kubectl get allすると「Throttling request took～」と表示される
kubeadmでv1.18のK8sクラスタを構築し、状態を見ようとkubectl get pod -A
$ kubectl get pod -A NAMESPACE NAME READY STATUS RESTARTS AGE kube-system calico-kube-controllers-5b87c97847-4ttpm 1/1 Running 0 20m kube-system calico-node-4sxf7 1/1 Running 0 20m kube-system calico-node-77j2m 1/1 Running 0 19m kube-system calico-node-mn9wv 1/1 Running 0 19m kube-system calico-node-snppb 1/1 Running 0 19m kube-system coredns-66bff467f8-2hnmq 1/1 Running 0 19m kube-system coredns-66bff467f8-tttfx 1/1 Running 0 19m kube-system etcd-k8s-master-ad-1-0 1/1 Running 0 21m kube-system kube-apiserver-k8s-master-ad-1-0 1/1 Running 0 21m kube-system kube-controller-manager-k8s-master-ad-1-0 1/1 Running 0 21m kube-system kube-proxy-pzws8 1/1 Running 0 19m kube-system kube-proxy-t6zlx 1/1 Running 0 19m kube-system kube-proxy-tsfbf 1/1 Running 0 20m kube-system kube-proxy-vm7gq 1/1 Running 0 19m kube-system kube-scheduler-k8s-master-ad-1-0 1/1 Running 0 21m kubernetes-dashboard dashboard-metrics-scraper-6b4884c9d5-rkngz 1/1 Running 0 20m kubernetes-dashboard kubernetes-dashboard-67768d44c-v8hk6 1/1 Running 0 20m
API ServerやController Manager、kube-proxyなどの主要なコンポーネントが起動しており問題はなさそうです。
さらに他のResourceも見ようとkubectl get allしたのですが、下記のような出力がされなんだか不穏な感じに。
$ kubectl get all I0621 08:47:27.465470 4117 request.go:621] Throttling request took 1.160347495s, request: GET:https://10.0.0.2:6443/apis/discovery.k8s.io/v1beta1?timeout=32s I0621 08:47:37.465472 4117 request.go:621] Throttling request took 1.997902393s, request: GET:https://10.0.0.2:6443/apis/authorization.k8s.io/v1?timeout=32s I0621 08:47:47.465490 4117 request.go:621] Throttling request took 4.797741119s, request: GET:https://10.0.0.2:6443/apis/scheduling.k8s.io/v1?timeout=32s I0621 08:47:58.265434 4117 request.go:621] Throttling request took 1.198248957s, request: GET:https://10.0.0.2:6443/apis/apiregistration.k8s.io/v1?timeout=32s I0621 08:48:08.465434 4117 request.go:621] Throttling request took 4.198026222s, request: GET:https://10.0.0.2:6443/apis/rbac.authorization.k8s.io/v1?timeout=32s I0621 08:48:19.865431 4117 request.go:621] Throttling request took 1.197977955s, request: GET:https://10.0.0.2:6443/apis/apps/v1?timeout=32s NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE service/kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 19m
実はこれ、v1.18のNew Featureである「API Request Throttling」の影響のようです。
API request throttling (due to a high rate of requests) is now reported in client-go logs at log level 2.
上記にあるように、client-go logsとして出力されるみたいですね。
The presence of these messages may indicate to the administrator the need to tune the cluster accordingly.
ともあるように、クラスタのチューニングはした方がよいのかもしれません。 どのパラメータを変更すればいいの？という部分についてはまた時間をとって調べてみます。
