Failed to run sample | Failed to create the Vulkan instance | Clara Holoscan SDK 0.4

[abdelkrim1]

Hi, I am trying to run : python3 ./apps/endoscopy_tool_tracking/python/endoscopy_tool_tracking.py --source=replayer using docker in x86 Platform, I got this error:
Vulkan Version:

• available: 1.2.131
• requesting: 1.2.0
  VK_ERROR_EXTENSION_NOT_PRESENT: VK_KHR_external_memory_capabilities - 0
  2023-01-20 10:56:08.963 ERROR /workspace/holoscan-sdk/gxf_extensions/holoviz/holoviz.cpp@513: Failed to create the Vulkan instance.
  2023-01-20 10:56:08.963 WARN gxf/std/greedy_scheduler.cpp@235: Error while executing entity 42 named ‘holoviz’: GXF_FAILURE
  2023-01-20 10:56:08.966 ERROR gxf/std/entity_executor.cpp@540: Entity [holoviz] must be in Lifecycle::kStarted or Lifecycle::kIdle stage before stopping. Current state is StartPending
  2023-01-20 10:56:08.966 INFO gxf/std/greedy_scheduler.cpp@367: Scheduler finished.
  2023-01-20 10:56:08.966 ERROR gxf/std/program.cpp@497: wait failed. Deactivating…
  2023-01-20 10:56:08.967 DEBUG gxf/cuda/cuda_stream.cpp@62: CudaStream destroyed
  2023-01-20 10:56:08.967 ERROR gxf/core/runtime.cpp@1251: Graph wait failed with error: GXF_FAILURE
  2023-01-20 10:56:08.967 PANIC /workspace/holoscan-sdk/src/core/executors/gxf/gxf_executor.cpp@278: GXF operation failed: GXF_FAILURE

any ideas?
Thanks

[aquraini]

Hello Abdelkrim1,

Thank you for your interest in Clara Holoscan and apologies for the troubles. This is a known issue and there’s a workaround that’s documented in the “Running the container” section of the NGC catalog page for the Holoscan container:

Holoscan Container | NVIDIA NGC

specifically:

# Find the nvidia_icd.json file which could reside at different paths
# Needed due to https://github.com/NVIDIA/nvidia-container-toolkit/issues/16
nvidia_icd_json=$(find /usr/share /etc -path '*/vulkan/icd.d/nvidia_icd.json' -type f 2>/dev/null | grep .) || (echo "nvidia_icd.json not found" >&2 && false)

Could you try running this command before launching docker? Hope this helps.

Thanks!

[abdelkrim1]

Thanks for the quick reply!