Update kubeconfig lookup with eksctl clusters

During eksctl bootstrap, the `/etc/systemd/system/kubelet.service` file still exists, although `kubelet.service.d/10-eksclt.al2.conf` and `/etc/eksctl/kubeconfig.yaml` are created and used.

This causes errors when running the `aws-cni-support.sh` script due to the precedence of looking up the kubeconfig file.

`error: unable to read certificate-authority /etc/kubernetes/pki/ca.crt for kubernetes due to open /etc/kubernetes/pki/ca.crt: no such file or directory`

scripts/aws-cni-support.sh

@@ -34,11 +34,11 @@ curl http://localhost:61678/metrics 2>&1 > ${LOG_DIR}/metrics.out
 # Collecting kubelet introspection data
 if [[ -n "${KUBECONFIG:-}" ]]; then
     command -v kubectl > /dev/null && kubectl get --kubeconfig=${KUBECONFIG} --raw=/api/v1/pods > ${LOG_DIR}/kubelet.out
+elif [[ -f /etc/eksctl/kubeconfig.yaml ]]; then
+    command -v kubectl > /dev/null && kubectl get --kubeconfig=/etc/eksctl/kubeconfig.yaml --raw=/api/v1/pods > ${LOG_DIR}/kubelet.out
 elif [[ -f /etc/systemd/system/kubelet.service ]]; then
     KUBECONFIG=`grep kubeconfig /etc/systemd/system/kubelet.service | awk '{print $2}'`
     command -v kubectl > /dev/null && kubectl get --kubeconfig=${KUBECONFIG} --raw=/api/v1/pods > ${LOG_DIR}/kubelet.out
-elif [[ -f /etc/eksctl/kubeconfig.yaml ]]; then
-    command -v kubectl > /dev/null && kubectl get --kubeconfig=/etc/eksctl/kubeconfig.yaml --raw=/api/v1/pods > ${LOG_DIR}/kubelet.out
 else
     echo "======== Unable to find KUBECONFIG, IGNORING POD DATA ========="
 fi