Unable to create pgie, nvosd

[samjith888]

Please provide complete information as applicable to your setup.

**• Hardware Platform : Jetson Tx2
**• DeepStream Version : 6.0.1 (deepstream-l4t:6.0.1-samples)
**• JetPack Version : 4.6.1

Getting the following error when try to run deepstream python app

Creating Pipeline 
 
 Unable to create pgie 
 Unable to create nvosd 
Creating EGLSink 

Traceback (most recent call last):
  File "deepstream_test_4.py", line 596, in ds_main
    main(sys.argv)
  File "deepstream_test_4.py", line 478, in main
    pgie.set_property('config-file-path', PGIE_CONFIG_FILE)
AttributeError: 'NoneType' object has no attribute 'set_property'

tried after removing gstreamer cache by using the below command, but still getting the same error
rm ${HOME}/.cache/gstreamer-1.0/*

ll /usr/lib/aarch64-linux-gnu/gstreamer-1.0/deepstream/

 libcustom2d_preprocess.so*
 libgstnvvideoconvert.so*
 libnvdsgst_audiotemplate.so*
 libnvdsgst_deepstream_bins.so*
 libnvdsgst_dewarper.so*
 libnvdsgst_dsanalytics.so*
 libnvdsgst_dsexample.so*
 libnvdsgst_infer.so*
 libnvdsgst_inferaudio.so*
 libnvdsgst_inferserver.so*
 libnvdsgst_logger.so*
 libnvdsgst_msgbroker.so*
 libnvdsgst_msgconv.so*
 libnvdsgst_multistream.so*
 libnvdsgst_multistreamtiler.so*
 libnvdsgst_of.so*
 libnvdsgst_ofvisual.so*
 libnvdsgst_osd.so*
 libnvdsgst_preprocess.so*
 libnvdsgst_segvisual.so*
 libnvdsgst_tracker.so*
 libnvdsgst_udp.so*
 libnvdsgst_videotemplate.so*

[yingliu]

Did you follow the installation guide to install python binding? (deepstream_python_apps/bindings at master · NVIDIA-AI-IOT/deepstream_python_apps · GitHub tag v1.1.1)
Did you success to run other python examples?

[samjith888]

Yes, I have installed python bindings as mentioned in the repo (v1.1.1).
But getting the same error

The same steps i have tried in a jetson nano (Jetpack 4.5, Deepstream 6.0.1), and I’m able to run deepstream python apps there

[Amycao]

Can you run success directly on the device?

[samjith888]

Sorry,

What do you meant by ‘ Can you run success directly on the device?’

[Amycao]

I see you run the python app within docker failed. so suggest you if or not can run directly on device?

[samjith888]

I’m wondering, how the same docker file and scripts is working on a jetson device!.

I forgot to mention the below error which I’m getting while building the container in the Jetson TX2 NX, but this error is not in the jetson nano

ERROR: for DS  Cannot start service DS_container: OCI runtime create failed: container_linux.go:380: starting container process caused: process_linux.go:545: container init caused: Running hook #0:: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as 'csv'
invoking the NVIDIA Container Runtime Hook directly (e.g. specifying the docker --gpus flag) is not supported. Please use the NVIDIA Container Runtime (e.g. specify the --runtime=nvidia flag) instead.: unknown

ERROR: for DS  Cannot start service DS_container: OCI runtime create failed: container_linux.go:380: starting container process caused: process_linux.go:545: container init caused: Running hook #0:: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as 'csv'
invoking the NVIDIA Container Runtime Hook directly (e.g. specifying the docker --gpus flag) is not supported. Please use the NVIDIA Container Runtime (e.g. specify the --runtime=nvidia flag) instead.: unknown
ERROR: Encountered errors while bringing up the project.

This link stating that the above error is happening since nvidia-container-toolkit is missing.

Then i tried to install nvidia container toolkit on the device by using below command , but its showing that it is already installed

Command
sudo apt-get install -y nvidia-container-toolkit

Output

nvidia-container-toolkit is already the newest version (1.11.0-1).
nvidia-container-toolkit set to manually installed.

[mchi]

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

Hi @samjith888 ,
The docker toolkit should be installed through SDKManager