在Openshift上通过yaml部署应用
1.通过直接执行yaml
通过如下命令直接执行
oc create -f nginx.yml
nginx.yml
apiVersion: v1
items:
- apiVersion: apps.openshift.io/v1
# okd 部署配置(dc)，与 k8s Deployment 资源对象类似，以启动多个容器的方式生成 pod
kind: DeploymentConfig
metadata:
# 标签，在查询时具体资源对象时非常重要，如：>oc get dc -l app=nginx
labels:
app: nginx
name: nginx
# 选定项目空间
namespace: test
spec:
# 副本数，即 nginx app 部署的实例数
replicas: 1
# pod 的计算资源配额
resources:
# pod 能分配的最大计算资源
limits:
cpu: 300m
memory: 1024Mi
# pod 分配的最少计算资源
requests:
cpu: 100m
memory: 200Mi
# 选择器 Service 根据此项来绑定到 dc
selector:
name: nginx
strategy:
# type: Recreate
type: Rolling
# dc 根据模板里内容创建 pod
template:
metadata:
labels:
app: nginx
name: nginx
deploymentconfig: nginx
spec:
# 容器集合
containers:
- capabilities: {}
# 容器内环境变量，下文给这个容器设置了时区和语言的环境变量
env:
- name: TZ
value: Asia/Shanghai
- name: LANG
value: en_US.UTF-8
# 容器使用什么镜像部署，在创建时需要替换成实际要部署的镜像
image: nginx:1.16
# 镜像下载策略，总是下载最新的(Always)
imagePullPolicy: IfNotPresent
ports:
- containerPort: 80
protocol: TCP
# 健康检查-pod 是否存活
livenessProbe:
failureThreshold: 2
# http get 请求的方式验证 pod-ip:80/
httpGet:
path: /
port: 80
initialDelaySeconds: 60
periodSeconds: 60
timeoutSeconds: 5
name: nginx
# 健康检查-pod 是否就绪
readinessProbe:
httpGet:
path: /
port: 80
initialDelaySeconds: 3
timeoutSeconds: 5
# 容器计算资源配额，与 pod 配额类似
resources:
limits:
cpu: 300m
memory: 1024Mi
requests:
cpu: 100m
memory: 200Mi
securityContext:
capabilities: {}
# privileged: true
terminationMessagePath: /dev/termination-log
# 将 pod 中配置的卷挂载到容器内
# volumeMounts:
# - mountPath: /data
# name: nginx
dnsPolicy: ClusterFirst
# 使用节点选择器将应用固定部署到 node1 计算节点上
# nodeSelector:
# kubernetes.io/hostname: node1.app.com
restartPolicy: Always
# 配置和定义使用实际计算节点主机文件夹地址这种类型的卷
# volumes:
# - persistentVolumeClaim:
# claimName: nginx
# name: nginx
# pod 触发器--配置变动触发更新
triggers:
- type: ConfigChange
- apiVersion: v1
# 服务，与 k8s 中 service 一样，将 dc 上部署的应用暴露给内部(多)或外部(少)
kind: Service
metadata:
labels:
app: nginx
name: nginx
spec:
# 应用中要暴露的端口信息
ports:
- name: http
# 对外暴露的端口
port: 80
protocol: TCP
# 应用实际端口
targetPort: 80
# 通过选择器选择 dc
selector:
name: nginx
sessionAffinity: None
type: ClusterIP
- apiVersion: route.openshift.io/v1
# OKD 平台特有资源，与 k8s 中 ingress 类似，用于将 service 正真暴露给外部使用，但只能使用域名访问
kind: Route
metadata:
labels:
app: nginx
name: nginx
spec:
# 配置域名
host: xinchen.app.com
to:
kind: Service
name: nginx
kind: List
metadata: {}
2. 通过创建template
参考文档: https://docs.okd.io/latest/dev_guide/templates.html#writing-templates
相关指令
# 上传模板
oc create -f <filename> -n <project>
# 使用模板
oc process -f nginx-template -p IMAGE_NAME=nginx:1.6 -p REPLICA_COUNT=1
# 编辑模板
oc edit template <templateName>
# 导出模板
oc get -o yaml --export all > <yaml_filename>
nginx-template.yml
kind: Template
apiVersion: v1
metadata:
name: nginx-template
objects:
- kind: DeploymentConfig
apiVersion: v1
metadata:
name: nginx
namespace: xinchen
labels:
app: nginx
spec:
replicas: ${REPLICA_COUNT}
selector:
name: nginx
resources:
limits:
cpu: 300m
memory: 1024Mi
requests:
cpu: 100m
memory: 200Mi
strategy:
type: Rolling
template:
metadata:
labels:
app: nginx
name: nginx
deploymentconfig: nginx
spec:
containers:
- capabilities: {}
env:
- name: TZ
value: Asia/Shanghai
- name: LANG
value: en_US.UTF-8
image: ${IMAGE_NAME}
imagePullPolicy: IfNotPresent
ports:
- containerPort: 80
protocol: TCP
livenessProbe:
failureThreshold: 2
httpGet:
path: /
port: 80
initialDelaySeconds: 60
periodSeconds: 60
timeoutSeconds: 5
name: nginx
readinessProbe:
httpGet:
path: /
port: 80
initialDelaySeconds: 3
timeoutSeconds: 5
resources:
limits:
cpu: 300m
memory: 1024Mi
requests:
cpu: 100m
memory: 200Mi
securityContext:
capabilities: {}
# privileged: true
terminationMessagePath: /dev/termination-log
# volumeMounts:
# - mountPath: /data
# name: nginx
dnsPolicy: ClusterFirst
restartPolicy: Always
# volumes:
# - persistentVolumeClaim:
# claimName: nginx
# name: nginx
# pod 触发器--配置变动触发更新
triggers:
- type: ConfigChange
- kind: Service
apiVersion: v1
metadata:
name: nginx
labels:
app: nginx
spec:
ports:
- name: 80-tcp
port : 80
protocol: TCP
targetPort: 80
selector:
name: nginx
sessionAffinity: None
type: ClusterIP
- apiVersion: route.openshift.io/v1
kind: Route
metadata:
labels:
app: nginx
name: nginx
spec:
host: ${HOST_NAME}
port:
targetPort: 80-tcp
to:
kind: Service
name: nginx
weight: 100
parameters:
- name: HOST_NAME
displayName: Host Name
required: true
- name: IMAGE_NAME
displayName: Image Name
required: true
- name: REPLICA_COUNT
displayName: Replica Count
value: "1"
required: true
3. 进入运行的容器
参考: https://docs.okd.io/latest/dev_guide/ssh_environment.html
# 查看pod名
oc get pods
# 进入容器
oc rsh <pod>
# 或者直接进入dc
oc rsh dc/nginx
4. 删除所有
# 删除所有
oc delete all -l app=nginx