"Target is not supported" using GUI to connect to Linux desktop

I am having issues connecting Nsight Systems to my Linux desktop.

I am running the docker container for nsys using run_nsight_streamer.py, downloaded today. I connect via the web gui.

This issue reproduces whether I visit the webpage from a different machine, or from the target itself. My SSH connection completes but I receive the error “Target is not supported.” as seen in the screenshots below.

Any advice for further troubleshooting would be helpful!

Target machine stats:

  • NVIDIA RTX 5090

  • nsys --version gives NVIDIA Nsight Systems version 2025.5.1.121-255136380782v0 (which is the version requested to install in the error message)

  • docker -- version gives Docker version 28.2.2, build e6534b4

  • uname -a gives Linux <hostname redacted> 6.8.0-85-generic #85-Ubuntu SMP PREEMPT_DYNAMIC Thu Sep 18 15:26:59 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux

  • nvidia-smi gives | NVIDIA-SMI 580.95.05 Driver Version: 580.95.05 CUDA Version: 13.0 |

  • I have CUDA Toolkit 12.9 installed.

  • docker logs for my Nsight container does not seem to have relevant info, but including it here just in case:

    docker-log.txt (143.2 KB)

@mpopov can you look into this?

Nsight Streamer currently does not support profiling (it is a view-only tool) due to security considerations. Please use Nsight Systems CLI or Nsight Operator instead.

Thanks for the quick response.

Just making sure - the instructions described in this tutorial, which involve using jupyterlab-nvidia-nsight are not supported? (the tutorial was created this year by an NVIDIA employee)