[toc]

1. k8s的yaml文件到底有多复杂

Kubernetes创建、更新、删除资源等操作时均可以使用json或yaml文件进行操作，更新和删除可以依赖之前的文件进行更改，但是创建具有多变形，往往编辑起来比较复杂，容器出错，而且k8s的配置项实在太多，稍微不注意就会犯错。要写好一个yaml文件，你需要了解yaml的语法，需要掌握k8s的各种配置，对于一个k8s的初学者而言，这将是一件很难的事情。

比如我们看一个同时创建一个Deployment、Service、Ingress的yaml文件内容：

---
apiVersion: v1
kind: Service
metadata:
labels:
app: test-yaml
name: test-yaml
namespace: freeswitch
spec:
ports:
- name: container-1-web-1
port: 8080
protocol: TCP
targetPort: 8080
selector:
app: test-yaml
sessionAffinity: None
type: ClusterIP
---
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
creationTimestamp: null
name: test-yaml
spec:
rules:
- host: test.com
http:
paths:
- backend:
serviceName: test-yaml
servicePort: 8080
path: /
---
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: test-yaml
name: test-yaml
namespace: freeswitch
spec:
replicas: 3
selector:
matchLabels:
app: test-yaml
strategy:
rollingUpdate:
maxSurge: 1
maxUnavailable: 0
type: RollingUpdate
template:
metadata:
annotations:
info: test for yaml
labels:
app: test-yaml
spec:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: app
operator: In
values:
- test-yaml
topologyKey: kubernetes.io/hostname
weight: 100
containers:
- env:
- name: TZ
value: Asia/Shanghai
- name: LANG
value: C.UTF-8
image: nginx
imagePullPolicy: Always
lifecycle: {}
livenessProbe:
failureThreshold: 2
initialDelaySeconds: 30
periodSeconds: 10
successThreshold: 1
tcpSocket:
port: 8080
timeoutSeconds: 2
name: test-yaml
ports:
- containerPort: 8080
name: web
protocol: TCP
readinessProbe:
failureThreshold: 2
initialDelaySeconds: 30
periodSeconds: 10
successThreshold: 1
tcpSocket:
port: 8080
timeoutSeconds: 2
resources:
limits:
cpu: 195m
memory: 375Mi
requests:
cpu: 10m
memory: 10Mi
securityContext:
allowPrivilegeEscalation: false
capabilities: {}
privileged: false
procMount: Default
readOnlyRootFilesystem: false
runAsNonRoot: false
volumeMounts:
- mountPath: /usr/share/zoneinfo/Asia/Shanghai
name: tz-config
- mountPath: /etc/localtime
name: tz-config
- mountPath: /etc/timezone
name: timezone
dnsPolicy: ClusterFirst
hostAliases:
- hostnames:
- www.baidu.com
ip: 114.114.114.114
imagePullSecrets:
- name: myregistrykey
- name: myregistrykey2
restartPolicy: Always
securityContext: {}
volumes:
- hostPath:
path: /usr/share/zoneinfo/Asia/Shanghai
type: ""
name: tz-config
- hostPath:
path: /etc/timezone
type: ""
name: timezone

这是一个包含了Service、Ingress、Deployment比较常用并且没有用到高级功能的yaml配置，就已经有上百行，如果是在添加了一些高级配置或者是Deployment中的容器不止一个，这个yaml会更大，就会造成一种视觉上疲劳，更改起来也比较麻烦而且非常容易出错。

2. 基于图形化的方式自动生成yaml

2.1 k8s图形化管理工具Ratel安装

本次采用Ratel自动生成yaml文件，Ratel安装文档：https://github.com/dotbalo/ratel-doc/blob/master/cluster/Install.md

2.2 使用Ratel创建生成yaml文件

2.2.1 基本配置

安装完成后，可以生成、创建管理常用的k8s核心资源，比如创建一个Deployment：
点击Deployment -- 创建如图所示：

之后可以填写一些基本的配置信息，比如Deployment名称、副本数、标签信息等，当然也可以点击必须/尽量部署至不同宿主机进行Pod亲和力的配置

同时也可添加一些复杂的配置，比如内核配置、容忍配置、节点亲和力快捷配置：

2.2.2 亲和力配置

基本配置编译完成以后，点击NEXT，下一个配置亲和力配置，如果上一页使用了亲和力快捷键，这边会自动生成亲和力配置，你可以再次编辑或者添加、删除：

2.2.3 存储配置

亲和力配置完成以后，可以点击NEXT进行存储配置，目前支持volume和projectedVolume配置，volume支持configMap、Secret、HostPath、PVC、NFS、Empty等常用类型的配置：

2.2.4 容器配置

接下来是容器配置，支持常用的容器配置，当然也可以添加多个容器：

 
稍微复制一点的配置：

2.2.4 初始化容器配置

初始化容器和容器配置类似

2.2.5 Service和Ingress配置

创建Deployment时可以一键添加Service和Ingress，添加Service时会自动读取容器的端口配置，添加Ingress时会自动读取Service配置

2.2.6 创建资源或生成yaml文件

上述配置完成以后，可以选择创建资源或生成yaml文件，假如点击生成yaml文件，会自动生成Service、Ingress、Deployment的yaml文件，可以直接拿着使用：

生成的内容如下：

---
apiVersion: v1
kind: Service
metadata:
creationTimestamp: null
labels:
app: test-yaml
name: test-yaml
namespace: default
spec:
ports:
- name: container-1-web-1
port: 8080
protocol: TCP
targetPort: 8080
selector:
app: test-yaml
sessionAffinity: None
type: ClusterIP
status:
loadBalancer: {}
---
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
creationTimestamp: null
name: test-yaml
spec:
rules:
- host: test.com
http:
paths:
- backend:
serviceName: test-yaml
servicePort: 8080
path: /
status:
loadBalancer: {}
---
apiVersion: apps/v1
kind: Deployment
metadata:
creationTimestamp: null
labels:
app: test-yaml
name: test-yaml
namespace: default
spec:
replicas: 3
selector:
matchLabels:
app: test-yaml
strategy:
rollingUpdate:
maxSurge: 1
maxUnavailable: 0
type: RollingUpdate
template:
metadata:
creationTimestamp: null
labels:
app: test-yaml
spec:
affinity:
nodeAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- preference:
matchExpressions:
- key: loki
operator: In
values:
- "true"
weight: 100
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- key: master
operator: NotIn
values:
- "true"
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: app
operator: In
values:
- test-yaml
topologyKey: kubernetes.io/hostname
containers:
- args:
- '*.jar --server.port=80'
command:
- java -jar
env:
- name: TZ
value: Asia/Shanghai
- name: LANG
value: C.UTF-8
- name: POD_IP
valueFrom:
fieldRef:
fieldPath: status.podIP
envFrom:
- configMapRef:
name: testcm
image: nginx
imagePullPolicy: IfNotPresent
lifecycle:
postStart:
exec:
command:
- echo "start"
preStop:
exec:
command:
- sleep 30
livenessProbe:
failureThreshold: 2
initialDelaySeconds: 30
periodSeconds: 10
successThreshold: 1
tcpSocket:
port: 8080
timeoutSeconds: 2
name: test-yaml
ports:
- containerPort: 8080
name: web
protocol: TCP
readinessProbe:
failureThreshold: 2
httpGet:
httpHeaders:
- name: a
value: b
path: /
port: 8080
scheme: HTTP
initialDelaySeconds: 30
periodSeconds: 10
successThreshold: 1
timeoutSeconds: 2
resources:
limits:
cpu: 493m
memory: 622Mi
requests:
cpu: 10m
memory: 10Mi
securityContext:
allowPrivilegeEscalation: false
capabilities: {}
privileged: false
procMount: Default
readOnlyRootFilesystem: false
runAsNonRoot: false
volumeMounts:
- mountPath: /usr/share/zoneinfo/Asia/Shanghai
name: tz-config
- mountPath: /etc/localtime
name: tz-config
- mountPath: /etc/timezone
name: timezone
- mountPath: /mnt
name: nfs-test
dnsPolicy: ClusterFirst
initContainers:
- args:
- init
command:
- echo
env:
- name: TZ
value: Asia/Shanghai
- name: LANG
value: C.UTF-8
image: nignx-init
imagePullPolicy: Always
name: init
resources:
limits:
cpu: 351m
memory: 258Mi
requests:
cpu: 10m
memory: 10Mi
securityContext:
allowPrivilegeEscalation: false
capabilities: {}
privileged: false
procMount: Default
readOnlyRootFilesystem: false
runAsNonRoot: false
volumeMounts:
- mountPath: /usr/share/zoneinfo/Asia/Shanghai
name: tz-config
- mountPath: /etc/localtime
name: tz-config
- mountPath: /etc/timezone
name: timezone
nodeSelector:
ratel: "true"
restartPolicy: Always
securityContext:
sysctls:
- name: net.core.somaxconn
value: "16384"
- name: net.ipv4.tcp_max_syn_backlog
value: "16384"
tolerations:
- effect: NoSchedule
key: node-role.kubernetes.io/master
operator: Exists
volumes:
- name: projected-test
projected:
defaultMode: 420
sources:
- downwardAPI:
items:
- fieldRef:
fieldPath: metadata.name
path: /opt/x
- hostPath:
path: /usr/share/zoneinfo/Asia/Shanghai
type: ""
name: tz-config
- hostPath:
path: /etc/timezone
type: ""
name: timezone
- name: nfs-test
nfs:
path: /data/nfs
server: 1.1.1.1
status: {}

这个yaml比之前的稍复杂，并且添加了一些高级配置，手动编写的还是比较麻烦的，所以用Ratel自动生成还是比较方便的，并且不会出错。

3. 其他资源文件自动生成

目前支持了很多资源文件的自动生成，比如：Deployment、StatefulSet、DaemonSet、Service、Ingress、CronJob、Secret、ConfigMap、PV、PVC等，可以大大减少我们的工作量和k8s的复杂度。

 
 
如果想要系统的学习k8s，可以专注下k8s的课程：

51CTO
 
腾讯课堂