Running Intro to Clara Train SDK fails in WSL 2

Hi,

I get the following error when running the setup script from here: https://ngc.nvidia.com/catalog/resources/nvidia:med:clara:getting_started/setup. I am trying to run in WSL 2.

docker: Error response from daemon: OCI runtime create failed: container_linux.go:349: starting container process caused “process_linux.go:449: container init caused “process_linux.go:432: running prestart hook 1 caused \“error running hook: exit status 1, stdout: , stderr: nvidia-container-cli: mount error: file creation failed: /var/lib/docker/overlay2/86c5edf835c1e591da3cf94566d2461420c0e55e3a38bfd213e250e6267e9d48/merged/usr/lib/x86_64-linux-gnu/libcuda.so.1: file exists\\n\”””: unknown.

This command works: $ docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark

Output:

Windowed mode
Simulation data stored in video memory
Single precision floating point simulation
1 Devices used for simulation
MapSMtoCores for SM 7.5 is undefined. Default to use 64 Cores/SM
GPU Device 0: “GeForce RTX 2070 with Max-Q Design” with compute capability 7.5

Compute 7.5 CUDA device: [GeForce RTX 2070 with Max-Q Design]
36864 bodies, total time for 10 iterations: 113.239 ms
= 120.008 billion interactions per second
= 2400.155 single-precision GFLOP/s at 20 flops per interaction

Did anyone succeed with this ?

Despite updating to the latest driver, version of windows (20211) I am still unable to run the the intro to clara script on WSL 2. Can NVIDIA shed any light on this ?!

Commands such as: sudo docker run --runtime=nvidia --rm -itd --ipc=host --name w2l wav2letter/wav2letter:cuda-latest
Unable to find image ‘wav2letter/wav2letter:cuda-latest’ locally

Also fail on WSL 2, but

docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark

works.

1 Like

I have the exact problem.
NVIDIA docs examples work perfectly

but my application gives: Cannot start service dynamic: OCI runtime create failed: container_linux.go:349: starting container process caused "process_linux.go:449: container init caused \"process_linux.go:432: running prestart hook 1 caused \\\"error running hook: exit status 1, stdout: , stderr: nvidia-container-cli: mount error: file creation failed: /var/lib/docker/overlay2/...blabla../merged/usr/lib/x86_64-linux-gnu/libcuda.so.1: file exists\\\\n\\\"\"": unknown

Hi, FYI one could modify image removiing or renaming duplicate lib as suggested in following to prevent error, but that doesn’t seem sufficient to make it actually work
Issues Running DeepStream on WSL2 (Docker Container): usr/lib/x86_64-linux-gnu/libcuda.so.1: file exists\\\\n\\\"\"": unknown