Lab 4.1 Upgrade Cluster - Error container.sock - no such file

mlh

hello,
i am working on the cluster upgrade in lab 4.1. my kubeadm version was 1.26.1-00. i installed 1.27.1-00 via apt-get install. when i try to apply the upgrade with sudo kubeadm upgrade apply v1.27.1 i get many of these errors:

[ERROR ImagePull]: failed to pull image registry.k8s.io/kube-apiserver:v1.27.1: output: E0910 07:06:33.434750 6948 remote_image.go:171] "PullImage from image service failed" err="rpc error: code = Unavailable desc = connection error: desc = \"transport: Error while dialing dial unix /run/containerd/container.sock: connect: no such file or directory\"" image="registry.k8s.io/kube-apiserver:v1.27.1"

i have the file /run/containerd/containerd.sock but no /run/containerd/container.sock.

why would the upgrade be looking for container.sock?

any help is appreciated.

thanks,
mel

mlh

i wanted to add the following:

• process data:
  root 929 1 2 04:54 ? 00:00:02 /usr/bin/containerd

root 2121 1 0 04:54 ? 00:00:00 /usr/bin/containerd-shim-runc-v2 -namespace k8s.io -id c5c88cabfdf7c032879428bc90a308b0338bc524cfe5c39a850d9f2ade7f3241 -address /run/containerd/containerd.sock

• /etc/containerd/config.toml file data:
  [grpc]
  address = "/run/containerd/containerd.sock"

i have a working cluster with an ngninx pod running. i have been able to successfully complete all of the labs up to this point without error or issue.

thanks,
mel

chrispokorni

Hi @mlh,

What are the contents of /etc/crictl.yaml?

Regards,
-Chris

mlh

thanks Chris. that was the issue. image endpoint was container.sock
image-endpoint: "unix:///run/containerd/container.sock"
regards,
mel