当前位置：网站首页>Kubernetes monitoring component metrics server deployment

Kubernetes monitoring component metrics server deployment

2022-07-25 07:27:00 【Brother Xing plays with the clouds】

from v1.8 Start , A measure of resource usage （ Like a container CPU And memory usage ） Can pass Metrics API obtain . Be careful ：

Metrics API Only the current measurement data can be queried , It doesn't save historical data
Metrics API URI by /apis/metrics.k8s.io/, stay k8s.io/metrics maintain
Must be deployed metrics-server To use the API,metrics-server By calling Kubelet Summary API get data

Open the aggregation layer (Aggregation Layer)

Some knowledge about the aggregation layer The aggregation layer runs on apiserver Inside the process , Allow users to colony Install additional Kubernetes Style API, Expand core API The function of . The aggregation layer needs to be started apiserver It can only be used when it is turned on .

Before users register extended resources , The aggregation layer does nothing . User to register API, You must add a APIService object , Used to declare API Of URL The path and the backend that handles the request APIService. thereafter , The aggregation layer will send all requests to that path (e.g. /apis/myextension.mycompany.io/v1/…) All forwarded to registered APIService.

In general ,APIService Object to extension-apiserver Running on the colony One of them pod in , If you need to actively manage the added resources ,extension-apiserver Also need to work with one or more controlller Association ,apiserver-builder It provides a framework for both parties .

Service Catalog yes Kubernetes A kind of API Extended implementation , convenient Kubernetes colony Internal application access colony external 、 Managed by a third party 、 Services provided , For example, database services provided by cloud providers .Service Catalog The installation of will provide for the services it provides extension-apiserver and controller Two extension components .

Open the aggregation layer modify apiserver The configuration file , Add the following startup parameters to enable aggregation layer：

--requestheader-client-ca-file=/etc/kubernetes/ssl/ca.pem --requestheader-allowed-names=aggregator --requestheader-extra-headers-prefix=X-Remote-Extra- --requestheader-group-headers=X-Remote-Group --requestheader-username-headers=X-Remote-User

Deploy metrics-server

git clone https://github.com/kubernetes-incubator/metrics-server cd metrics-server kubectl create -f deploy/1.8+/

You can see later metrics-server Run up ：

kubectl -n kube-system get pods -l k8s-app=metrics-server

remarks ： If gcr.io Can't access the image of metrics-server-deployment.yaml Replace the image in with ：daocloud.io/liukuan73/metrics-server-amd64:v0.2.1