Centos 9 kubeadm k8s cluster pod to pod communication is not working

Question

Description: I am setting up Kubernetes 1.29 on CentOS 9 using kubeadm and Calico v3.26.1 as the CNI. I followed the installation steps from this article:

Install a Kubernetes Cluster on RHEL 9 | CentOS 9

I applied the following Calico manifests:

tigera-operator.yaml

custom-resources.yaml

Problem: 🚨 Issues observed:

1. Pods cannot access the internet (e.g., curl google.com fails).
2. Pod-to-pod communication is broken (pods on the same and different nodes cannot reach each other).

What I Have Tried:

1. Verified Calico pods are running:

   kubectl get pods -n calico-system

2. Checked if IP forwarding is enabled:

   sysctl net.ipv4.ip_forward ✅ It is set to 1.

3. Examined calico-node logs:

   kubectl logs -n calico-system -l k8s-app=calico-node

Environment Details:

1. Kubernetes Version: 1.29
2. Calico Version: v3.26.1
3. OS: CentOS Stream 9
4. Installation Method: kubeadm

Expected Behavior:

1. Pods should be able to access external networks.
2. Pod-to-pod communication should work seamlessly.

Actual Behavior:

1. No internet access from inside the pod.
2. Pods cannot reach each other (both within and across nodes).

Questions:

1. Does Calico v3.26.1 fully support Kubernetes 1.29 on CentOS 9?
2. Are there additional configurations required for networking on CentOS 9?
3. How can I further debug and fix this issue?