Setting up cluster on VMs

farazoman

Hi there, I've been following the guide installing all the needed dependencies and following the required steps. But it seems the pods can never get into a ready state and stay that way.

When getting the pods this is what is shown:

faraz@k8s-cp-ubu:~$ kubectl get pods --all-namespaces
NAMESPACE     NAME                                       READY   STATUS             RESTARTS          AGE
kube-system   calico-kube-controllers-55fc758c88-fws4q   0/1     CrashLoopBackOff   6 (3m15s ago)     11m
kube-system   calico-node-p9pxg                          0/1     CrashLoopBackOff   6 (2m37s ago)     11m
kube-system   coredns-57575c5f89-9vbxf                   0/1     CrashLoopBackOff   5 (26s ago)       11m
kube-system   coredns-57575c5f89-xgt4r                   1/1     Running            6 (3m40s ago)     11m
kube-system   etcd-k8s-cp-ubu                            1/1     Running            146 (9m36s ago)   13m
kube-system   kube-apiserver-k8s-cp-ubu                  1/1     Running            159 (53s ago)     13m
kube-system   kube-controller-manager-k8s-cp-ubu         0/1     CrashLoopBackOff   209 (2m13s ago)   13m
kube-system   kube-proxy-zn5vd                           0/1     CrashLoopBackOff   5 (2m32s ago)     11m
kube-system   kube-scheduler-k8s-cp-ubu                  0/1     CrashLoopBackOff   212 (101s ago)    13m

It seems all the pods are restarting to some extent. Even though the control plane is in ready state.

faraz@k8s-cp-ubu:~$ kubectl get nodes -o wide
NAME         STATUS   ROLES           AGE   VERSION    INTERNAL-IP     EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION      CONTAINER-RUNTIME
k8s-cp-ubu   Ready    control-plane   70m   v1.24.10   192.168.100.6   <none>        Ubuntu 22.04.1 LTS   5.15.0-60-generic   containerd://1.6.16

In general, looking at the logs for each pod either have no error logs or in one case (I can't reproduce it) mentioned a timeout for UDP on 192.168.100.6, on the core-dns pod.

Would anyone have any additional troubleshooting tips or suggestions for how to pass this state. I've been stuck for a while and haven't been able to get to an actual running control plane.

chrispokorni

Hi @farazoman,

First, I'd recommend ensuring that the local VMs are assigned IP addresses that are not overlapping the 192.168.0.0/16 network, which will eventually be used for the application pods network layer implemented by the Calico CNI plugin (please revisit the Summary section of the Lab exercise 2.1: Overview and Preliminaries).

Second, please use the guest OS noted in the lab guide - Ubuntu 20.04 LTS (in the same exercise Overview and Preliminaries)

And third, please install the Kubernetes version recommended by the lab guide - v1.26.1 (step 4 of Lab exercise 2.2, and as provided in the k8scp.sh installation and config script).

Regards,
-Chris

farazoman

thanks for the advice, I'll review the docs but I just noticed I messed up in posting in the wrong class, I"m actually taking LFS258.

Few things to take away though,
1. use ubuntu 20.04
2. update ips (assuming the same problem of ips applies to LFS258
3. use 1.25 (from 258 slides), although I don't imagine this being critical to the rest of the course

Addtionally, just to avoid double posting, I won't put it on the other class forum unless you suggest otherwise. Thanks!

fcioanca

This thread was moved to the correct forum, LFS258.