环境：
- kubeadm v1.13.4部署
从 v1.8 开始，资源使用情况的度量（如容器的 CPU 和内存使用）可以通过 Metrics API 获取。注意
- Metrics API 只可以查询当前的度量数据，并不保存历史数据
- Metrics API URI 为
/apis/metrics.k8s.io/，在 k8s.io/metrics 维护 - 必须部署 metrics-server 才能使用该 API，metrics-server 通过调用 Kubelet Summary API 获取数据
开始部署metrics-server
克隆github上的项目
git clone https://github.com/kubernetes-incubator/metrics-server.git
项目克隆下来，官方的教程里面非常简单，直接kubectl apply -f 1.8+/就好了。但是实际操作中远不止这些，我在操作过程中就出现证书错误，无法访问metircs等问题。
error: metrics not available yet
查看日志后发现各种证书错误，获取不到每个node的10250端口。
所以我们需要对官方的deployment文件进行修改下：
$ cd metrics-server/deploy/1.8+
$ ls
aggregated-metrics-reader.yaml auth-delegator.yaml auth-reader.yaml metrics-apiservice.yaml metrics-server-deployment.yaml metrics-server-service.yaml resource-reader.yaml
里面有一个deployment文件，原内容如下:
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: metrics-server
namespace: kube-system
---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
name: metrics-server
namespace: kube-system
labels:
k8s-app: metrics-server
spec:
selector:
matchLabels:
k8s-app: metrics-server
template:
metadata:
name: metrics-server
labels:
k8s-app: metrics-server
spec:
serviceAccountName: metrics-server
volumes:
# mount in tmp so we can safely use from-scratch images and/or read-only containers
- name: tmp-dir
emptyDir: {}
containers:
- name: metrics-server
image: k8s.gcr.io/metrics-server-amd64:v0.3.1
imagePullPolicy: Always
volumeMounts:
- name: tmp-dir
mountPath: /tmp
将其修改一下,修改后如下：
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: metrics-server
namespace: kube-system
---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
name: metrics-server
namespace: kube-system
labels:
k8s-app: metrics-server
spec:
selector:
matchLabels:
k8s-app: metrics-server
template:
metadata:
name: metrics-server
labels:
k8s-app: metrics-server
spec:
serviceAccountName: metrics-server
volumes:
# mount in tmp so we can safely use from-scratch images and/or read-only containers
- name: tmp-dir
emptyDir: {}
containers:
- name: metrics-server
image: mirrorgooglecontainers/metrics-server-amd64:v0.3.1
imagePullPolicy: IfNotPresent
command:
- /metrics-server
- --kubelet-insecure-tls
- --kubelet-preferred-address-types=InternalDNS,InternalIP,ExternalDNS,ExternalIP,Hostname
volumeMounts:
- name: tmp-dir
mountPath: /tmp
resources:
limits:
cpu: 300m
memory: 200Mi
requests:
cpu: 200m
memory: 100Mi
修改后我们增加了资源限制和参数， 默认镜像是需要梯子的，这里改成国内可访问的镜像仓库，资源限制的requests其实还可以在低点。
kubectl apply -f metrics-server/deploy/1.8+/
等待几分钟后就可以查看node信息了
$ kubectl top node
NAME CPU(cores) CPU% MEMORY(bytes) MEMORY%
10.1.65.104 119m 2% 3982Mi 50%
10.1.65.105 119m 2% 2502Mi 31%
10.1.65.110 63m 1% 1676Mi 21%
10.1.65.111 61m 1% 1634Mi 20%
10.1.65.123 86m 2% 1476Mi 18%
10.1.65.124 81m 2% 1480Mi 18%
10.1.65.125 56m 1% 1421Mi 18%
10.1.65.126 105m 2% 1474Mi 18%
10.1.65.135 104m 2% 660Mi 8%
10.1.65.136 66m 1% 1424Mi 18%
10.1.65.137 89m 2% 1459Mi 18%
10.1.65.76 183m 4% 2361Mi 29%
10.1.65.77 121m 3% 2150Mi 27%
10.1.65.78 183m 4% 2579Mi 32%
10.1.65.79 1967m 24% 10222Mi 64%
10.1.65.80 76m 0% 8165Mi 51%
10.1.65.81 737m 9% 8537Mi 53%
10.1.65.82 129m 3% 694Mi 8%
10.1.65.83 162m 4% 4657Mi 59%
10.1.65.84 117m 2% 4649Mi 59%
10.1.65.85 135m 3% 692Mi 8%
10.1.65.86 145m 3% 4554Mi 57%
10.1.65.88 215m 2% 11599Mi 72%
10.1.65.89 1894m 23% 13540Mi 84%
