Fix error log when agent fails to connect to K8s API

[tnqn]

When agent fails to connect to K8s API, the previous error logs about empty PodCIDR has no basis, and it shouldn't suggest users to check kube-controller-manager configuration.

E0803 11:57:06.000640 1 agent.go:926] "Spec.PodCIDR is empty for Node. Please make sure --allocate-node-cidrs is enabled for kube-controller-manager and --cluster-cidr specifies a sufficient CIDR range" err="failed to get Node with name kind-worker from K8s: Get "https://1.1.1.1/api/v1/nodes/kind-worker\": dial tcp 1.1.1.1:443: i/o timeout" nodeName="kind-worker"
F0803 11:57:06.001470 1 main.go:53] Error running agent: error initializing agent: CIDR string is empty for Node kind-worker

This patch fixes the error logs by differentiating ErrWaitTimeout from other errors. Only the former means the PodCIDR is empty.

When the PodCIDR is empty:

I0803 12:23:33.380739 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:23:38.388378 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:23:43.387476 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:23:48.388400 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:23:53.388225 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:23:58.387538 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:24:03.387824 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
I0803 12:24:03.393946 1 agent.go:921] "Waiting for Node PodCIDR configuration to complete" nodeName="kind-worker2"
E0803 12:24:03.394044 1 agent.go:928] "Spec.PodCIDR is empty for Node. Please make sure --allocate-node-cidrs is enabled for kube-controller-manager and --cluster-cidr specifies a sufficient CIDR range, or nodeIPAM is enabled for antrea-controller" err="timed out waiting for the condition" nodeName="kind-worker2"
F0803 12:24:03.396010 1 main.go:53] Error running agent: error initializing agent: Spec.PodCIDR is empty for Node kind-worker2

When connection fails:

F0803 12:17:02.228931 1 main.go:53] Error running agent: error initializing agent: failed to get Node with name kind-worker from K8s: Get "https://1.1.1.1/api/v1/nodes/kind-worker": dial tcp 1.1.1.1:443: i/o timeout

[tnqn]

/test-all