1、Pod控制器
1.1 介绍
Pod控制器是用于实现管理pod的中间层,确保pod资源符合预期的状态,pod的资源出现故障时,会尝试 进行重启,当根据重启策略无效,则会重新新建pod的资源。
1.2 pod控制器有多种类型
- ReplicationController(RC):RC保证了在所有时间内,都有特定数量的Pod副本正在运行,如果太多了,RC就杀死几个,如果太少了,RC会新建几个
- ReplicaSet(RS):代用户创建指定数量的pod副本数量,确保pod副本数量符合预期状态,并且支持滚动式自动扩容和缩容功能。
- Deployment(重要):工作在ReplicaSet之上,用于管理无状态应用,目前来说最好的控制器。支持滚动更新和回滚功能,还提供声明式配置。
- DaemonSet:用于确保集群中的每一个节点只运行特定的pod副本,通常用于实现系统级后台任务。比如ELK服务
- Job:只要完成就立即退出,不需要重启或重建。
- CronJob:周期性任务控制,不需要持续后台运行
- StatefulSet:管理有状态应用
本文主要讲解ReplicaSet、Deployment、DaemonSet 三中类型的pod控制器。
2、ReplicaSet
2.1 认识ReplicaSet
(1)什么是ReplicaSet?
ReplicaSet是下一代复本控制器,是Replication Controller(RC)的升级版本。ReplicaSet和 Replication Controller之间的唯一区别是对选择器的支持。ReplicaSet支持labels user guide中描述的set-based选择器要求, 而Replication Controller仅支持equality-based的选择器要求。
(2)如何使用ReplicaSet
大多数kubectl 支持Replication Controller 命令的也支持ReplicaSets。rolling-update命令除外,如果要使用rolling-update,请使用Deployments来实现。
虽然ReplicaSets可以独立使用,但它主要被 Deployments用作pod 机制的创建、删除和更新。当使用Deployment时,你不必担心创建pod的ReplicaSets,因为可以通过Deployment实现管理ReplicaSets。
(3)何时使用ReplicaSet?
ReplicaSet能确保运行指定数量的pod。然而,Deployment 是一个更高层次的概念,它能管理ReplicaSets,并提供对pod的更新等功能。因此,我们建议你使用Deployment来管理ReplicaSets,除非你需要自定义更新编排。
这意味着你可能永远不需要操作ReplicaSet对象,而是使用Deployment替代管理 。后续讲解Deployment会详细描述。
2.2 ReplicaSet定义资源清单几个字段
- apiVersion: app/v1 版本
- kind: ReplicaSet 类型
- metadata 元数据
- spec 期望状态
- minReadySeconds:应为新创建的pod准备好的最小秒数
- replicas:副本数; 默认为1
- selector:标签选择器
- template:模板(必要的)
- metadata:模板中的元数据
- spec:模板中的期望状态
- status 当前状态
2.3 演示:创建一个简单的ReplicaSet
(1)编写yaml文件,并创建启动
简单创建一个replicaset:启动2个pod
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
[root@master manifests] # vim rs-damo.yaml apiVersion: apps /v1 kind: ReplicaSet metadata: name: myapp namespace: default spec: replicas: 2 selector: matchLabels: app: myapp release: canary template: metadata: name: myapp-pod labels: app: myapp release: canary environment: qa spec: containers: - name: myapp-container image: ikubernetes /myapp :v1 ports: - name: http containerPort: 80 [root@master manifests] # kubectl create -f rs-damo.yaml replicaset.apps /myapp created |
(2)查询验证
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
---查询replicaset(rs)信息 [root@master manifests] # kubectl get rs NAME DESIRED CURRENT READY AGE myapp 2 2 2 23s ---查询pod信息 [root@master manifests] # kubectl get pods NAME READY STATUS RESTARTS AGE myapp-r4ss4 1 /1 Running 0 25s myapp-zjc5l 1 /1 Running 0 26s ---查询pod详细信息;模板中的label都生效了 [root@master manifests] # kubectl describe pod myapp-r4ss4 Name: myapp-r4ss4 Namespace: default Priority: 0 PriorityClassName: <none> Node: node2 /192 .168.130.105 Start Time: Thu, 06 Sep 2018 14:57:23 +0800 Labels: app=myapp environment=qa release=canary ... ... ---验证服务 [root@master manifests] # curl 10.244.2.13 Hello MyApp | Version: v1 | <a href= "hostname.html" >Pod Name< /a > |
(3)生成pod原则:“多退少补”
① 删除pod,会立即重新构建,生成新的pod
1
2
3
4
5
6
|
[root@master manifests] # kubectl delete pods myapp-zjc5l pod "myapp-k4j6h" deleted [root@master ~] # kubectl get pods NAME READY STATUS RESTARTS AGE myapp-r4ss4 1 /1 Running 0 33s myapp-mdjvh 1 /1 Running 0 10s |
② 若另一个pod,不小心符合了rs的标签选择器,就会随机干掉一个此标签的pod
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
---随便启动一个pod [root@master manifests] # kubectl get pods --show-labels NAME READY STATUS RESTARTS AGE LABELS myapp-hxgbh 1 /1 Running 0 7m app=myapp,environment=qa,release=canary myapp-mdjvh 1 /1 Running 0 6m app=myapp,environment=qa,release=canary pod- test 1 /1 Running 0 13s app=myapp,tier=frontend ---将pod- test 打上release=canary标签 [root@master manifests] # kubectl label pods pod-test release=canary pod /pod-test labeled ---随机停掉一个pod [root@master manifests] # kubectl get pods --show-labels NAME READY STATUS RESTARTS AGE LABELS myapp-hxgbh 1 /1 Running 0 8m app=myapp,environment=qa,release=canary myapp-mdjvh 1 /1 Running 0 7m app=myapp,environment=qa,release=canary pod- test 0 /1 Terminating 0 1m app=myapp,release=canary,tier=frontend |
2.4 ReplicaSet动态扩容/缩容
(1)使用edit 修改rs 配置,将副本数改为5;即可实现动态扩容
1
2
3
4
5
6
|
[root@master manifests] # kubectl edit rs myapp ... ... spec: replicas: 5 ... ... replicaset.extensions /myapp edited |
(2)验证
1
2
3
4
5
6
7
8
|
[root@master manifests] # kubectl get pods NAME READY STATUS RESTARTS AGE client 0 /1 Error 0 1d myapp-bck7l 1 /1 Running 0 16s myapp-h8cqr 1 /1 Running 0 16s myapp-hfb72 1 /1 Running 0 6m myapp-r4ss4 1 /1 Running 0 9m myapp-vvpgf 1 /1 Running 0 16s |
2.5 ReplicaSet在线升级版本
(1)使用edit 修改rs 配置,将容器的镜像改为v2版;即可实现在线升级版本
1
2
3
4
5
6
7
|
[root@master manifests] # kubectl edit rs myapp ... ... spec: containers: - image: ikubernetes /myapp :v2 ... ... replicaset.extensions /myapp edited |
(2)查询rs,已经完成修改
1
2
3
|
[root@master manifests] # kubectl get rs -o wide NAME DESIRED CURRENT READY AGE CONTAINERS IMAGES SELECTOR myapp 5 5 5 11m myapp-container ikubernetes /myapp :v2 app=myapp,release=canary |
(3)但是,修改完并没有升级
需删除pod,再自动生成新的pod时,就会升级成功;
即可以实现灰度发布:删除一个,会自动启动一个版本升级成功的pod
1
2
3
4
5
6
7
8
9
10
11
|
---访问没有删除pod的服务,显示是V1版 [root@master manifests] # curl 10.244.2.15 Hello MyApp | Version: v1 | <a href= "hostname.html" >Pod Name< /a > ---删除一个pod,访问新生成pod的服务,版本升级为v2版 [root@master manifests] # kubectl delete pod myapp-bck7l pod "myapp-bck7l" deleted [root@master ~] # kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE myapp-hxgbh 1 /1 Running 0 20m 10.244.1.17 node1 [root@master manifests] # curl 10.244.1.17 Hello MyApp | Version: v2 | <a href= "hostname.html" >Pod Name< /a > |
3、Deployment
3.1 Deployment简述
(1)介绍
Deployment 为 Pod和Replica Set 提供了一个声明式定义(declarative)方法,用来替代以前的ReplicationController来方便的管理应用
你只需要在 Deployment 中描述您想要的目标状态是什么,Deployment controller 就会帮您将 Pod 和ReplicaSet 的实际状态改变到您的目标状态。您可以定义一个全新的 Deployment 来创建 ReplicaSet 或者删除已有的 Deployment 并创建一个新的来替换。
注意:您不该手动管理由 Deployment 创建的 Replica Set,否则您就篡越了 Deployment controller 的职责!
(2)典型的应用场景包括
- 使用Deployment来创建ReplicaSet。ReplicaSet在后台创建pod。检查启动状态,看它是成功还是失败。
- 然后,通过更新Deployment 的 PodTemplateSpec 字段来声明Pod的新状态。这会创建一个新的ReplicaSet,Deployment会按照控制的速率将pod从旧的ReplicaSet移动到新的ReplicaSet中。
- 滚动升级和回滚应用:如果当前状态不稳定,回滚到之前的Deployment revision。每次回滚都会更新Deployment的revision。
- 扩容和缩容:扩容Deployment以满足更高的负载。
- 暂停和继续Deployment:暂停Deployment来应用PodTemplateSpec的多个修复,然后恢复上线。
- 根据Deployment 的状态判断上线是否hang住了。
- 清除旧的不必要的 ReplicaSet。
3.2 Deployment定义资源清单几个字段
- apiVersion: app/v1 版本
- kind: Deployment 类型
- metadata 元数据
- spec 期望状态
- --------------replicaset 也有的选项---------------
- minReadySeconds:应为新创建的pod准备好的最小秒数
- replicas:副本数; 默认为1
- selector:标签选择器
- template:模板(必须的)
- metadata:模板中的元数据
- spec:模板中的期望状态
- --------------deployment 独有的选项---------------
- strategy:更新策略;用于将现有pod替换为新pod的部署策略
- Recreate:重新创建
- RollingUpdate:滚动更新
- maxSurge:可以在所需数量的pod之上安排的最大pod数;例如:5、10%
- maxUnavailable:更新期间可用的最大pod数;
- revisionHistoryLimit:要保留以允许回滚的旧ReplicaSet的数量,默认10
- paused:表示部署已暂停,部署控制器不处理部署
- progressDeadlineSeconds:在执行此操作之前部署的最长时间被认为是失败的
- status 当前状态
3.3 演示:创建一个简单的Deployment
(1)创建一个简单的ReplicaSet,启动2个pod
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
[root@master manifests] # vim deploy-damo.yaml apiVersion: apps /v1 kind: Deployment metadata: name: myapp-deploy namespace: default spec: replicas: 2 selector: matchLabels: app: myapp release: canary template: metadata: labels: app: myapp release: canary spec: containers: - name: myapp image: ikubernetes /myapp :v1 ports: - name: http containerPort: 80 [root@master manifests] # kubectl apply -f deploy-damo.yaml deployment.apps /myapp-deploy configured |
注:apply 声明式创建启动;和create差不多;但是可以对一个文件重复操作;create不可以。
(2)查询验证
1
2
3
4
5
6
7
8
9
10
11
12
13
|
---查询deployment信息 [root@master manifests] # kubectl get deploy NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE myapp-deploy 2 2 2 2 14s ---查询replicaset信息;deployment会先生成replicaset [root@master manifests] # kubectl get rs NAME DESIRED CURRENT READY AGE myapp-deploy-69b47bc96d 2 2 2 28s ---查询pod信息;replicaset会再创建pod [root@master manifests] # kubectl get pods NAME READY STATUS RESTARTS AGE myapp-deploy-69b47bc96d-bm8zc 1 /1 Running 0 18s myapp-deploy-69b47bc96d-pjr5v 1 /1 Running 0 18s |
3.4 Deployment动态扩容/缩容
有2中方法实现
(1)方法1:直接修改yaml文件,将副本数改为3
1
2
3
4
5
6
7
|
[root@master manifests] # vim deploy-damo.yaml ... ... spec: replicas: 3 ... ... [root@master manifests] # kubectl apply -f deploy-damo.yaml deployment.apps /myapp-deploy configured |
查询验证成功:有3个pod
1
2
3
4
5
|
[root@master manifests] # kubectl get pods NAME READY STATUS RESTARTS AGE myapp-deploy-69b47bc96d-bcdnq 1 /1 Running 0 25s myapp-deploy-69b47bc96d-bm8zc 1 /1 Running 0 2m myapp-deploy-69b47bc96d-pjr5v 1 /1 Running 0 2m |
(2)通过patch命令打补丁命令扩容
与方法1的区别:不需修改yaml文件;平常测试时使用方便;
但列表格式复杂,极容易出错
1
2
|
[root@master manifests] # kubectl patch deployment myapp-deploy -p '{"spec":{"replicas":5}}' deployment.extensions /myapp-deploy patched |
查询验证成功:有5个pod
1
2
3
4
5
6
7
|
[root@master ~] # kubectl get pods NAME READY STATUS RESTARTS AGE myapp-deploy-67f6f6b4dc-2756p 1 /1 Running 0 26s myapp-deploy-67f6f6b4dc-2lkwr 1 /1 Running 0 26s myapp-deploy-67f6f6b4dc-knttd 1 /1 Running 0 21m myapp-deploy-67f6f6b4dc-ms7t2 1 /1 Running 0 21m myapp-deploy-67f6f6b4dc-vl2th 1 /1 Running 0 21m |
3.5 Deployment在线升级版本
(1)直接修改deploy-damo.yaml
1
2
3
4
5
6
7
|
[root@master manifests] # vim deploy-damo.yaml ... ... spec: containers: - name: myapp image: ikubernetes /myapp :v2 ... ... |
(2)可以动态监控版本升级
1
|
[root@master ~] # kubectl get pods -w |
发现是滚动升级,先停一个,再开启一个新的(升级);再依次听一个...
(3)验证:访问服务,版本升级成功
1
2
3
4
5
|
[root@master ~] # kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE myapp-deploy-67f6f6b4dc-6lv66 1 /1 Running 0 2m 10.244.1.75 node1 [root@master ~] # curl 10.244.1.75 Hello MyApp | Version: v2 | <a href= "hostname.html" >Pod Name< /a > |
3.6 Deployment修改版本更新策略
(1)方法1:修改yaml文件
1
2
3
4
5
6
7
|
[root@master manifests] # vim deploy-damo.yaml ... ... strategy: rollingUpdate: maxSurge: 1 #每次更新一个pod maxUnavailable: 0 #最大不可用pod为0 ... ... |
(2)打补丁:修改更新策略
1
2
|
[root@master manifests] # kubectl patch deployment myapp-deploy -p '{"spec":{"strategy":{"rollingUpdate":{"maxSurge":1,"maxUnavailable":0}}}}' deployment.extensions /myapp-deploy patched |
(3)验证:查询详细信息
1
2
3
4
|
[root@master manifests] # kubectl describe deployment myapp-deploy ... ... RollingUpdateStrategy: 0 max unavailable, 1 max surge ... ... |
(4)升级到v3版
① 金丝雀发布:先更新一个pod,然后立即暂停,等版本运行没问题了,再继续发布
1
2
3
|
[root@master manifests] # kubectl set image deployment myapp-deploy myapp=ikubernetes/myapp:v3 && kubectl rollout pause deployment myapp-deploy deployment.extensions /myapp-deploy image updated #一个pod更新成功 deployment.extensions /myapp-deploy paused #暂停更新 |
② 等版本运行没问题了,解除暂停,继续发布更新
1
2
|
[root@master manifests] # kubectl rollout resume deployment myapp-deploy deployment.extensions /myapp-deploy resumed |
③ 中间可以一直监控过程
1
2
3
4
5
6
7
8
9
10
|
[root@master ~] # kubectl rollout status deployment myapp-deploy #输出版本更新信息 Waiting for deployment "myapp-deploy" rollout to finish: 1 out of 5 new replicas have been updated... Waiting for deployment spec update to be observed... Waiting for deployment spec update to be observed... Waiting for deployment "myapp-deploy" rollout to finish: 1 out of 5 new replicas have been updated... Waiting for deployment "myapp-deploy" rollout to finish: 1 out of 5 new replicas have been updated... Waiting for deployment "myapp-deploy" rollout to finish: 2 out of 5 new replicas have been updated... Waiting for deployment "myapp-deploy" rollout to finish: 2 out of 5 new replicas have been updated... ---也可以使用get查询pod 更新过程 [root@master ~] # kubectl get pods -w |
④ 验证:随便访问一个pod的服务,版本升级成功
1
2
3
4
5
|
[root@master ~] # kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE myapp-deploy-6bdcd6755d-2bnsl 1 /1 Running 0 1m 10.244.1.77 node1 [root@master ~] # curl 10.244.1.77 Hello MyApp | Version: v3 | <a href= "hostname.html" >Pod Name< /a > |
3.7 Deployment版本回滚
(1)命令
查询版本变更历史
1
|
$ kubectl rollout history deployment deployment_name |
undo回滚版本;--to-revision= 回滚到第几版本
1
|
$ kubectl rollout undo deployment deployment_name --to-revision=N |
(2)演示
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
---查询版本变更历史 [root@master manifests] # kubectl rollout history deployment myapp-deploy deployments "myapp-deploy" REVISION CHANGE-CAUSE 1 <none> 2 <none> 3 <none> ---回滚到第1版本 [root@master manifests] # kubectl rollout undo deployment myapp-deploy --to-revision=1 deployment.extensions /myapp-deploy [root@master manifests] # kubectl rollout history deployment myapp-deploy deployments "myapp-deploy" REVISION CHANGE-CAUSE 2 <none> 3 <none> 4 <none> |
(3)查询验证,已经回到v1版了
1
2
3
4
5
|
[root@master manifests] # kubectl get rs -o wide NAME DESIRED CURRENT READY AGE CONTAINERS IMAGES SELECTOR myapp-deploy-67f6f6b4dc 0 0 0 18h myapp ikubernetes /myapp :v2 app=myapp,pod-template- hash =2392926087,release=canary myapp-deploy-69b47bc96d 5 5 5 18h myapp ikubernetes /myapp :v1 app=myapp,pod-template- hash =2560367528,release=canary myapp-deploy-6bdcd6755d 0 0 0 10m myapp ikubernetes /myapp :v3 app=myapp,pod-template- hash =2687823118,release=canary |
4、DaemonSet
4.1 DaemonSet简述
(1)介绍
DaemonSet保证在每个Node上都运行一个容器副本,常用来部署一些集群的日志、监控或者其他系统管理应用
(2)典型的应用包括
- 日志收集,比如fluentd,logstash等
- 系统监控,比如Prometheus Node Exporter,collectd,New Relic agent,Ganglia gmond等
- 系统程序,比如kube-proxy, kube-dns, glusterd, ceph等
4.2 DaemonSet定义资源清单几个字段
- apiVersion: app/v1 版本
- kind: DaemonSet 类型
- metadata 元数据
- spec 期望状态
- --------------replicaset 也有的选项---------------
- minReadySeconds:应为新创建的pod准备好的最小秒数
- selector:标签选择器
- template:模板(必须的)
- metadata:模板中的元数据
- spec:模板中的期望状态
- --------------daemonset 独有的选项---------------
- revisionHistoryLimit:要保留以允许回滚的旧ReplicaSet的数量,默认10
- updateStrategy:用新pod替换现有DaemonSet pod的更新策略
- status 当前状态
4.3 演示:创建一个简单的DaemonSet
(1)创建并创建一个简单的DaemonSet,启动pod,只后台运行filebeat手机日志服务
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
|
[root@master manifests] # vim ds-demo.yaml apiVersion: apps /v1 kind: DaemonSet metadata: name: filebeat-ds namespace: default spec: selector: matchLabels: app: filebeat release: stable template: metadata: labels: app: filebeat release: stable spec: containers: - name: filebeat image: ikubernetes /filebeat :5.6.5-alpine env : - name: REDIS_HOST value: redis.default.svc.cluster. local - name: REDIS_LOG_LEVEL value: info [root@master manifests] # kubectl apply -f ds-demo.yaml daemonset.apps /myapp-ds created |
(2)查询验证
1
2
3
4
5
6
7
8
9
10
11
|
[root@master ~] # kubectl get ds NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE filebeat-ds 2 2 2 2 2 <none> 6m [root@master ~] # kubectl get pods NAME READY STATUS RESTARTS AGE filebeat-ds-r25hh 1 /1 Running 0 4m filebeat-ds-vvntb 1 /1 Running 0 4m [root@master ~] # kubectl exec -it filebeat-ds-r25hh -- /bin/sh / # ps aux PID USER TIME COMMAND 1 root 0:00 /usr/local/bin/filebeat -e -c /etc/filebeat/filebeat .yml |
4.4 DaemonSet动态版本升级
(1)使用kubectl set image 命令更新pod的镜像;实现版本升级
1
2
|
[root@master ~] # kubectl set image daemonsets filebeat-ds filebeat=ikubernetes/filebeat:5.6.6-alpine daemonset.extensions /filebeat-ds image updated |
(2)验证,升级成功
1
2
3
|
[root@master ~] # kubectl get ds -o wide NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE CONTAINERS IMAGES SELECTOR filebeat-ds 2 2 2 2 2 <none> 7m filebeat ikubernetes /filebeat :5.6.6-alpine app=filebeat,release=stable |