveglglessink

Thanks

[yuweiw]

The approach is OK. You can refer to our demo code: opt\nvidia\deepstream\deepstream\sources\apps\sample_apps\deepstream-test4\deepstream_test4_app.c. The tee plugin is followed by two queue plugins.

[PrabakaranR]

Thanks yuweiw
Appreciate your support