bug(nvidia-container-toolkit): unsupported IMEX channel

[ugurgural]

Hi, since with the version v20241109, I started to get the error I shared below. Initially thought that my nvidia-device-plugin could be an older version, but even updating to the latest one(0.17.0) didn't solve it. As a workaround, I reverted the karpenter configuration pointing out to the previous version v20241016. Does anyone have an idea what this could be?

Note: configuration is almost barebone with the original eks GPU AMI, I only enable docker service in the user data for DIND scenarios.

Error: failed to create containerd task: 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: error parsing IMEX info: unsupported IMEX channel value: all: unknown

Environment:

• AWS Region: eu-central-1
• Instance Type(s): g4dn.8xlarge
• EKS Platform version (use aws eks describe-cluster --name <name> --query cluster.platformVersion): eks.8
• Kubernetes version (use aws eks describe-cluster --name <name> --query cluster.version): 1.30
• AMI Version: v20241109
• Kernel (e.g. uname -a): Linux ip-xxx-xx-xx-xx.eu-central-1.compute.internal 5.10.227-219.884.amzn2.x86_64 1 SMP Tue Oct 22 16:38:23 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
• Release information (run cat /etc/eks/release on a node):

BASE_AMI_ID="ami-07d4ef2c1e42a1ce8"
BUILD_TIME="Sat Nov  9 22:55:23 UTC 2024"
BUILD_KERNEL="5.10.227-219.884.amzn2.x86_64"
ARCH="x86_64"

[Issacwww]

I think this issue is related to NVIDIA/nvidia-container-toolkit#797, we will have another release (the one after v20241115) soon to have the updated nvidia-contianer-toolkit 1.17.2