Kubernetes nodeName Manual Scheduling practice (K8S节点名称绑定以及手工调度)

最新推荐文章于 2025-05-06 13:35:51 发布

Moshow郑锴

最新推荐文章于 2025-05-06 13:35:51 发布

阅读量746

点赞数 17

分类专栏： Cloud|Docker|Kubernetes 文章标签： kubernetes 容器云原生

本文链接：https://blog.youkuaiyun.com/moshowgame/article/details/147168347

版权

Cloud|Docker|Kubernetes 专栏收录该内容

67 篇文章

订阅专栏

Manual Scheduling

在 Kubernetes 中，手动调度框架允许您将 Pod 分配到特定节点，而无需依赖默认调度器。这对于测试、调试或处理特定工作负载非常有用。您可以通过在 Pod 的规范中设置 nodeName 字段来实现手动调度。以下是一个示例：

apiVersion: v1
kind: Pod
metadata:
  name: manual-scheduled-pod
spec:
  containers:
  - name: nginx
    image: nginx
  nodeName: your-node-name

将 your-node-name 替换为您希望 Pod 运行的节点名称。应用此配置后，Kubernetes 会直接将 Pod 放置在指定的节点上。

nodeName

nodeName 是比亲和性或者 nodeSelector 更为直接的形式。nodeName 是 Pod 规约中的一个字段。如果 nodeName 字段不为空，调度器会忽略该 Pod，而指定节点上的 kubelet 会尝试将 Pod 放到该节点上。使用 nodeName 规则的优先级会高于使用 nodeSelector 或亲和性与非亲和性的规则。

使用 nodeName 来选择节点的方式有一些局限性：

如果所指代的节点不存在，则 Pod 无法运行，而且在某些情况下可能会被自动删除。
如果所指代的节点无法提供用来运行 Pod 所需的资源，Pod 会失败，而其失败原因中会给出是否因为内存或 CPU 不足而造成无法运行。
在云环境中的节点名称并不总是可预测的，也不总是稳定的。

警告：nodeName 旨在供自定义调度器或需要绕过任何已配置调度器的高级场景使用。如果已分配的 Node 负载过重，绕过调度器可能会导致 Pod 失败。你可以使用节点亲和性或 nodeselector 字段将 Pod 分配给特定 Node，而无需绕过调度器。

下面是一个使用 nodeName 字段的 Pod 规约示例：

apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  containers:
  - name: nginx
    image: nginx
  nodeName: kube-01

上面的 Pod 只能运行在节点 kube-01 之上。

Practice Detail

        Welcome to the KodeKloud Hands-On lab                                                                          
    __ ______  ____  ________ __ __    ____  __  ______ 
   / //_/ __ \/ __ \/ ____/ //_// /   / __ \/ / / / __ \
  / ,< / / / / / / / __/ / ,<  / /   / / / / / / / / / /
 / /| / /_/ / /_/ / /___/ /| |/ /___/ /_/ / /_/ / /_/ / 
/_/ |_\____/_____/_____/_/ |_/_____/\____/\____/_____/  
                                                        
            All rights reserved                                                                                        

controlplane ~ ➜  kubectl create -f nginx.yaml
pod/nginx created

controlplane ~ ➜  kubectl get pods
NAME    READY   STATUS    RESTARTS   AGE
nginx   0/1     Pending   0          8s

controlplane ~ ✖ kubectl describe pod nginx
Name:             nginx
Namespace:        default
Priority:         0
Service Account:  default
Node:             <none>
Labels:           <none>
Annotations:      <none>
Status:           Pending
IP:               
IPs:              <none>
Containers:
  nginx:
    Image:        nginx
    Port:         <none>
    Host Port:    <none>
    Environment:  <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-bvfbc (ro)
Volumes:
  kube-api-access-bvfbc:
    Type:                    Projected (a volume that contains injected data from multiple sources)
    TokenExpirationSeconds:  3607
    ConfigMapName:           kube-root-ca.crt
    ConfigMapOptional:       <nil>
    DownwardAPI:             true
QoS Class:                   BestEffort
Node-Selectors:              <none>
Tolerations:                 node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                             node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:                      <none>

controlplane ~ ➜  kubectl get pods -n kube-system
NAME                                   READY   STATUS    RESTARTS   AGE
coredns-7484cd47db-mhzfg               1/1     Running   0          10m
coredns-7484cd47db-pq6v7               1/1     Running   0          10m
etcd-controlplane                      1/1     Running   0          10m
kube-apiserver-controlplane            1/1     Running   0          10m
kube-controller-manager-controlplane   1/1     Running   0          10m
kube-proxy-dd6zp                       1/1     Running   0          9m48s
kube-proxy-mtmxb                       1/1     Running   0          10m

controlplane ~ ➜  echo "no scheduler present"
no scheduler present

controlplane ~ ➜  cat nginx.yaml 
---
apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  containers:
  -  image: nginx
     name: nginx


controlplane ~ ➜  vi nginx.yaml

controlplane ~ ➜  cat nginx.yaml 
---
apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  nodeName: node01
  containers:
  -  image: nginx
     name: nginx


controlplane ~ ➜  echo "Manually schedule the pod on node01"
Manually schedule the pod on node01

controlplane ~ ➜  cat nginx.yaml
---
apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  nodeName: node01
  containers:
  -  image: nginx
     name: nginx


controlplane ~ ➜  kubectl replace --force -f nginx.yaml
pod "nginx" deleted
pod/nginx replaced

controlplane ~ ➜  kubectl get pods -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP           NODE     NOMINATED NODE   READINESS GATES
nginx   1/1     Running   0          43s   172.17.1.3   node01   <none>           <none>

controlplane ~ ➜  vi nginx.yaml

controlplane ~ ➜  cat nginx.yaml
---
apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  nodeName: controlplane
  containers:
  -  image: nginx
     name: nginx


controlplane ~ ➜  kubectl replace --force -f nginx.yaml
pod "nginx" deleted
pod/nginx replaced

controlplane ~ ➜  kubectl get pods -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP           NODE           NOMINATED NODE   READINESS GATES
nginx   1/1     Running   0          8s    172.17.0.4   controlplane   <none>           <none>

controlplane ~ ➜  echo "Now schedule the same pod on the controlplane node."
Now schedule the same pod on the controlplane node.

Powered by Moshow@https://zhengkai.blog.youkuaiyun.com/