Could not get cuda device count (cudaErrorUnknown) in deepstream

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• 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)

cuda isssue
cuda isssue2040×1210 174 KB

Hii …

i am facing some difficulty on when i starting my pipeline :

INFO:pipeline:Creating Stream Muxer
INFO:pipeline:Creating Inference
Error: Could not get cuda device count (cudaErrorUnknown)
Failed to parse group property
** ERROR: <gst_nvinfer_parse_config_file:1392>: failed
INFO:pipeline:Creating Inference-2
Error: Could not get cuda device count (cudaErrorUnknown)
Failed to parse group property
** ERROR: <gst_nvinfer_parse_config_file:1392>: failed
INFO:pipeline:Creating Fake Sink
INFO:pipeline:Creating Video Converter
INFO:pipeline:Creating Caps filter 1
INFO:pipeline:Starting DeepStream pipeline…
INFO: Started server process [430]
Unable to set device in gst_nvstreammux_change_state
Unable to set device in gst_nvstreammux_change_state
INFO: Waiting for application startup.
0:00:00.042169211 430 0x747938004bc0 WARN bin gstbin.c:2808:reset_state: Failed to switch back down to NULL
INFO: Application startup complete.
INFO:pipeline:Pipeline set to READY state.

after i execute these two lines like nvcc --version and nvidia-smi in container and as well a inhost then after i started my pipeline script and its running fine…

do you have any idea regarding this error

i am using below version (drivers , cudatoolkit)

cuda_device 2
cuda_device 22442×1534 309 KB

3

Could you attach the command you used to launch your docker?

4

sudo docker run -it --privileged --name=deepstream_yolov9 --net=host -e CUDA_CACHE_DISABLE=0 -v pwd/Deepstream:/opt/nvidia/deepstream/deepstream-7.1/deepstream_yolov9 nvcr.io/nvidia/deepstream:7.1-triton-multiarch

5

Do you mean that after you run these two commands, there will be no errors?
Also you can try to set the --privileged=true in your command.

6

there is no such type of error after executing this in both host cmd prompt as well as in container cmd…

7

This is weird. We haven’t had this similar problem on our side before. So if you run a new docker, will this issue be occurred again?

8

yes still same issue … not resolved

if i use --gpus all
i am encountering this below error :

docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as ‘legacy’
nvidia-container-cli: initialization error: load library failed: libnvidia-ml.so.1: cannot open shared object file: no such file or directory: unknown.

9

So will the problem still be solved after you run the following command nvcc --version and nvidia-smi this time?

Have you followed our step-by-step Guide docker-containers to configuring your environment?

10

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.