About batched-push-timeout for nvstreammux in DeepStream 6.1

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

• Hardware Platform (Jetson / GPU) Jetson AGX Orin
• DeepStream Version 6.1
• JetPack Version (valid for Jetson only) 5.0.1
• TensorRT Version 8.4.0
• Issue Type( questions, new requirements, bugs) questions

Hi, I set 4 to batch-size and 40000 to batched-push-timeout for nvstreammux in DeepStream 6.0.1.

[streammux]
gpu-id=0
live-source=0
batch-size=4
batched-push-timeout=40000
width=1920
height=1080
enable-padding=0
nvbuf-memory-type=0

On this configuration, deepstream-application ran successfully when I set 1 input stream(mp4, 25fps).

But, in DeepStream 6.1, the application occured an error that not execute inference and not exit the application on same configuration.

** INFO: <bus_callback:183>: Pipeline running

**PERF:  FPS 0 (Avg)
Thu Sep 29 10:47:36 2022
**PERF:  0.00 (0.00)
Thu Sep 29 10:47:41 2022
**PERF:  inf (0.60)
Thu Sep 29 10:47:46 2022
**PERF:  0.00 (0.30)
Thu Sep 29 10:47:51 2022
**PERF:  0.00 (0.20)
...

In addition, I set USE_NEW_NVSTREAMMUX=yes and ran the application on same configuration in DeepStream 6.1, then the application ran successfully.

Did the specifications of batched-push-timeout for nvstreammux changed by updating DeepStream 6.1?

3

No.

Are you working with deepstream-app? Can you post the complete configuration?

4

Thank you for your response.

I post the configuration that modified samples/configs/deepstream-app/source30_1080p_dec_infer-resnet_tiled_display_int8.txt inside nvcr.io/nvidia/deepstream-l4t:6.1-samples.
I don’t modified other configurations.

I run the following command inside nvcr.io/nvidia/deepstream-l4t:6.1-samples, then deepstream-test5-app occur the error.

$ cd /opt/nvidia/deepstream/deepstream-6.1
$ deepstream-test5-app -c samples/configs/deepstream-app/deepstream-app-config.txt

deepstream-app-config.txt (5.1 KB)

5

The configuration file works well on my Orin board with DeepStream 6.1.

6

The configuration file does not work well on my board with Jetpack 5.0.1 DP and DeepStream 6.1.
But, it works well on my board with Jetpack 5.0.2 and DeepStream 6.1.1.

Do you mean that it works well on your board with Jetpack 5.0.1 DP and DeepStream 6.1?

7

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

Yes.