Cannot start pipeline running python test app 1

[joyce]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) NVIDIA Orin NX Developer Kit
• DeepStream Version 6.3.0
• JetPack Version (valid for Jetson only) R35 (release), REVISION: 4.1, GCID: 33958178, BOARD: t186ref, EABI: aarch64, DATE: Tue Aug 1 19:57:35 UTC 2023
• TensorRT Version 8.5.2-1+cuda11.4
• NVIDIA GPU Driver Version (valid for GPU only)

Note: I can currently on DeepStream_Python_Apps_Bindings_v1.1.8

I am trying to run Deepstream-test1 with python ./deepstream_test_1.py …/…/…/…/samples/streams/sample_720p.h264. I get the error:

...
Starting pipeline 

Error: gst-core-error-quark: GStreamer error: state change failed and some element failed to post a proper error message with the reason for the failure. (4): gstbasesink.c(5367): gst_base_sink_change_state (): /GstPipeline:pipeline0/GstNv3dSink:nv3d-sink:
Failed to start

if I change the sink to nveglstreamsrc, the it get

Starting pipeline 

Opening in BLOCKING MODE 
0:00:00.205987578 63555     0x119dfac0 WARN                 nvinfer gstnvinfer.cpp:679:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::initialize() <nvdsinfer_context_impl.cpp:1174> [UID = 1]: Warning, OpenCV has been deprecated. Using NMS for clustering instead of cv::groupRectangles with topK = 20 and NMS Threshold = 0.5
0:00:02.958693413 63555     0x119dfac0 INFO                 nvinfer gstnvinfer.cpp:682:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1988> [UID = 1]: deserialized trt engine from :/opt/nvidia/deepstream/deepstream-6.3/samples/models/Primary_Detector/resnet10.caffemodel_b1_gpu0_int8.engine
INFO: [Implicit Engine Info]: layers num: 3
0   INPUT  kFLOAT input_1         3x544x960       
1   OUTPUT kFLOAT conv2d_bbox     16x34x60        
2   OUTPUT kFLOAT conv2d_cov/Sigmoid 4x34x60         

0:00:03.128983386 63555     0x119dfac0 INFO                 nvinfer gstnvinfer.cpp:682:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:2091> [UID = 1]: Use deserialized engine model: /opt/nvidia/deepstream/deepstream-6.3/samples/models/Primary_Detector/resnet10.caffemodel_b1_gpu0_int8.engine
0:00:03.252724591 63555     0x119dfac0 INFO                 nvinfer gstnvinfer_impl.cpp:328:notifyLoadModelStatus:<primary-inference> [UID 1]: Load new model:dstest1_pgie_config.txt sucessfully
NvEglStreamSrc: Can not start because egl_display and stream are not set
NvMMLiteOpen : Block : BlockType = 261 
NvMMLiteBlockCreate : Block : BlockType = 261 
Error: gst-core-error-quark: GStreamer error: state change failed and some element failed to post a proper error message with the reason for the failure. (4): gstbasesrc.c(3468): gst_base_src_start (): /GstPipeline:pipeline0/GstNvEglStreamSrc:nv3d-sink:
Failed to start

[yingliu]

Can you check if the method in this topic works for your case? Faced an error running “deepstream_python_apps/apps/runtime_source_add_delete” - Intelligent Video Analytics / DeepStream SDK - NVIDIA Developer Forums

[joyce]

I’ve tried that solution, it doesn’t work. thanks

[junshengy]

How do you run the program? Is it in ssh?

What is the result of the following command?
The value that you usually set DISPLAY to in ssh

cat /proc/$(pidof "gnome-terminal-server")/environ | tr '\0' '\n' | grep ^DISPLAY=

[joyce]

Yes it is in ssh. I run it as such sudo python3 ./deepstream_test_1.py ../../../../samples/streams/sample_720p.h264

When I run cat /proc/$(pidof "gnome-terminal-server")/environ | tr '\0' '\n' | grep ^DISPLAY= it says DISPLAY=:1
I’ve tried setting DISPLAY=:0 but it gives the same error.

When I run it on the Jetson, it does it no problem.

[junshengy]

I think you should try setting export DISPLAY=:1.

DeepStream does not support x11-forwarding, so when you use ssh, you need to execute export DISPLAY=:x

[joyce]

I’m getting the same error for DISPLAY=:1.

[junshengy]

Do not use sudo, otherwise environment variables will be lost.

If the /opt directory is not writable, use sudo chmod -R 777

[joyce]

thanks, the pipeline runs successfully but the actual display of the cars and people do not show.

[junshengy]

@joyce

but the actual display of the cars and people do not show.

This is normal since x11-forwarding is not supported.
The program will not be displayed on ssh, it can only be displayed on the host side