【k8s资源调度-Deployment】
无状态服务器资源调度
·
1、标签和选择器
1.1 标签Label
- 配置文件：在各类资源的sepc.metadata.label 中进行配置
- 通过kubectl 命令行创建修改标签，语法如下
- 创建临时label：kubectl label po <资源名称> app=hello -n <命令空间（可不加）>
- 修改已经存在的label： kubectl label po <资源名称> app=hello -n <命令空间（可不加）> --overwrite
- 通过查看lable的时候修改label
- selector 按照 label 单值查找节点： kubectl get po -A -l app=hello
- 查看所有节点的labels ：kubectl get po --show-labels
## 创建pod
[root@k8s-master ~]# kubectl create -f nginx-po.yml
pod/nginx-liveness-po created
# 查看pod状态
[root@k8s-master ~]# kubectl get po nginx-liveness-po
NAME READY STATUS RESTARTS AGE
nginx-liveness-po 1/1 Running 0 51s
# 查看pod状态，显示label信息
[root@k8s-master ~]# kubectl get po nginx-liveness-po --show-labels
NAME READY STATUS RESTARTS AGE LABELS
nginx-liveness-po 1/1 Running 0 68s test=l.0.0,type=app
# 临时给pod添加label
[root@k8s-master ~]# kubectl label po nginx-liveness-po auth=xiaobai
pod/nginx-liveness-po labeled
[root@k8s-master ~]# kubectl get po nginx-liveness-po --show-labels
NAME READY STATUS RESTARTS AGE LABELS
nginx-liveness-po 1/1 Running 0 105s auth=xiaobai,test=l.0.0,type=app
# 修改pod的label，但是不加 --overwrite 会提示失败，需要加 --overweite
[root@k8s-master ~]# kubectl label po nginx-liveness-po auth=xiaohong
error: 'auth' already has a value (xiaobai), and --overwrite is false
[root@k8s-master ~]# kubectl label po nginx-liveness-po auth=xiaohong --overwrite
pod/nginx-liveness-po labeled
# 修改label属性已经更改
[root@k8s-master ~]# kubectl get po nginx-liveness-po --show-labels
NAME READY STATUS RESTARTS AGE LABELS
nginx-liveness-po 1/1 Running 0 2m3s auth=xiaohong,test=l.0.0,type=app
# 通过单个labe信息查找pod
[root@k8s-master ~]# kubectl get po -A -l type=app
NAMESPACE NAME READY STATUS RESTARTS AGE
default nginx-liveness-po 1/1 Running 0 8m27s
1.2 选择器Seletor
- 通过yaml配置文件，在各对象的配置 spec.seletor 或其他可以写selector的属性中编写。
- 通过命令行模式，语法如下
- 匹配但个值，查找app=hello的pod：kubectl get po -A -I app=hello
- 匹配多个值：kubectl get po -A -I ‘k8s-app in (metrics-server,kubernetes-dashboard)’
- 查找version!=1 and app=nginx的pod信息：kubectl get po -I version!=1,app=nginx
# 单值匹配的时候
[root@k8s-master ~]# kubectl get po -A -l type=app --show-labels
NAMESPACE NAME READY STATUS RESTARTS AGE LABELS
default nginx-liveness-po 1/1 Running 0 18m auth=xiaohong,test=l.0.0,type=app
# 显示所有的pod的label信息
[root@k8s-master ~]# kubectl get po -A --show-labels
NAMESPACE NAME READY STATUS RESTARTS AGE LABELS
default nginx-liveness-po 1/1 Running 0 18m auth=xiaohong,test=l.0.0,type=app
kube-flannel kube-flannel-ds-glkkb 1/1 Running 4 (24h ago) 3d19h app=flannel,controller-revision-hash=7cfb6d964b,pod-template-generation=1,tier=node
kube-flannel kube-flannel-ds-pdmtw 1/1 Running 1 (24h ago) 44h app=flannel,controller-revision-hash=7cfb6d964b,pod-template-generation=1,tier=node
kube-flannel kube-flannel-ds-tpm8x 1/1 Running 2 (24h ago) 3d19h app=flannel,controller-revision-hash=7cfb6d964b,pod-template-generation=1,tier=node
kube-system coredns-c676cc86f-pdsl6 1/1 Running 3 (24h ago) 2d7h k8s-app=kube-dns,pod-template-hash=c676cc86f
kube-system coredns-c676cc86f-q7hcw 1/1 Running 1 (24h ago) 2d7h k8s-app=kube-dns,pod-template-hash=c676cc86f
kube-system etcd-k8s-master 1/1 Running 2 (24h ago) 3d20h component=etcd,tier=control-plane
kube-system kube-apiserver-k8s-master 1/1 Running 3 (24h ago) 3d20h component=kube-apiserver,tier=control-plane
kube-system kube-controller-manager-k8s-master 1/1 Running 4 (24h ago) 3d20h component=kube-controller-manager,tier=control-plane
kube-system kube-proxy-n2w92 1/1 Running 4 (24h ago) 3d19h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-proxy-p8fhs 1/1 Running 1 (24h ago) 44h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-proxy-xtllb 1/1 Running 2 (24h ago) 3d20h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-scheduler-k8s-master 1/1 Running 4 (24h ago) 3d20h component=kube-scheduler,tier=control-plane
# 匹配单个值 app!=flannel
[root@k8s-master ~]# kubectl get po -A -l app!=flannel
NAMESPACE NAME READY STATUS RESTARTS AGE
default nginx-liveness-po 1/1 Running 0 19m
kube-system coredns-c676cc86f-pdsl6 1/1 Running 3 (24h ago) 2d7h
kube-system coredns-c676cc86f-q7hcw 1/1 Running 1 (24h ago) 2d7h
kube-system etcd-k8s-master 1/1 Running 2 (24h ago) 3d20h
kube-system kube-apiserver-k8s-master 1/1 Running 3 (24h ago) 3d20h
kube-system kube-controller-manager-k8s-master 1/1 Running 4 (24h ago) 3d20h
kube-system kube-proxy-n2w92 1/1 Running 4 (24h ago) 3d19h
kube-system kube-proxy-p8fhs 1/1 Running 1 (24h ago) 44h
kube-system kube-proxy-xtllb 1/1 Running 2 (24h ago) 3d20h
kube-system kube-scheduler-k8s-master 1/1 Running 4 (24h ago) 3d20h
[root@k8s-master ~]# kubectl get po -A -l app!=flannel --show-labels
NAMESPACE NAME READY STATUS RESTARTS AGE LABELS
default nginx-liveness-po 1/1 Running 0 19m auth=xiaohong,test=l.0.0,type=app
kube-system coredns-c676cc86f-pdsl6 1/1 Running 3 (24h ago) 2d7h k8s-app=kube-dns,pod-template-hash=c676cc86f
kube-system coredns-c676cc86f-q7hcw 1/1 Running 1 (24h ago) 2d7h k8s-app=kube-dns,pod-template-hash=c676cc86f
kube-system etcd-k8s-master 1/1 Running 2 (24h ago) 3d20h component=etcd,tier=control-plane
kube-system kube-apiserver-k8s-master 1/1 Running 3 (24h ago) 3d20h component=kube-apiserver,tier=control-plane
kube-system kube-controller-manager-k8s-master 1/1 Running 4 (24h ago) 3d20h component=kube-controller-manager,tier=control-plane
kube-system kube-proxy-n2w92 1/1 Running 4 (24h ago) 3d19h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-proxy-p8fhs 1/1 Running 1 (24h ago) 44h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-proxy-xtllb 1/1 Running 2 (24h ago) 3d20h controller-revision-hash=dd4c999cf,k8s-app=kube-proxy,pod-template-generation=1
kube-system kube-scheduler-k8s-master 1/1 Running 4 (24h ago) 3d20h component=kube-scheduler,tier=control-plane
# 多值匹配
[root@k8s-master ~]# kubectl get po -A -l app!=flannel,test=1.0.0 --show-labels
No resources found
# 多值匹配，多值匹配是“与”的关系，不是“或”的关系
[root@k8s-master ~]# kubectl get po -A -l app!=flannel,type=app --show-labels
NAMESPACE NAME READY STATUS RESTARTS AGE LABELS
default nginx-liveness-po 1/1 Running 0 20m auth=xiaohong,test=l.0.0,type=app
2、部署对象Deployment（无状态应用）
2.1 创建deploy
2.1.1 通过命令行创建一个deployment
# 创建一个deployment
[root@k8s-master ~]# kubectl create deploy nginx-deploy --image=nginx:1.20
deployment.apps/nginx-deploy created
创建一个deployment
使用命令行模式：kubectl create deploy nginx-deploy --image=nginx:1.20
或执行 kubectl create -f xxx.yaml --record
-record会在annotation中记录当前命令创建或升级了资源，后续可以查看做过哪变动操作。
2.1.2 查看deployment 信息
通过kubectl get 查看deployment，可以使用deployment，也可以用deploy
# 通过kubectl get 查看deployment，可以使用deployment，也可以用deploy
[root@k8s-master ~]# kubectl get deployments
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 1/1 1 1 10s
[root@k8s-master ~]# kubectl get deploy
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 1/1 1 1 14s
2.1.3 查看replicasets 信息
由于创建的deployment资源包含 replicasets信息（replicasets是可以自动扩容和缩容），可以查看到 replicasets的名字中包含deploy资源的名字信息
# 由于deployment包含 replicasets（replicasets是可以自动扩容和缩容），可以查看到 replicasets的名字中包含deploy的名字信息
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-86b7d8c46d 1 1 1 18s
[root@k8s-master ~]# kubectl get replicasets.apps
NAME DESIRED CURRENT READY AGE
nginx-deploy-86b7d8c46d 1 1 1 38s
2.1.4 查看pod 信息
可以查看到deployment创建的pod种包含 replicasets 资源的名字信息
# 下面的pod可以查看到pod的名字包含 replicasets 名字的信息
[root@k8s-master ~]# kubectl get po
NAME READY STATUS RESTARTS AGE
nginx-deploy-86b7d8c46d-78rj9 1/1 Running 0 46s
2.1.5 通过命令行模式创建的deployment 生成 yaml信息
# 通过创建好的deploy生成yaml文件
[root@k8s-master ~]# kubectl get deploy nginx-deploy -o yaml
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
creationTimestamp: "2024-02-23T10:59:04Z"
generation: 1
labels:
app: nginx-deploy
name: nginx-deploy
namespace: default
resourceVersion: "235341"
uid: def47aae-13f7-415a-a9d1-18ef72e5a925
spec:
progressDeadlineSeconds: 600
replicas: 1
revisionHistoryLimit: 10
selector:
matchLabels:
app: nginx-deploy
strategy:
rollingUpdate:
maxSurge: 25%
maxUnavailable: 25%
type: RollingUpdate
template:
metadata:
creationTimestamp: null
labels:
app: nginx-deploy
spec:
containers:
- image: nginx:1.20
imagePullPolicy: IfNotPresent
name: nginx
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
dnsPolicy: ClusterFirst
restartPolicy: Always
schedulerName: default-scheduler
securityContext: {}
terminationGracePeriodSeconds: 30
status:
availableReplicas: 1
conditions:
- lastTransitionTime: "2024-02-23T10:59:05Z"
lastUpdateTime: "2024-02-23T10:59:05Z"
message: Deployment has minimum availability.
reason: MinimumReplicasAvailable
status: "True"
type: Available
- lastTransitionTime: "2024-02-23T10:59:04Z"
lastUpdateTime: "2024-02-23T10:59:05Z"
message: ReplicaSet "nginx-deploy-86b7d8c46d" has successfully progressed.
reason: NewReplicaSetAvailable
status: "True"
type: Progressing
observedGeneration: 1
readyReplicas: 1
replicas: 1
updatedReplicas: 1
2.1.6 查看po、rs、deploy中所有的label信息
通过以下信息，可以查看到：
deploy、rs、pod种都包含同一个label信息：app=nginx-deploy；
rs和pod中同时还有一个相同的pod标签信息：pod-template-hash=86b7d8c46d，这个标签信息就是为了我们能更方便的动态创建pod使用。
[root@k8s-master ~]# kubectl get po,rs,deploy --show-labels
NAME READY STATUS RESTARTS AGE LABELS
pod/nginx-deploy-86b7d8c46d-78rj9 1/1 Running 0 30m app=nginx-deploy,pod-template-hash=86b7d8c46d
NAME DESIRED CURRENT READY AGE LABELS
replicaset.apps/nginx-deploy-86b7d8c46d 1 1 1 30m app=nginx-deploy,pod-template-hash=86b7d8c46d
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
deployment.apps/nginx-deploy 1/1 1 1 30m app=nginx-deploy
2.1.7 通过（2.1.5）中生成的yaml信息创建yaml文件
通过命令行创建的deployment信息生成yaml信息，根据我们所需信息来创建yaml文件
apiVersion: apps/v1 # deployment api版本
kind: Deployment # 资源类型为 Deployment
metadata: # 原信息
labels: # 标签
app: nginx-deploy # 具体的标签信息：app=nginx-deploy key: value 配置形式
name: nginx-deploy # deployment的名字
namespace: default # 所在的命名空间
spec:
replicas: 1 # 期望副本数
revisionHistoryLimit: 10 # 进行滚动更新后，保留的历史版本数量
selector: # 选择器，用于找到匹配的RS
matchLabels: # 按照标签匹配
app: nginx-deploy # 匹配的标签
strategy: #更新策略
rollingUpdate: # 滚动更新配置
maxSurge: 25% # 滚动更新时，更新的个数最多超过多少个期望副本数，或者比例
maxUnavailable: 25% # 进行滚动更新时，最大不可用比例更新比例，表示在所有副本数中，最多可以有多少个不更新成功
type: RollingUpdate # 更新类型，采用滚动更新
template: # pod 模板
metadata: # pod模板的元信息
labels: # pod模板的标签
app: nginx-deploy # pod模板的标签信息
spec: # pod 期望信息
containers: # pod 的容器信息
- image: nginx:1.20 # 镜像信息
imagePullPolicy: IfNotPresent # 镜像拉取策略
name: nginx # 容器名字
restartPolicy: Always # pod的重启策略
terminationGracePeriodSeconds: 30 # pod的过期时间
2.2 滚动更新
只有修改了deployment 配置文件文件中的template中的属性后，才会触发滚动更新操作
2.2.1 修改deploy中的其他属性查看pod是否会更新
[root@k8s-master ~]# kubectl get deployments --show-labels
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
nginx-deploy 1/1 1 1 44m app=nginx-deploy
[root@k8s-master ~]# kubectl edit deployments nginx-deploy
deployment.apps/nginx-deploy edited
可以通过下面新的看到我们的deploy更新并没有导致pod更新，只是把新加的标签给更新到了deploy上。
[root@k8s-master ~]# kubectl get deployments --show-labels
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
nginx-deploy 1/1 1 1 49m app=nginx-deploy,new-version=test
# rs和pod并没有这个标签信息
[root@k8s-master ~]# kubectl get deployments,rs,pod --show-labels
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
deployment.apps/nginx-deploy 1/1 1 1 51m app=nginx-deploy,new-version=test
NAME DESIRED CURRENT READY AGE LABELS
replicaset.apps/nginx-deploy-86b7d8c46d 1 1 1 51m app=nginx-deploy,pod-template-hash=86b7d8c46d
NAME READY STATUS RESTARTS AGE LABELS
pod/nginx-deploy-86b7d8c46d-78rj9 1/1 Running 0 51m app=nginx-deploy,pod-template-hash=86b7d8c46d
2.2.2 修改这个deploy的副本数
[root@k8s-master ~]# kubectl get deploy nginx-deploy --show-labels
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
nginx-deploy 3/3 3 3 55m app=nginx-deploy,new-version=test
[root@k8s-master ~]# kubectl describe deployments nginx-deploy
Name: nginx-deploy
Namespace: default
CreationTimestamp: Fri, 23 Feb 2024 18:59:04 +0800
Labels: app=nginx-deploy
new-version=test
Annotations: deployment.kubernetes.io/revision: 1
Selector: app=nginx-deploy
Replicas: 3 desired | 3 updated | 3 total | 3 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 25% max unavailable, 25% max surge
Pod Template:
Labels: app=nginx-deploy
Containers:
nginx:
Image: nginx:1.20
Port: <none>
Host Port: <none>
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Progressing True NewReplicaSetAvailable
Available True MinimumReplicasAvailable
OldReplicaSets: <none>
NewReplicaSet: nginx-deploy-86b7d8c46d (3/3 replicas created)
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal ScalingReplicaSet 55m deployment-controller Scaled up replica set nginx-deploy-86b7d8c46d to 1
Normal ScalingReplicaSet 27s deployment-controller Scaled up replica set nginx-deploy-86b7d8c46d to 3 from 1
[root@k8s-master ~]#
通过增加副本数后，我们可以看到deploy变为了3，rs为1，pod为3，pod的模版使用的都是同一个模版。
[root@k8s-master ~]# kubectl get deployments,rs,po --show-labels
NAME READY UP-TO-DATE AVAILABLE AGE LABELS
deployment.apps/nginx-deploy 3/3 3 3 57m app=nginx-deploy,new-version=test
NAME DESIRED CURRENT READY AGE LABELS
replicaset.apps/nginx-deploy-86b7d8c46d 3 3 3 57m app=nginx-deploy,pod-template-hash=86b7d8c46d
NAME READY STATUS RESTARTS AGE LABELS
pod/nginx-deploy-86b7d8c46d-6cf95 1/1 Running 0 93s app=nginx-deploy,pod-template-hash=86b7d8c46d
pod/nginx-deploy-86b7d8c46d-78rj9 1/1 Running 0 57m app=nginx-deploy,pod-template-hash=86b7d8c46d
pod/nginx-deploy-86b7d8c46d-vjncw 1/1 Running 0 93s app=nginx-deploy,pod-template-hash=86b7d8c46d
2.2.3 通过edit选项编辑template属性后我们看看会发生那些变化
通过实时监控deploy的时候可以看到deploy的变化
[root@k8s-master ~]# kubectl get deployments.apps nginx-deploy -w
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 3/3 3 3 61m
nginx-deploy 3/3 3 3 61m
nginx-deploy 3/3 3 3 61m
nginx-deploy 3/3 0 3 61m
nginx-deploy 3/3 1 3 61m
nginx-deploy 4/3 1 4 62m
nginx-deploy 3/3 1 3 62m
nginx-deploy 3/3 2 3 62m
nginx-deploy 4/3 2 4 62m
nginx-deploy 3/3 2 3 62m
nginx-deploy 3/3 3 3 62m
nginx-deploy 4/3 3 4 62m
nginx-deploy 3/3 3 3 62m
# 通过这个命令可以看到更新的状态为成功了
[root@k8s-master ~]# kubectl rollout status deployment nginx-deploy
deployment "nginx-deploy" successfully rolled out
2.2.4 修改set选项编辑刚才的镜像信息看看更新如何
[root@k8s-master ~]# kubectl set image deployment/nginx-deploy nginx=nginx:1.20
deployment.apps/nginx-deploy image updated
[root@k8s-master ~]# kubectl get deployments.apps nginx-deploy -w
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 3/3 3 3 72m
nginx-deploy 3/3 3 3 72m
nginx-deploy 3/3 3 3 72m
nginx-deploy 3/3 0 3 72m
nginx-deploy 3/3 1 3 72m
nginx-deploy 4/3 1 4 72m
nginx-deploy 3/3 1 3 72m
nginx-deploy 3/3 2 3 72m
nginx-deploy 4/3 2 4 72m
nginx-deploy 3/3 2 3 72m
nginx-deploy 3/3 3 3 72m
nginx-deploy 4/3 3 4 72m
nginx-deploy 3/3 3 3 72m
通过如下信息，可以看到新的nginx pod信息可以看到他滚动更新到刚开始的rs信息
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-6bf65c4445 0 0 0 17m
nginx-deploy-86b7d8c46d 3 3 3 79m
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-deploy-86b7d8c46d-4s98c 1/1 Running 0 7m17s
nginx-deploy-86b7d8c46d-7s9gz 1/1 Running 0 7m19s
nginx-deploy-86b7d8c46d-w8ljk 1/1 Running 0 7m15s
2.3 回滚操作
有时候你可能想回退一个Deployment，例如，当Deployment不稳定时，比如一直crash looping。
默认侍况下，kubernetes会在系统中保存前两次的Deployment的rollout历史记录，以便你可以随时会退（你可以修改revisionhistory limit来更改保存的revision数)，
2.3.1 通过set选项修改一个错误的images信息
[root@k8s-master ~]# kubectl set image deployment/nginx-deploy nginx=nginx:1.200
deployment.apps/nginx-deploy image updated
# 可以看到一个更新的这个nginx的状态是ImagePullBackOff
[root@k8s-master ~]# kubectl get pod -w
NAME READY STATUS RESTARTS AGE
nginx-deploy-85bc5c8fdb-h4rmn 0/1 ImagePullBackOff 0 83s
nginx-deploy-86b7d8c46d-4s98c 1/1 Running 0 25m
nginx-deploy-86b7d8c46d-7s9gz 1/1 Running 0 25m
nginx-deploy-86b7d8c46d-w8ljk 1/1 Running 0 25m
使用命令：kubectl describe pod nginx-deploy-85bc5c8fdb-h4rmn 可以查看到这个新的pod信息
2.3.2 通过edit选项查看这个deploy的信息
2.3.3 查看历史的版本信息
[root@k8s-master ~]# kubectl rollout history deployment/nginx-deploy --revision=0
deployment.apps/nginx-deploy
REVISION CHANGE-CAUSE
2 <none>
3 <none>
4 <none>
2.3.4 查看历史更新的具体信息
[root@k8s-master ~]# kubectl rollout history deployment/nginx-deploy --revision=4
deployment.apps/nginx-deploy with revision #4
Pod Template:
Labels: app=nginx-deploy
pod-template-hash=85bc5c8fdb
Containers:
nginx:
Image: nginx:1.200
Port: <none>
Host Port: <none>
Environment: <none>
Mounts: <none>
Volumes: <none>
[root@k8s-master ~]# kubectl rollout history deployment/nginx-deploy --revision=3
deployment.apps/nginx-deploy with revision #3
Pod Template:
Labels: app=nginx-deploy
pod-template-hash=86b7d8c46d
Containers:
nginx:
Image: nginx:1.20
Port: <none>
Host Port: <none>
Environment: <none>
Mounts: <none>
Volumes: <none>
[root@k8s-master ~]# kubectl rollout history deployment/nginx-deploy --revision=2
deployment.apps/nginx-deploy with revision #2
Pod Template:
Labels: app=nginx-deploy
pod-template-hash=6bf65c4445
Containers:
nginx:
Image: nginx:1.21
Port: <none>
Host Port: <none>
Environment: <none>
Mounts: <none>
Volumes: <none>
2.3.5 回滚为revision=3的版本
[root@k8s-master ~]# kubectl rollout undo deployment/nginx-deploy --to-revision=3
deployment.apps/nginx-deploy rolled back
[root@k8s-master ~]# kubectl rollout status deployments.apps nginx-deploy
deployment "nginx-deploy" successfully rolled out
如下已经回滚为了revision=3的版本。
[root@k8s-master ~]# kubectl describe deployments nginx-deploy
Name: nginx-deploy
Namespace: default
CreationTimestamp: Fri, 23 Feb 2024 18:59:04 +0800
Labels: app=nginx-deploy
new-version=test
Annotations: deployment.kubernetes.io/revision: 5
Selector: app=nginx-deploy
Replicas: 3 desired | 3 updated | 3 total | 3 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 25% max unavailable, 25% max surge
Pod Template:
Labels: app=nginx-deploy
Containers:
nginx:
Image: nginx:1.20
Port: <none>
Host Port: <none>
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Available True MinimumReplicasAvailable
Progressing True NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet: nginx-deploy-86b7d8c46d (3/3 replicas created)
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal ScalingReplicaSet 57m deployment-controller Scaled up replica set nginx-deploy-86b7d8c46d to 3 from 1
Normal ScalingReplicaSet 51m deployment-controller Scaled up replica set nginx-deploy-6bf65c4445 to 1
Normal ScalingReplicaSet 50m deployment-controller Scaled down replica set nginx-deploy-86b7d8c46d to 2 from 3
Normal ScalingReplicaSet 50m deployment-controller Scaled up replica set nginx-deploy-6bf65c4445 to 2 from 1
Normal ScalingReplicaSet 49m deployment-controller Scaled down replica set nginx-deploy-86b7d8c46d to 1 from 2
Normal ScalingReplicaSet 49m deployment-controller Scaled up replica set nginx-deploy-6bf65c4445 to 3 from 2
Normal ScalingReplicaSet 49m deployment-controller Scaled down replica set nginx-deploy-86b7d8c46d to 0 from 1
Normal ScalingReplicaSet 40m deployment-controller Scaled up replica set nginx-deploy-86b7d8c46d to 1 from 0
Normal ScalingReplicaSet 40m deployment-controller Scaled down replica set nginx-deploy-6bf65c4445 to 2 from 3
Normal ScalingReplicaSet 40m (x4 over 40m) deployment-controller (combined from similar events): Scaled down replica set nginx-deploy-6bf65c4445 to 0 from 1
Normal ScalingReplicaSet 16m deployment-controller Scaled up replica set nginx-deploy-85bc5c8fdb to 1
Normal ScalingReplicaSet 116s deployment-controller Scaled down replica set nginx-deploy-85bc5c8fdb to 0 from 1
2.3.6 为啥可以回退为revision=3呢？
可以通过设置，spec.revisonHistoryLimit来指定deployment保留多少revision。如果revisonHistoryLimit设置为0，则不允许deployment回退了。
2.4 扩容和缩容
- 扩缩容命令： kubectl scale --replicas=6 deployment nginx-deploy
- 扩容和缩容一样，只需要通过–replicas 指定数量即可
2.4.1 扩容
# 目前有3个pod，扩容到6个
[root@k8s-master ~]# kubectl scale --replicas=6 deployment nginx-deploy
deployment.apps/nginx-deploy scaled
[root@k8s-master ~]# kubectl get deployments nginx-deploy
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 6/6 6 6 5h43m
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-6bf65c4445 0 0 0 4h42m
nginx-deploy-85bc5c8fdb 0 0 0 4h7m
nginx-deploy-86b7d8c46d 6 6 6 5h43m
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-deploy-86b7d8c46d-4s98c 1/1 Running 0 4h31m
nginx-deploy-86b7d8c46d-7s9gz 1/1 Running 0 4h31m
nginx-deploy-86b7d8c46d-n2m6j 1/1 Running 0 33s
nginx-deploy-86b7d8c46d-w8ljk 1/1 Running 0 4h31m
nginx-deploy-86b7d8c46d-wgdnv 1/1 Running 0 33s
nginx-deploy-86b7d8c46d-z86rx 1/1 Running 0 33s
2.4.1 缩容
# 缩容，有6个pod，现在缩容到4个pod
[root@k8s-master ~]# kubectl scale --replicas=4 deployment nginx-deploy
deployment.apps/nginx-deploy scaled
[root@k8s-master ~]# kubectl get deployments nginx-deploy
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deploy 4/4 4 4 5h45m
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-6bf65c4445 0 0 0 4h44m
nginx-deploy-85bc5c8fdb 0 0 0 4h9m
nginx-deploy-86b7d8c46d 4 4 4 5h45m
[root@k8s-master ~]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-deploy-86b7d8c46d-4s98c 1/1 Running 0 4h33m 10.2.2.12 k8s-node-01 <none> <none>
nginx-deploy-86b7d8c46d-7s9gz 1/1 Running 0 4h33m 10.2.1.28 k8s-node-02 <none> <none>
nginx-deploy-86b7d8c46d-w8ljk 1/1 Running 0 4h33m 10.2.1.29 k8s-node-02 <none> <none>
nginx-deploy-86b7d8c46d-z86rx 1/1 Running 0 2m33s 10.2.2.14 k8s-node-01 <none> <none>
2.5 暂停和恢复
由于每次对pod 的 template中的信息发生修改后，都会触发更新deployment操作，那么此时如果频繁修改信息，就会产生多次更新，而实际上只需要执行最后一次更新即可，当出现此类情况时我们就可以暂停deployment的 rollout。
2.5.1 给template添加一个最小cpu和内存的参数
2.5.2 暂停deploy更新
通过kubectl rollout pause deployment <name> 就可以实现暂停，直到你下次恢复后才会继续进行滚动更新。
# 暂时deploy更新
[root@k8s-master ~]# kubectl rollout pause deployment nginx-deploy
deployment.apps/nginx-deploy paused
2.5.3 再次更新template，给cpu和内存添加一个最大参数
这次更新过之后，rs信息和pod信息都没有发生改变，是由于我们暂停了deploy的更新
[root@k8s-master ~]# kubectl edit deploy nginx-deploy
deployment.apps/nginx-deploy edited
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-6bf65c4445 0 0 0 5h7m
nginx-deploy-6dc7697cfb 3 3 3 8m33s
nginx-deploy-85bc5c8fdb 0 0 0 4h32m
nginx-deploy-86b7d8c46d 0 0 0 6h8m
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-deploy-6dc7697cfb-6p9mj 1/1 Running 0 8m38s
nginx-deploy-6dc7697cfb-lc472 1/1 Running 0 8m37s
nginx-deploy-6dc7697cfb-w295s 1/1 Running 0 8m35s
2.5.4 查看更新信息
通过下面的信息，可以看到刚才添加cpu和内存参数在更新历史列表里面没有显示，说明我们的deploy没有更新
[root@k8s-master ~]# kubectl rollout history deployment nginx-deploy --revision=0
deployment.apps/nginx-deploy
REVISION CHANGE-CAUSE
2 <none>
4 <none>
5 <none>
6 <none>
[root@k8s-master ~]# kubectl rollout history deployment nginx-deploy --revision=6
deployment.apps/nginx-deploy with revision #6
Pod Template:
Labels: app=nginx-deploy
pod-template-hash=6dc7697cfb
Containers:
nginx:
Image: nginx:1.20
Port: <none>
Host Port: <none>
Requests:
cpu: 100m
memory: 128Mi
Environment: <none>
Mounts: <none>
Volumes: <none>
2.5.5 恢复deploy的更新
启动deploy的更新
[root@k8s-master ~]# kubectl rollout resume deployment nginx-deploy
deployment.apps/nginx-deploy resumed
这里 rs 已经发生了改变
[root@k8s-master ~]# kubectl get rs
NAME DESIRED CURRENT READY AGE
nginx-deploy-57bf686b9d 3 3 3 12s
nginx-deploy-6bf65c4445 0 0 0 5h15m
nginx-deploy-6dc7697cfb 0 0 0 17m
nginx-deploy-85bc5c8fdb 0 0 0 4h41m
nginx-deploy-86b7d8c46d 0 0 0 6h17m
刚才查询的历史更新版本只有4个，现在又多了一个
[root@k8s-master ~]# kubectl rollout history deployment nginx-deploy --revision=0
deployment.apps/nginx-deploy
REVISION CHANGE-CAUSE
2 <none>
4 <none>
5 <none>
6 <none>
7 <none>
通过历史更新版本信息可以看到，我们添加的最大cpu和内存参数都发生了改变。
[root@k8s-master ~]# kubectl rollout history deployment nginx-deploy --revision=7
deployment.apps/nginx-deploy with revision #7
Pod Template:
Labels: app=nginx-deploy
pod-template-hash=57bf686b9d
Containers:
nginx:
Image: nginx:1.20
Port: <none>
Host Port: <none>
Limits:
cpu: 500m
memory: 512Mi
Requests:
cpu: 100m
memory: 128Mi
Environment: <none>
Mounts: <none>
Volumes: <none>
2.6 配置文件（部署无状态服务nginx的yaml文件）
[root@k8s-master ~]# cat nginx-deploy.yaml
apiVersion: apps/v1 # deployment api版本
kind: Deployment # 资源类型为 Deployment
metadata: # 原信息
labels: # 标签
app: nginx-deploy # 具体的标签信息：app=nginx-deploy key: value 配置形式
name: nginx-deploy # deployment的名字
namespace: default # 所在的命名空间
spec:
replicas: 1 # 期望副本数
revisionHistoryLimit: 10 # 进行滚动更新后，保留的历史版本数量
selector: # 选择器，用于找到匹配的RS
matchLabels: # 按照标签匹配
app: nginx-deploy # 匹配的标签
strategy: #更新策略
rollingUpdate: # 滚动更新配置
maxSurge: 25% # 滚动更新时，更新的个数最多超过多少个期望副本数，或者比例
maxUnavailable: 25% # 进行滚动更新时，最大不可用比例更新比例，表示在所有副本数中，最多可以有多少个不更新成功
type: RollingUpdate # 更新类型，采用滚动更新
template: # pod 模板
metadata: # pod模板的元信息
labels: # pod模板的标签
app: nginx-deploy # pod模板的标签信息
spec: # pod 期望信息
containers: # pod 的容器信息
- image: nginx:1.20 # 镜像信息
imagePullPolicy: IfNotPresent # 镜像拉取策略
name: nginx # 容器名字
restartPolicy: Always # pod的重启策略
terminationGracePeriodSeconds: 30 # pod的过期时间
K8S/Kubernetes社区为您提供最前沿的新闻资讯和知识内容
更多推荐
24
0- 0
已为社区贡献18条内容
所有评论(0)