Solution Networking Weave optional

In this lesson, we explore key aspects of the Kubernetes cluster networking configuration. We will answer several questions regarding node count, networking solutions, Weave deployment details, and pod network settings.

1. How Many Nodes are in the Cluster?

To determine the total number of nodes in the Kubernetes cluster, use the following command:

kubectl get node

The output shows two nodes—one named "controlplane" and one named "node01":

controlplane ➜  kubectl get node
NAME         STATUS   ROLES         AGE   VERSION
controlplane Ready    control-plane 37m   v1.27.0
node01      Ready    <none>        37m   v1.27.0

2. What is the Networking Solution Used by the Cluster?

The networking solution is identified by inspecting the CNI (Container Network Interface) configuration files. First, navigate to the directory containing these configuration files:

controlplane ➜  cd /etc/cni/net.d
controlplane /etc/cni/net.d ➜  ls
10-weave.conflist

Now, view the contents of the Weave configuration file:

controlplane /etc/cni/net.d ➜  cat 10-weave.conflist
{
  "cniVersion": "0.3.0",
  "name": "weave",
  "plugins": [
    {
      "name": "weave",
      "type": "weave-net",
      "hairpinMode": true
    },
    {
      "type": "portmap",
      "capabilities": {"portMappings": true},
      "snat": true
    }
  ]
}

controlplane /etc/cni/net.d ➜

From the configuration, it is clear that the cluster uses Weave as its networking solution.

3. How Many Weave Agents are Deployed in the Cluster?

To find out how many Weave agents (pods) are running in the cluster, list all pods in the kube-system namespace:

kubectl get pods -n kube-system

The output includes two Weave pods:

NAME                                      READY   STATUS    RESTARTS   AGE
coredns-5d78c9869d-597bx                  1/1     Running   0          38m
coredns-5d78c9869d-cqgw8                  1/1     Running   0          38m
etcd-controlplane                         1/1     Running   0          38m
kube-apiserver-controlplane               1/1     Running   0          38m
kube-controller-manager-controlplane      1/1     Running   0          38m
kube-proxy-45bqq                          1/1     Running   0          38m
kube-proxy-wjvbv                           1/1     Running   0          38m
kube-scheduler-controlplane               1/1     Running   0          38m
weave-net-mknr5                           2/2     Running   0          38m
weave-net-qgk9                            2/2     Running   0          38m

There are two Weave pods deployed, indicating one Weave agent per node.

4. On Which Nodes are the Weave Pods Running?

To check on which nodes the Weave agents are running, include the -o wide option when listing pods:

kubectl get pods -n kube-system -o wide

The output shows:

NAME                                      READY   STATUS    RESTARTS   AGE     IP            NODE          NOMINATED NODE   READINESS GATES
coredns-5d78c9869d-597bx                  1/1     Running   0          38m     10.244.0.2    controlplane   <none>           <none>
coredns-5d78c9869d-cqgw8                  1/1     Running   0          38m     10.244.0.3    controlplane   <none>           <none>
etcd-controlplane                         1/1     Running   0          38m     192.6.255.3   controlplane   <none>           <none>
kube-apiserver-controlplane               1/1     Running   0          38m     192.6.255.3   controlplane   <none>           <none>
kube-controller-manager-controlplane      1/1     Running   0          38m     192.6.255.3   controlplane   <none>           <none>
kube-proxy-45bqq                          1/1     Running   0          38m     192.6.255.3   node01         <none>           <none>
kube-proxy-wjvbv                           1/1     Running   0          38m     192.6.255.3   node01         <none>           <none>
kube-scheduler-controlplane               1/1     Running   0          38m     192.6.255.3   controlplane   <none>           <none>
weave-net-mknr5                           2/2     Running   0          38m     192.6.255.3   controlplane   <none>           <none>
weave-net-qgk9                            2/2     Running   0          38m     192.6.255.5   node01         <none>           <none>

Here, "weave-net-mknr5" is running on the "controlplane" node, and "weave-net-qgk9" is running on "node01". This confirms that each node hosts one Weave agent.

5. What is the Name of the Bridge Network Interface Created by Weave on Each Node?

To verify the network interfaces on a node, run:

ip add

A sample output from the "controlplane" node appears as follows:

controlplane /etc/cni/net.d ➜ ip add
 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
 2: datapath <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1376 qdisc noqueue state UNKNOWN group default
    link/ether 7a:30:47:3e:64:61 brd ff:ff:ff:ff:ff:ff
 4: weave: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1376 qdisc noqueue state UP group default
    link/ether 76:54:21:d6:20:ba brd ff:ff:ff:ff:ff:ff
    inet 10.244.0.1/16 brd 10.244.255.255 scope global weave
       valid_lft forever preferred_lft forever
...
controlplane /etc/cni/net.d ➜

The interface named "weave" is the bridge network interface created by Weave.

6. What is the Pod IP Address Range Configured by Weave?

The assigned IP address range for pods can be observed from the network interface details of the "weave" bridge:

inet 10.244.0.1/16 brd 10.244.255.255 scope global weave

To further verify, check the logs of one of the Weave pods:

kubectl logs -n kube-system weave-net-mknr5

Within the logs, you will find an entry such as:

INFO: 2023/07/26 03:42:17.199330 Command line options: map[... ipalloc-range:10.244.0.0/16 ... ]
...
10.244.0.1

This confirms that the pod IP address range is set to 10.244.0.0/16.

7. What is the Default Gateway for Pods Deployed on Node "node01"?

To determine the default gateway for pods running on "node01," follow these steps:

Step 1: Create a Busybox Pod Manifest

Generate a YAML configuration for a simple Busybox pod using the following command:

kubectl run busybox --image=busybox --dry-run=client -o yaml -- sleep 1000 > busybox.yaml

Next, edit the busybox.yaml file to ensure the pod is scheduled on "node01" by adding the nodeName field within the spec:

apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: null
  labels:
    run: busybox
  name: busybox
spec:
  nodeName: node01
  containers:
  - args:
    - sleep
    - "1000"
    image: busybox
    name: busybox
    resources: {}
  dnsPolicy: ClusterFirst
  restartPolicy: Always
status: {}

Step 2: Deploy the Busybox Pod

Apply the configuration file:

kubectl apply -f busybox.yaml

Verify that the pod is running:

kubectl get pod

Step 3: Inspect the Routing Table Inside the Pod

Once the Busybox pod is running, execute this command to view its routing table:

kubectl exec busybox -- ip route

The output should display the default route, for example:

default via 10.244.192.0 dev eth0
10.244.0.0/16 dev eth0 scope link src 10.244.192.1

This shows that the default gateway for pods on node "node01" is 10.244.192.0.

Summary

In this lab, we verified several important network configurations:

For additional information on Kubernetes networking, you can explore the Kubernetes Documentation.

Happy networking!