swartz-k

I properly misunderstand. The right way may be `kubelet` - `containerd` - `firecracker-containerd`. Trying on this.

> Thanks for opening it. @swartz-k , if you'd like open an PR on fixing it, we will be very appreciated. Sure, I will do that. And also should we...

可以尝试检查文件 `/etc/nvidia-container-runtime/config.toml`中的 `ldconfig`值，默认是 `"@/sbin/ldconfig"` 改为 `"/sbin/ldconfig"` 试试？

It looks like scheduler `gpu-manager` failed, In my situation restart will be helpful.

gpu-count: number of GPU devices gpu-mem: GPU memory If you wanna use 50% of 3090Ti 24GB，you can set ``` limits: aliyun.com/gpu-mem: 12 ```

In my case, the error message is as below. ``` 2023-01-11T06:56:35.919540Z ERROR vector::topology: Configuration error. error=Source "task_log": Could not create subdirectory "task_log" inside of data dir "/var/lib/vector/": Read-only file system...

```Error 401: Anonymous caller does not have storage.objects.list access to the Google Cloud Storage bucket., required","stacktrace":"github.com/kubeflow/kfserving/pkg/agent.(*Puller).modelProcessor\n\t/go/src/github.com/kubeflow/kfserving/pkg/agent/puller.go:149"}``` looks like Anonymous cannot access`gs://kfserving-custom-models/model_repo`