Kubernetes install with kubespray

업데이트: 2021.09.13

Kubespray 를 이용한 Kubernetes 설치 (Baremetal)

사전준비

서버 노드 준비

3대 Node

168.28.15
168.28.16
168.28.17

서버 노드 환경설정 : kubernetes-101 참고

Swap 영역을 비활성화
SELinux Disable
방화벽 Disable
브릿지 네트워크 할성화
설치 노드에서의 ssh 접속 허용 : SSH 키복사 ssh-copy-id

설치 준비

Git : Repository Clone
Python3 : Inventory 및 환경 설정을 위한 스크립트 실행
Ansible : 원격 실행(설치) ansible-playbook

Repository clone 및 Python package install

$ git clone https://github.com/kubernetes-sigs/kubespray
$ cd kubespray

# Package install
$ pip3 install -r requirements.txt

설치 방법 1 : `inventory_builder`

inventory_builder Python script 활용 inventory 구성 및 설치

# Install 을 위한 Inventory 생성을 위한 공간 생성(복사)
$ cp -rfp inventory/sample inventory/glass

# IP 3개 지정, Node 설정 → inventory/glass/hosts.yaml
$ declare -a IPS=(192.168.28.15 192.168.28.16 192.168.28.17)
$ CONFIG_FILE=inventory/glass/hosts.yaml python3 contrib/inventory_builder/inventory.py ${IPS[@]}
DEBUG: Adding group all
DEBUG: Adding group kube_control_plane
DEBUG: Adding group kube_node
DEBUG: Adding group etcd
DEBUG: Adding group k8s_cluster
DEBUG: Adding group calico_rr
DEBUG: adding host node1 to group all
DEBUG: adding host node2 to group all
DEBUG: adding host node3 to group all
DEBUG: adding host node1 to group etcd
DEBUG: adding host node2 to group etcd
DEBUG: adding host node3 to group etcd
DEBUG: adding host node1 to group kube_control_plane
DEBUG: adding host node2 to group kube_control_plane
DEBUG: adding host node1 to group kube_node
DEBUG: adding host node2 to group kube_node
DEBUG: adding host node3 to group kube_node

# inventory 자동 구성 - 3개 노드 
#   kube_control_plane: node1,node2 
#   workder: node3
$ cat inventory/glass/hosts.yaml
all:
  hosts:
    node1:
      ansible_host: 192.168.28.15
      ip: 192.168.28.15
      access_ip: 192.168.28.15
    node2:
      ansible_host: 192.168.28.16
      ip: 192.168.28.16
      access_ip: 192.168.28.16
    node3:
      ansible_host: 192.168.28.17
      ip: 192.168.28.17
      access_ip: 192.168.28.17
  children:
    kube_control_plane:
      hosts:
        node1:
        node2:
    kube_node:
      hosts:
        node1:
        node2:
        node3:
    etcd:
      hosts:
        node1:
        node2:
        node3:
    k8s_cluster:
      children:
        kube_control_plane:
        kube_node:
    calico_rr:
      hosts: {}


## Install // 오래 걸림 
$ ansible-playbook -i inventory/glass/hosts.yaml  --become --become-user=root cluster.yml

Cluster 확인

$ kubectl get node
NAME    STATUS   ROLES                  AGE   VERSION
node1   Ready    control-plane,master   40m   v1.18.6
node2   Ready    control-plane,master   40m   v1.18.6
node3   Ready    <none>                 38m   v1.18.6

$ kubectl get pod -A
NAMESPACE     NAME                                       READY   STATUS             RESTARTS   AGE
kube-system   calico-kube-controllers-8575b76f66-kx2sv   0/1     CrashLoopBackOff   12         40m
kube-system   calico-node-g9xmm                          1/1     Running            0          40m
kube-system   calico-node-pg7dt                          1/1     Running            0          40m
kube-system   calico-node-rwbcf                          1/1     Running            0          40m
kube-system   coredns-8474476ff8-26p2b                   1/1     Running            0          39m
kube-system   coredns-8474476ff8-j5k4w                   1/1     Running            0          39m
kube-system   dns-autoscaler-7df78bfcfb-7r474            1/1     Running            0          39m
kube-system   kube-apiserver-node1                       1/1     Running            0          42m
kube-system   kube-apiserver-node2                       1/1     Running            0          41m
kube-system   kube-controller-manager-node1              1/1     Running            0          42m
kube-system   kube-controller-manager-node2              1/1     Running            0          41m
kube-system   kube-proxy-4lqh5                           1/1     Running            0          40m
kube-system   kube-proxy-6tg8t                           1/1     Running            0          40m
kube-system   kube-proxy-brljv                           1/1     Running            0          40m
kube-system   kube-scheduler-node1                       1/1     Running            0          42m
kube-system   kube-scheduler-node2                       1/1     Running            0          41m
kube-system   nginx-proxy-node3                          1/1     Running            0          40m
kube-system   nodelocaldns-76jw6                         1/1     Running            0          39m
kube-system   nodelocaldns-k87js                         1/1     Running            0          39m
kube-system   nodelocaldns-lnlqt                         1/1     Running            0          39m

서비스 설치 및 확인

mvcapp-deploy-service.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: mvcapp
spec:
  selector:
    matchLabels:
      app: mvcapp
  replicas: 4 
  template: 
    metadata:
      labels:
        app: mvcapp
    spec:
      containers:
      - name: mvcapp
        image: cdecl/mvcapp:0.6
        imagePullPolicy: Always
        ports:
        - containerPort: 80

---
apiVersion: v1
kind: Service
metadata:
  name: mvcapp
spec:
  type: NodePort
  selector:
    app: mvcapp
  ports:
  - port: 80
    targetPort: 80
    nodePort: 30010

$ kubectl apply -f mvcapp-deploy-service.yaml
deployment.apps/mvcapp created
service/mvcapp created

$ kubectl get pod -owide
NAME                      READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
mvcapp-5c56c55f76-c245s   1/1     Running   0          44s   10.233.96.4   node2   <none>           <none>
mvcapp-5c56c55f76-nt7bw   1/1     Running   0          44s   10.233.92.5   node3   <none>           <none>
mvcapp-5c56c55f76-pjr8c   1/1     Running   0          44s   10.233.96.3   node2   <none>           <none>
mvcapp-5c56c55f76-xthgs   1/1     Running   0          44s   10.233.92.6   node3   <none>           <none>]

$ kubectl get svc
NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.233.0.1      <none>        443/TCP        8m42s
mvcapp       NodePort    10.233.12.111   <none>        80:30010/TCP   83s

# cluster ip 로 호출
$ curl -s 10.233.12.111:80
    * Project           : mvcapp
    * Version           : 0.6 / net5.0
    * Hostname          : mvcapp-5c56c55f76-frbfs
    * Sleep(sync)       : 0
    * RemoteAddr        : 10.233.90.0
    * X-Forwarded-For   : 
    * Request Count     : 1
    * User-Agent        : curl/7.29.0

# node port (kube-proxy) 호출 
$ curl -s node1:30010
    * Project           : mvcapp
    * Version           : 0.6 / net5.0
    * Hostname          : mvcapp-5c56c55f76-c82ws
    * Sleep(sync)       : 0
    * RemoteAddr        : 10.233.90.0
    * X-Forwarded-For   : 
    * Request Count     : 1
    * User-Agent        : curl/7.29.0

설치 방법 2 : inventory 직접 수정

inventory.ini 수정 및 설치

# Install 을 위한 Inventory 생성을 위한 공간 생성(복사)
$ cp -rfp inventory/sample inventory/glass

inventory/glass/inventory.ini
- kube_control_plane 1개만 설정 : node1

# ## Configure 'ip' variable to bind kubernetes services on a
# ## different ip than the default iface
# ## We should set etcd_member_name for etcd cluster. The node that is not a etcd member do not need to set the value, or can set the empty string value.
[all]
node1 ansible_host=192.168.28.15
node2 ansible_host=192.168.28.16
node3 ansible_host=192.168.28.17

# ## configure a bastion host if your nodes are not directly reachable
# [bastion]
# bastion ansible_host=x.x.x.x ansible_user=some_user

[kube_control_plane]
node1

[etcd]
node1
node2
node3

[kube_node]
node2
node3
# node4
# node5
# node6

[calico_rr]

[k8s_cluster:children]
kube_control_plane
kube_node
calico_rr

## Install
$ ansible-playbook -i inventory/glass/inventory.ini  --become --become-user=root cluster.yml

Cluster 확인

$ kubectl get node
NAME    STATUS   ROLES                  AGE     VERSION
node1   Ready    control-plane,master   3m41s   v1.18.6
node2   Ready    <none>                 2m37s   v1.18.6
node3   Ready    <none>                 2m38s   v1.18.6

$ kubectl get pod -A
NAMESPACE     NAME                                       READY   STATUS             RESTARTS   AGE
kube-system   calico-kube-controllers-8575b76f66-nm6ll   0/1     CrashLoopBackOff   5          4m35s
kube-system   calico-node-9xqmz                          1/1     Running            0          4m55s
kube-system   calico-node-dbr4p                          1/1     Running            0          4m55s
kube-system   calico-node-h4xcv                          1/1     Running            0          4m55s
kube-system   coredns-8474476ff8-6rx9f                   1/1     Running            0          4m20s
kube-system   coredns-8474476ff8-r7nrf                   1/1     Running            0          4m15s
kube-system   dns-autoscaler-7df78bfcfb-g482g            1/1     Running            0          4m17s
kube-system   kube-apiserver-node1                       1/1     Running            0          6m6s
kube-system   kube-controller-manager-node1              1/1     Running            0          6m6s
kube-system   kube-proxy-6rxsz                           1/1     Running            0          5m12s
kube-system   kube-proxy-jt2dl                           1/1     Running            0          5m12s
kube-system   kube-proxy-kdxch                           1/1     Running            0          5m12s
kube-system   kube-scheduler-node1                       1/1     Running            0          6m6s
kube-system   nginx-proxy-node2                          1/1     Running            0          5m12s
kube-system   nginx-proxy-node3                          1/1     Running            0          5m12s
kube-system   nodelocaldns-2tjm8                         1/1     Running            0          4m16s
kube-system   nodelocaldns-pzkjq                         1/1     Running            0          4m16s
kube-system   nodelocaldns-r9gbv                         1/1     Running            0          4m16s

$ kubectl apply -f mvcapp-deploy-service.yaml
deployment.apps/mvcapp created
service/mvcapp created

$ kubectl get pod -owide
NAME                      READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
mvcapp-5c56c55f76-c82ws   1/1     Running   0          48s   10.233.96.3   node2   <none>           <none>
mvcapp-5c56c55f76-frbfs   1/1     Running   0          48s   10.233.92.2   node3   <none>           <none>
mvcapp-5c56c55f76-kx2bm   1/1     Running   0          48s   10.233.96.2   node2   <none>           <none>
mvcapp-5c56c55f76-xvlbp   1/1     Running   0          48s   10.233.92.1   node3   <none>           <none>

클러스터 고가용성 구조

https://github.com/kubernetes-sigs/kubespray/blob/master/docs/ha-mode.md

Etcd 클러스터 구성

Kube-apiserver

kubeadm 고가용성 구조의 경우 kube-apiserver 서버접근을 위한 HA (loadbalancer) 가 필요
kubespray 의 경우 nginx-based 기반 reverse proxy 각 node 에 내장되어 있어서 별도 LB 구성이 필요 없음
- local internal loadbalancer 를 사용하지 않은 경우 별도 HA 구성 가능

Node 추가 및 Cluster 변경

https://github.com/kubernetes-sigs/kubespray/blob/master/docs/nodes.md

1) Control-plane 변경

Change order of current control planes
Upgrade the cluster

inventory 수정 후, cluster.yml 실행

$ ansible-playbook -i inventory/glass/hosts.yaml cluster.yml

2) Adding/replacing a worker node

inventory 수정 후, scale.yml 실행

# 전체 갱신 
$ ansible-playbook -i inventory/glass/hosts.yaml scale.yml

# 추가된 node3 만 갱신 
$ ansible-playbook -i inventory/glass/hosts.yaml scale.yml --limit=node3

3) Remove a worker node

$ ansible-playbook -i inventory/glass/hosts.yaml remove-node.yml -e node=node3
...

$ kubectl get node
NAME    STATUS                        ROLES                  AGE   VERSION
node1   Ready                         control-plane,master   86m   v1.18.6
node2   Ready                         control-plane,master   85m   v1.18.6
node3   NotReady,SchedulingDisabled   <none>                 19m   v1.18.6


$ kubectl delete node node3
node "node3" deleted

$ kubectl get node
NAME    STATUS   ROLES                  AGE   VERSION
node1   Ready    control-plane,master   88m   v1.18.6
node2   Ready    control-plane,master   87m   v1.18.6

초기화 : 클러스터 삭제

$ ansible-playbook -i inventory/glass/hosts.yaml  --become --become-user=root reset.yml

Byung Kyu KIM

Kubernetes install with kubespray

사전준비

서버 노드 준비

서버 노드 환경설정 : kubernetes-101 참고

설치 준비

설치 방법 1 : `inventory_builder`

Cluster 확인

서비스 설치 및 확인

설치 방법 2 : inventory 직접 수정

Cluster 확인

클러스터 고가용성 구조

Etcd 클러스터 구성

Kube-apiserver

Node 추가 및 Cluster 변경

1) Control-plane 변경

2) Adding/replacing a worker node

3) Remove a worker node

초기화 : 클러스터 삭제

댓글남기기

참고

TLS와 SSH 프로토콜 기능

GNU C++로 정적 라이브러리 링크하는 방법

유니코드 NFD to NFC 변환

Ollama를 이용한 Mistral 로컬 실행 가이드

Byung Kyu KIM

사전준비

서버 노드 준비

서버 노드 환경설정 : kubernetes-101 참고

설치 준비

설치 방법 1 : inventory_builder

Cluster 확인

서비스 설치 및 확인

설치 방법 2 : inventory 직접 수정

Cluster 확인

클러스터 고가용성 구조

Etcd 클러스터 구성

Kube-apiserver

Node 추가 및 Cluster 변경

1) Control-plane 변경

2) Adding/replacing a worker node

3) Remove a worker node

초기화 : 클러스터 삭제

댓글남기기

참고

TLS와 SSH 프로토콜 기능

GNU C++로 정적 라이브러리 링크하는 방법

유니코드 NFD to NFC 변환

Ollama를 이용한 Mistral 로컬 실행 가이드

설치 방법 1 : `inventory_builder`