kubernetes高级实战

一、模拟企业环境进行一个实战部署

[root@master node]# kubectl apply -f pod-tomcat.yaml
pod/tomcat-test created
[root@master node]# kubectl get pods
NAME          READY   STATUS    RESTARTS   AGE
tomcat-test   2/2     Running   0          2s
[root@master node]# kubectl get pods -o wide
NAME          READY   STATUS    RESTARTS   AGE   IP               NODE    NOMINATED NODE   READINESS GATES
tomcat-test   2/2     Running   0          8s    10.244.166.137   node1   <none>           <none>
[root@master node]# cat pod-tomcat.yaml
apiVersion: v1
kind: Pod
metadata:
name: tomcat-test
namespace: default
labels:
    app: tomcat
spec:
nodeName: node1
containers:
- name: tomcat-java
    ports:
    - containerPort: 8080
    image: xianchao/tomcat-8.5-jre8:v1
    imagePullPolicy: IfNotPresent
- name: busybox-test
    image: busybox:1.28
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 3600"

[root@node1 ~]# kubectl label node node2 version=v1

[root@master node]# kubectl apply -f pod-tomcat.yaml
pod/tomcat-test created
[root@master node]# kubectl get pod -owide
NAME          READY   STATUS    RESTARTS   AGE   IP              NODE    NOMINATED NODE   READINESS GATES
tomcat-test   2/2     Running   0          6s    10.244.104.10   node2   <none>           <none>
[root@master node]# cat pod-tomcat.yaml
apiVersion: v1
kind: Pod
metadata:
name: tomcat-test
namespace: default
labels:
    app: tomcat
spec:
nodeSelector:
    version: v1
containers:
- name: tomcat-java
    ports:
    - containerPort: 8080
    image: xianchao/tomcat-8.5-jre8:v1
    imagePullPolicy: IfNotPresent
- name: busybox-test
    image: busybox:1.28
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 3600"

[root@node1 ~]# kubectl label node node1 zone=foo

[root@master node]# kubectl apply -f affinity.yaml
pod/pod-affinity created
[root@master node]# kubectl get pod
NAME READY STATUS RESTARTS AGE
pod-affinity 1/1 Running 0 24s

[root@master node]# cat affinity.yaml
apiVersion: v1
kind: Pod
metadata:
name: pod-affinity
namespace: default
labels:
    app: myapp
spec:
affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar
containers:
- name: myapp
    image: docker.io/ikubernetes/myapp:v1
    imagePullPolicy: IfNotPresent

[root@master node]# kubectl apply -f pod-nodeaffinity-demo-2.yaml
pod/pod-node-affinity-demo-2 created
[root@master node]# cat pod-nodeaffinity-demo-2.yaml
apiVersion: v1
kind: Pod
metadata:
name: pod-node-affinity-demo-2
namespace: default
labels:
    app: myapp
    tier: frontend
spec:
containers:
- name: myapp
    image: docker.io/ikubernetes/myapp:v1
    imagePullPolicy: IfNotPresent
affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - preference:
          matchExpressions:
          - key: zone1
            operator: In
            values:
            - foo1
            - bar1
        weight: 10
      - preference:
          matchExpressions:
          - key: zone2
            operator: In
            values:
            - foo2
            - bar2
        weight: 20

[root@master node]# kubectl get pods
NAME                       READY   STATUS    RESTARTS   AGE
pod-affinity               1/1     Running   0          8m3s
pod-node-affinity-demo-2   1/1     Running   0          10s

二、命令＋命令解析

### 1. **部署第一个 Pod（`tomcat-test`）**
#### 操作：
```bash
kubectl apply -f pod-tomcat.yaml
```

#### 配置文件 (`pod-tomcat.yaml`)：
```yaml
apiVersion: v1
kind: Pod
metadata:
name: tomcat-test
namespace: default
labels:
    app: tomcat
spec:
nodeName: node1
containers:
- name: tomcat-java
    ports:
    - containerPort: 8080
    image: xianchao/tomcat-8.5-jre8:v1
    imagePullPolicy: IfNotPresent
- name: busybox-test
    image: busybox:1.28
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 3600"
```

#### 解释：
- **`nodeName: node1`**：指定 Pod 必须调度到 `node1` 节点。
- **`containers`**：
- `tomcat-java`：运行 Tomcat 应用，监听 8080 端口。
- `busybox-test`：运行一个 BusyBox 容器，执行 `sleep 3600` 命令（保持容器运行 1 小时）。
- **`imagePullPolicy: IfNotPresent`**：如果本地已有镜像，则不会从远程仓库拉取。

#### 查看 Pod：
```bash
kubectl get pods -o wide
```
- 输出：
```
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
tomcat-test 2/2 Running 0 8s 10.244.166.137 node1 <none> <none>
```
- Pod 状态为 `Running`，表示已成功调度到 `node1` 节点并运行。

---

### 2. **为 `node2` 添加标签**
#### 操作：
```bash
kubectl label node node2 version=v1
```

#### 解释：
- 为 `node2` 节点添加一个标签 `version=v1`。
- 标签可以用于节点选择器（`nodeSelector`）或亲和性规则（`affinity`）。

---

### 3. **部署第二个 Pod（`tomcat-test`）**
#### 操作：
```bash
kubectl apply -f pod-tomcat.yaml
```

#### 配置文件 (`pod-tomcat.yaml`)：
```yaml
apiVersion: v1
kind: Pod
metadata:
name: tomcat-test
namespace: default
labels:
    app: tomcat
spec:
nodeSelector:
    version: v1
containers:
- name: tomcat-java
    ports:
    - containerPort: 8080
    image: xianchao/tomcat-8.5-jre8:v1
    imagePullPolicy: IfNotPresent
- name: busybox-test
    image: busybox:1.28
    imagePullPolicy: IfNotPresent
    command:
    - "/bin/sh"
    - "-c"
    - "sleep 3600"
```

#### 解释：
- **`nodeSelector`**：指定 Pod 必须调度到带有 `version=v1` 标签的节点（即 `node2`）。
- 其他配置与第一个 Pod 相同。

#### 查看 Pod：
```bash
kubectl get pods -o wide
```
- 输出：
```
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
tomcat-test 2/2 Running 0 6s 10.244.104.10 node2 <none> <none>
```
- Pod 状态为 `Running`，表示已成功调度到 `node2` 节点并运行。

---

### 4. **为 `node1` 添加标签**
#### 操作：
```bash
kubectl label node node1 zone=foo
```

#### 解释：
- 为 `node1` 节点添加一个标签 `zone=foo`。
- 该标签将用于后续的亲和性调度。

---

### 5. **部署第三个 Pod（`pod-affinity`）**
#### 操作：
```bash
kubectl apply -f affinity.yaml
```

#### 配置文件 (`affinity.yaml`)：
```yaml
apiVersion: v1
kind: Pod
metadata:
name: pod-affinity
namespace: default
labels:
    app: myapp
spec:
affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar
containers:
- name: myapp
    image: docker.io/ikubernetes/myapp:v1
    imagePullPolicy: IfNotPresent
```

#### 解释：
- **`nodeAffinity`**：定义节点亲和性规则。
- **`requiredDuringSchedulingIgnoredDuringExecution`**：调度时必须满足的条件。
- **`matchExpressions`**：匹配节点的标签。
- `key: zone`：匹配标签键为 `zone` 的节点。
- `operator: In`：标签值必须在 `values` 列表中（即 `foo` 或 `bar`）。
- 由于 `node1` 有 `zone=foo` 标签，Pod 将被调度到 `node1`。

#### 查看 Pod：
```bash
kubectl get pods
```
- 输出：
```
NAME READY STATUS RESTARTS AGE
pod-affinity 1/1 Running 0 24s
```
- Pod 状态为 `Running`，表示已成功调度到 `node1` 节点并运行。

---

### 6. **部署第四个 Pod（`pod-node-affinity-demo-2`）**
#### 操作：
```bash
kubectl apply -f pod-nodeaffinity-demo-2.yaml
```

#### 配置文件 (`pod-nodeaffinity-demo-2.yaml`)：
```yaml
apiVersion: v1
kind: Pod
metadata:
name: pod-node-affinity-demo-2
namespace: default
labels:
    app: myapp
    tier: frontend
spec:
containers:
- name: myapp
    image: docker.io/ikubernetes/myapp:v1
    imagePullPolicy: IfNotPresent
affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - preference:
          matchExpressions:
          - key: zone1
            operator: In
            values:
            - foo1
            - bar1
        weight: 10
      - preference:
          matchExpressions:
          - key: zone2
            operator: In
            values:
            - foo2
            - bar2
        weight: 20
```

#### 解释：
- **`nodeAffinity`**：定义节点亲和性规则。
- **`preferredDuringSchedulingIgnoredDuringExecution`**：调度时优先满足的条件（非强制）。
- **`preference`**：定义偏好规则。
    - `key: zone1` 和 `key: zone2`：匹配节点的标签。
    - `operator: In`：标签值必须在 `values` 列表中。
    - `weight`：规则的权重（值越大，优先级越高）。
- 由于没有节点满足 `zone1` 或 `zone2` 的标签，Pod 将被调度到任意可用节点。

#### 查看 Pod：
```bash
kubectl get pods
```
- 输出：
```
NAME                       READY   STATUS    RESTARTS   AGE
pod-affinity               1/1     Running   0          8m3s
pod-node-affinity-demo-2   1/1     Running   0          10s
```
- Pod 状态为 `Running`，表示已成功调度并运行。

---

### 总结
- **`nodeName`**：强制 Pod 调度到指定节点。
- **`nodeSelector`**：根据节点标签调度 Pod。
- **`nodeAffinity`**：定义更复杂的节点亲和性规则（强制或偏好）。
- 通过这些机制，可以灵活控制 Pod 的调度行为，满足不同的部署需求。

如果还有其他问题，欢迎继续提问！