Running standalone examples

Omniverse Isaac Sim
1

Hi!
I’ve a container running on a remote AWS server, where for doing this I followed the steps here:
https://docs.omniverse.nvidia.com/isaacsim/latest/installation/install_container.html#isaac-sim-app-install-container
(Ubuntu 20.04, IsaacSim 4.2.0)

‘./runheadless.native.sh’ works fine, but when I try to run a standalone example as in Isaac Sim Workflows — Omniverse IsaacSim, I get the following error lines in my output:
2024-11-19 07:52:18 [21,773ms] [Error] [carb.graphics-vulkan.plugin] VkResult: ERROR_INITIALIZATION_FAILED
2024-11-19 07:52:18 [21,773ms] [Error] [carb.graphics-vulkan.plugin] vkEnumeratePhysicalDevices failed. No physical device is found.
2024-11-19 07:52:18 [21,773ms] [Error] [carb.graphics-vulkan.plugin] No physical device is found.
2024-11-19 07:52:18 [21,947ms] [Error] [gpu.foundation.plugin] carb::graphics::createInstance failed.
2024-11-19 07:52:21 [24,988ms] [Error] [carb.graphics-vulkan.plugin] VkResult: ERROR_INITIALIZATION_FAILED
2024-11-19 07:52:21 [24,988ms] [Error] [carb.graphics-vulkan.plugin] vkEnumeratePhysicalDevices failed. No physical device is found.
2024-11-19 07:52:21 [24,989ms] [Error] [carb.graphics-vulkan.plugin] No physical device is found.
2024-11-19 07:52:21 [25,187ms] [Error] [gpu.foundation.plugin] carb::graphics::createInstance failed.
2024-11-19 07:52:21 [25,187ms] [Error] [omni.gpu_foundation_factory.plugin] Failed to create any GPU devices, including an attempt with compatibility mode.

2024-11-19 07:52:22 [25,801ms] [Error] [omni.kit.renderer.plugin] GPU Foundation is not initialized!

2024-11-19 07:53:16 [25,709ms] [Error] [omni.kit.imgui_renderer.plugin] _prepareDefaultWindow: invalid device attached to the default window!

2024-11-19 07:54:24 [36,546ms] [Error] [omni.physx.plugin] CUDA libs are present, but no suitable CUDA GPU was found!
2024-11-19 07:54:24 [36,630ms] [Warning] [omni.graph.core.plugin] OmniGraphSettings::getCudaDeviceOrdinal: unable to get a valid CUDA device id from the renderer. Defaulting to GPU0.
2024-11-19 07:54:24 [36,648ms] [Warning] [omni.graph.core.plugin] OmniGraphSettings::getCudaDeviceOrdinal: unable to get a valid CUDA device id from the renderer. Defaulting to GPU0.
[36.764s] app ready
2024-11-19 07:54:25 [37,832ms] [Warning] [omni.kit.imgui_renderer.plugin] _createExtendCursor: No app window.
2024-11-19 07:54:25 [37,832ms] [Warning] [omni.kit.imgui_renderer.plugin] _createExtendCursor: No app window.
2024-11-19 07:54:25 [37,838ms] [Warning] [omni.graph.core.plugin] OmniGraphSettings::getCudaDeviceOrdinal: unable to get a valid CUDA device id from the renderer. Defaulting to GPU0.

The command was:
./python.sh standalone_examples/api/omni.isaac.franka/follow_target_with_rmpflow.py

2

What is the GPU driver version? Is it one of the production versions, either 535 or 550?
Does the compatibility checker look good?

3

Thank you for the fast reply.
It’s 535, which is specified in the first step in the guide.

4

Please provide the complete log.

6

Thank you! I looked at the script ‘omni.isaac.franka/follow_target_with_rmpflow.py’ itself and it wasn’t in a headless mode. When I replace the instantiation of SImulationApp with a relevant code from the script omni.isaac.kit/livestreaming.py it worked fine. I think you need to clarify that these examples are not in headless mode (which is motioned indirectly in ’ Using Jupyter to Develop a Standalone Application’ section)

7

I’m glad to hear you resolved the problem.

Yes, some standalone examples do not run in headless mode by default.

8

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