一、pod的进阶
pod的进阶:
1.1、pod的生命周期当中的状态:
1、Running运行中,pod已经分配到节点上且pod内的容器正常运行。正常状态(ready 1/1)。
2、complete:完成之后退出,容器内的返回码是0,echo $?(表示容器正常运行结束)
3、pending:挂起状态,pod已经创建好了,但是没有被分配到节点上。
面试题:出现pending状态如何解决?
- 1、节点上的资源不足 nginx------node1 node1---->换个节点
- 2、污点,节点上设置了污点标签,导致节点不可部署
- 3、pv,节点上没有合适的pv挂载点(手动),创建pv失败。(手动,自动)
- 4、网络原因,防火墙导致节点不可用。
- 5、swap没有关闭,k8s禁止使用交换分区。
- 6、HostPort已经被占用,NodePort节点上的端口被占用,也会pending。
- 7、ImagePullBackOff:镜像拉取失败
- 8、CrashLoopBackOff:容器已经启动了,但是异常退出了,可以看日志,或者查看详细信息情况。
- 9、error:pod启动过程中报错,日志可以查询。
- 10、PodInitializing:初始化中(pod内部有初始化init容器)
- 11、Evicte:pod被驱逐。
pod状态一览:
CrashLoopBackOff: 容器退出,kubelet正在将它重启
InvalidImageName: 无法解析镜像名称
ImageInspectError: 无法校验镜像
ErrImageNeverPull: 策略禁止拉取镜像
ImagePullBackOff: 正在重试拉取
RegistryUnavailable: 连接不到镜像中心
ErrImagePull: 通用的拉取镜像出错
CreateContainerConfigError: 不能创建kubelet使用的容器配置
CreateContainerError: 创建容器失败
m.internalLifecycle.PreStartContainer 执行hook报错
RunContainerError: 启动容器失败
PostStartHookError: 执行hook报错
ContainersNotInitialized: 容器没有初始化完毕
ContainersNotReady: 容器没有准备完毕
ContainerCreating: 容器创建中
PodInitializing:pod 初始化中
DockerDaemonNotReady: docker还没有完全启动
NetworkPluginNotReady: 网络插件还没有完全启动
Evicte: pod被驱赶
4、Failed:失败:容器内的返回码是非0状态退出,进入失败状态。
logs -f 可以查看pod的日志 describe pod 查看pod的详细情况,也可以查询到错误原因。
5、Terminating(终止中)
pod正在删除中。
6、Unknown:未知
集群出现问题了,API出现了问题,或者是API server和调度器之间通信有问题(证书过期)。
1.2、资源限制
docker cpu 100000
内存 m g
k8s cpu 最小单位 100m 0.1 一个cpu的10%
1000m 1 沾满一个cpu
500m 0.5 占cpu的50%
2 占2个cpu
内存:单位 Ki Mi Gi Ti
1、cpu和内存做资源限制
[root@master01 k8s-yaml]# vim test1.yml#定义api接口的版本
apiVersion: v1
kind: Pod
metadata:name: nginx1labels:app: nginx1
spec:containers:- name: nginximage: nginx:1.22resources:requests:cpu: "0.5"memory: "512Mi"
#软限制,最低的要求,可以不做limits:cpu: "1"memory: "1Gi"
#硬策略,最多使用这么多[root@master01 k8s-yaml]# kubectl apply -f test1.yml
pod/nginx1 created[root@master01 k8s-yaml]# kubectl describe pod nginx1 Limits:cpu: 1memory: 1GiRequests:cpu: 500mmemory: 512Mi
二、探针probe:(面试必问)
探针是对容器执行定期的检查。
2.1、启动探针
探针:启动探针,在容器启动时,根据条件判断容器是否成功。如果有启动探针和其他探针并列。只有启动探针执行完毕(成功),后续的探针才会执行,启动探针失败,整个容器判定为失败,pod也会进入失败状态。
在整个容器的生命周期当中,只有启动探针在启动时执行,执行成功之后,后续不再执行。
startupProbe
2.2、存活探针
存活探针:livenessProbe 探测容器是否正常运行(Running),如果探测失败,会根据pod的重启策略来决定是否重启。
将伴随整个pod的生命周期。
2.3、就绪探针
就绪探针:readinessProbe 探测pod的状态是否进入ready,如果进入ready状态失败,service将会把这个pod的ip从转发中移除。
service不会把请求转发到这个pod
nginx1 --3 node1 node2 node3
nginx1-pod 10.244.0.10
nginx1-pod2 10.244.0.11-------------service----------NodePort 192.168.168.81:30001
nginx1-pod3 10.244.0.12----------ready 0/1
就绪探针没有检测成功或者失败,pod可能是running,但是ready一定是0/1。
存活探针一般用于容器内的配置文件或者是关键组件是否正常。
就绪探针一般用于指定端口的服务,需要对外提供访问的业务。
这两个探针都会伴随整个pod的生命周期。
2.4、probe的检测方法:
1、exec方法:就是进入容器内,指定命令,命令的返回码是0就是成功,非0都是失败。
在容器内使用自定义命令来检测容器内的健康状况,判断关键配置文件是否存在,依赖环境是否完整等等。
2、tcpSocket方法:对容器的ip地址进行tcp检查(三次握手),和指定的端口进行连接,如果三次握手和端口通信建立连接正常。
则认为成功,判断容器的端口是否正常启动,端口是否处于监听状态。
3、httpGet方法:对容器内的ip+端口进行http请求,请求的方式是get。响应码大于等于200且小于400,都是成功。
200=<x<400,主要应用于web容器。
结果:
1、成功。
3、未知。探针失败,但是不会采取任何行动。
2.5、lifecycle字段:
2.5、容器钩子:
1、postStart----启动钩子
postStart:容器启动时立即执行的命令,执行容器内需要执行的初始化命令,等待依赖环境。
2、preStop----停止钩子
preStop:停止之前执行的任务,清理任务,同步文件等等(导出容器内的数据)。
1、启动任务失败,pod能否进入正常状态
2、停止任务失败,pod能否退出
3、如果和探针一起,启动失败的任务影响探针嘛?
影响,优先级大于启动探针
容器的启动的钩子----->启动探针-------->存活和就绪探针
存活探针
[root@master01 k8s-yaml]# vim test2.yamlapiVersion: apps/v1
kind: Deployment
metadata:name: centoslabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: centos7image: centos:7command: ["/bin/bash","-c","touch /opt/123.txt && sleep 3600"]livenessProbe:exec:command: ["/usr/bin/test","-e","/opt/123.txt"]initialDelaySeconds: 1
#initialDelaySeconds: 表示容器启动之后多少秒开始第一次探测,1是秒,要等待应用程
#序准备好之后再探测。以避免结果有误。periodSeconds: 3
#在pod的生命周期内,探针的检测时间间隔是3秒 ,也没有固定的范围,根据业务容器的>情况来看,比较敏感的检测时间,可以缩短时间间隔
#在pod的生命周期内,探针的检测时间间隔是3秒,也没有固定的范围,根据业务容器的情
况来看,比较敏感的检测时间,可以缩短时间间隔failureThreshold: 2
#表示次数,表示探针检测容器失败几次就把容器标记为不健康。timeoutSeconds: 1
#timeoutSeconds的时间必须小于periodSeconds,表示探针在多少时间之内完成探测。 successThreshold: 1
#只要探针成功一次,就把容器标记为健康,这个值只能是1,默认也是1,可以不写。[root@master01 k8s-yaml]# kubectl apply -f test2.yaml
deployment.apps/centos created
[root@master01 k8s-yaml]# kubectl get pod
[root@master01 k8s-yaml]# kubectl describe pod centos-6746885856-pvd88 Liveness: exec [/usr/bin/test -e /opt/123.txt] delay=1s timeout=1s period=3s #success=1 #failure=2##删除文件,触发存活探针
[root@master01 k8s-yaml]# kubectl exec -it centos-6746885856-nvndp
error: you must specify at least one command for the container
[root@master01 k8s-yaml]# kubectl exec -it centos-6746885856-nvndp bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
[root@centos-6746885856-nvndp /]# cd /opt/
[root@centos-6746885856-nvndp opt]# ls
123.txt
[root@centos-6746885856-nvndp opt]# rm -rf *
[root@centos-6746885856-nvndp opt]# exit
exit
[root@master01 k8s-yaml]# kubectl describe pod centos-6746885856-nvndp Events:Type Reason Age From Message---- ------ ---- ---- -------Normal Scheduled 71s default-scheduler Successfully assigned default/centos-6746885856-nvndp to node02Normal Pulled 70s kubelet Container image "centos:7" already present on machineNormal Created 70s kubelet Created container centos7Normal Started 70s kubelet Started container centos7Warning Unhealthy 1s (x2 over 4s) kubelet Liveness probe failed:Normal Killing 1s kubelet Container centos7 failed liveness probe, will be restarted
探针的优化(面试)
initialDelaySeconds: 1
#initialDelaySeconds: 表示容器启动之后多少秒开始第一次探测,1是秒,要等待应用程
#序准备好之后再探测。以避免结果有误。periodSeconds: 3
#在pod的生命周期内,探针的检测时间间隔是3秒 ,也没有固定的范围,根据业务容器的>情况来看,比较敏感的检测时间,可以缩短时间间隔
#在pod的生命周期内,探针的检测时间间隔是3秒,也没有固定的范围,根据业务容器的情
况来看,比较敏感的检测时间,可以缩短时间间隔failureThreshold: 2
#表示次数,表示探针检测容器失败几次就把容器标记为不健康。timeoutSeconds: 1
#timeoutSeconds的时间必须小于periodSeconds,表示探针在多少时间之内完成探测。 successThreshold: 1
#只要探针成功一次,就把容器标记为健康,这个值只能是1,默认也是1,可以不写。
存活探针不设置参数
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: centoslabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: centos7image: centos:7command: ["/bin/bash","-c","touch /opt/123.txt && sleep 3600"]livenessProbe:exec:command: ["/usr/bin/test","-e","/opt/123.txt"]# initialDelaySeconds: 1
#initialDelaySeconds: 表示容器启动之后多少秒开始第一次探测,1是秒,要等待应用程
#序准备好之后再探测。以避免结果有误。# periodSeconds: 3
#在pod的生命周期内,探针的检测时间间隔是3秒 ,也没有固定的范围,根据业务容器的>情况来看,比较敏感的检测时间,可以缩短时间间隔
#在pod的生命周期内,探针的检测时间间隔是3秒,也没有固定的范围,根据业务容器的情
况来看,比较敏感的检测时间,可以缩短时间间隔# failureThreshold: 2
#表示次数,表示探针检测容器失败几次就把容器标记为不健康。# timeoutSeconds: 1
#timeoutSeconds的时间必须小于periodSeconds,表示探针在多少时间之内完成探测。
# successThreshold: 1
#只要探针成功一次,就把容器标记为健康,这个值只能是1,默认也是1,可以不写。[root@master01 k8s-yaml]# kubectl get pod
[root@master01 k8s-yaml]# kubectl describe pod centos-596c98dd98-wwfbl Liveness: exec [/usr/bin/test -e /opt/123.txt] delay=0s timeout=1s period=10s #success=1 #failure=3
存活探针检测方法为tcp-80
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:tcpSocket:port: 80
[root@master01 k8s-yaml]# kubectl apply -f test3.yaml
deployment.apps/nginx created
[root@master01 k8s-yaml]# kubectl get pod
[root@master01 k8s-yaml]# kubectl describe pod nginx-585c6b6f4b-kfrjv Liveness: tcp-socket :80 delay=0s timeout=1s period=10s #success=1 #failure=3
存活探针检测方法为tcp-81
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:tcpSocket:port: 81[root@master01 k8s-yaml]# kubectl get pod -o wide
nginx-654cfc659-nlkzh 0/1 CrashLoopBackOff 4 2m28s 10.244.2.91 node02 <none> [root@master01 k8s-yaml]# kubectl describe pod nginx-654cfc659-nlkzh
存活探针检测方法为http-81
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:httpGet:port: 80scheme: HTTPpath: /index.html[root@master01 k8s-yaml]# kubectl apply -f test3.yaml
[root@master01 k8s-yaml]# kubectl get pod -o wide
[root@master01 k8s-yaml]# kubectl describe pod nginx-5859bfdf9f-xbqnn Liveness: http-get http://:80/index.html delay=0s timeout=1s period=10s #success=1 #failure=3
启动探针检测方法为http-80+就绪探针–exec指定命令检测文件是否存在
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:startupProbe:httpGet:port: 80scheme: HTTPpath: /index.htmlreadinessProbe:exec: command: ["/usr/bin/test","-e","/etc/passwd"][root@master01 k8s-yaml]# kubectl apply -f test3.yaml
[root@master01 k8s-yaml]# kubectl get pod -o wide
[root@master01 k8s-yaml]# kubectl describe pod nginx-76f8b6d4f7-xt4mp#scheme:调用的协议(http)
#path:path: /index.html curl 192.168.168.81
# initialDelaySeconds: 1
#initialDelaySeconds: 表示容器启动之后多少秒开始第一次探测,1是秒,要等待应用程
#序准备好之后再探测。以避免结果有误。# periodSeconds: 3
#在pod的生命周期内,探针的检测时间间隔是3秒 ,也没有固定的范围,根据业务容器的>情况来看,比较敏感的检测时间,可以缩短时间间隔
#在pod的生命周期内,探针的检测时间间隔是3秒,也没有固定的范围,根据业务容器的情
况来看,比较敏感的检测时间,可以缩短时间间隔# failureThreshold: 2
#表示次数,表示探针检测容器失败几次就把容器标记为不健康。# timeoutSeconds: 1
#timeoutSeconds的时间必须小于periodSeconds,表示探针在多少时间之内完成探测。
# successThreshold: 1
#只要探针成功一次,就把容器标记为健康,这个值只能是1,默认也是1,可以不写。
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:startupProbe:httpGet:port: 80scheme: HTTPpath: /index.htmlreadinessProbe:exec:command: ["/usr/bin/test","-e","/etc/passwd"]
---
#表示分段,上一个yml结束,下一个新的yml
apiVersion: v1
kind: Service
metadata:name: nginx-1
# namespacelabels:app: nginx1
spec:type: NodePortports:- port: 80targetPort: 80nodePort: 30000selector:app: centos7[root@master01 k8s-yaml]# kubectl apply -f test3.yaml [root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 12h
nginx-76f8b6d4f7-hhvdq 1/1 Running 0 41s
[root@master01 k8s-yaml]# kubectl exec -it nginx-76f8b6d4f7-hhvdq bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
root@nginx-76f8b6d4f7-hhvdq:/# cd /usr/share/nginx/html/
root@nginx-76f8b6d4f7-hhvdq:/usr/share/nginx/html# ls
50x.html index.html
root@nginx-76f8b6d4f7-hhvdq:/usr/share/nginx/html# cat index.html
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
html { color-scheme: light dark; }
body { width: 35em; margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif; }
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p><p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p><p><em>Thank you for using nginx.</em></p>
</body>
</html>
root@nginx-76f8b6d4f7-hhvdq:/usr/share/nginx/html# echo 123456 > index.html
root@nginx-76f8b6d4f7-hhvdq:/usr/share/nginx/html# cat index.html
123456
root@nginx-76f8b6d4f7-hhvdq:/usr/share/nginx/html# exit
exit
[root@master01 k8s-yaml]# curl 192.168.168.81
curl: (7) Failed connect to 192.168.168.81:80; 拒绝连接
[root@master01 k8s-yaml]# curl 192.168.168.81:30000
123456[root@master01 k8s-yaml]# vim test3.yaml
[root@master01 k8s-yaml]# kubectl exec -it nginx-76f8b6d4f7-hhvdq bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
root@nginx-76f8b6d4f7-hhvdq:/# rm -rf /etc/passwd
root@nginx-76f8b6d4f7-hhvdq:/# exit
exit
[root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 12h
nginx-76f8b6d4f7-hhvdq 1/1 Running 0 5m48s
[root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 12h
nginx-76f8b6d4f7-hhvdq 0/1 Running 0 8m18s就绪探针生效,文件丢失,ready变化
启动钩子—指定命令检测文件不存在,导致后续容器启动失败
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:startupProbe:httpGet:port: 80scheme: HTTPpath: /index.htmlreadinessProbe:exec:command: ["/usr/bin/test","-e","/etc/passwd"]lifecycle:postStart:exec:command: ["/bin/bash","-c","cat /opt/123.txt"]
---
#表示分段,上一个yml结束,下一个新的yml
apiVersion: v1
kind: Service
metadata:name: nginx-1
# namespacelabels:app: nginx1
spec:type: NodePortports:- port: 80targetPort: 80[root@master01 k8s-yaml]# kubectl apply -f test3.yaml --force
[root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 12h
nginx-df44cb667-kt4xj 0/1 PostStartHookError: command '/bin/bash -c cat /opt/123.txt' exited with 1: cat: /opt/123.txt: No such file or directory1 3s[root@master01 k8s-yaml]# kubectl describe pod nginx-df44cb667-p4vspWarning FailedPostStartHook 4m49s (x4 over 5m33s) kubelet Exec lifecycle hook ([/bin/bash -c cat /opt/123.txt]) for Container "nginx" in Pod "nginx-df44cb667-p4vsp_default(8ee9249a-e118-4e25-8b10-cb53e8cc189f)" failed - error: command '/bin/bash -c cat /opt/123.txt' exited with 1: cat: /opt/123.txt: No such file or directory
, message: "cat: /opt/123.txt: No such file or directory\n"Normal Killing 4m49s (x4 over 5m33s) kubelet FailedPostStartHookWarning BackOff 25s (x27 over 5m31s) kubelet Back-off restarting failed container
lifecycle字段:
容器钩子:
postStart:容器启动时立即执行的命令,执行容器内需要执行的初始化命令,等待依赖环境。
preStop:停止之前执行的任务,清理任务,同步文件等等(导出容器内的数据)。
1、启动任务失败,pod能否进入正常状态
2、停止任务失败,pod能否退出
3、如果和探针一起,启动失败的任务影响探针嘛?
影响,优先级大于启动探针
容器的启动的钩子----->启动探针-------->存活和就绪探针
4、容器的钩子,不论是启动还是停止之前的命令,只能使用exec。
容器的钩子,不论是启动还是停止之前的命令,只能使用exec。
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 3selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:startupProbe:httpGet:port: 81scheme: HTTPpath: /index.htmlreadinessProbe:exec:command: ["/usr/bin/test","-e","/etc/passwd"]lifecycle:postStart:tcpSocket:port: 80---
#表示分段,上一个yml结束,下一个新的yml
apiVersion: v1
kind: Service
metadata:name: nginx-1
# namespacelabels:app: nginx1
spec:type: NodePortports:- port: 80targetPort: 80
[root@master01 k8s-yaml]# kubectl apply -f test3.yaml
[root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 13h
nginx-56f444c575-qq29d 0/1 CrashLoopBackOff 5 7m17s
[root@master01 k8s-yaml]# kubectl describe pod nginx-56f444c575-qq29d Warning FailedPostStartHook 4m6s (x4 over 6m21s) kubelet Cannot run handler: invalid handler: &Handler{Exec:nil,HTTPGet:nil,TCPSocket:&TCPSocketAction{Port:{0 80 },Host:,},}Normal Killing 4m6s (x4 over 6m21s) kubelet FailedPostStartHookWarning BackOff 66s (x17 over 5m19s) kubelet Back-off restarting failed container容器的钩子,不论是启动还是停止之前的命令,只能使用exec。
容器的钩子,不论是启动还是停止之前的命令,只能使用exec。
[root@master01 k8s-yaml]# vim test3.yaml apiVersion: apps/v1
kind: Deployment
metadata:name: nginxlabels:app: centos7
spec:replicas: 1selector:matchLabels:app: centos7template:metadata:labels:app: centos7spec:
#定义pod的容器参数containers:- name: nginximage: nginx:1.22livenessProbe:startupProbe:httpGet:port: 81scheme: HTTPpath: /index.htmlreadinessProbe:exec:command: ["/usr/bin/test","-e","/etc/passwd"]lifecycle:postStart:httpGet:port: 80scheme: HTTPpath: /index.html
---
#表示分段,上一个yml结束,下一个新的yml
apiVersion: v1
kind: Service
metadata:name: nginx-1
# namespacelabels:app: nginx1
spec:type: NodePortports:- port: 80targetPort: 80[root@master01 k8s-yaml]# kubectl apply -f test3.yaml
deployment.apps/nginx created
service/nginx-1 unchanged
[root@master01 k8s-yaml]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nfs1-76f66b958-68wpl 1/1 Running 0 13h
nginx-cbf59dc9f-s4tx6 0/1 PostStartHookError: Get "http://10.244.2.213:80//index.html": dial tcp 10.244.2.213:80: connect: connection refused 0 2s[root@master01 k8s-yaml]# kubectl describe pod nginx-cbf59dc9f-s4tx6 Warning FailedPostStartHook 24s (x3 over 38s) kubelet Http lifecycle hook (/index.html) for Container "nginx" in Pod "nginx-cbf59dc9f-s4tx6_default(ab801d25-2c1f-4ede-a9ac-351d843b8067)" failed - error: Get "http://10.244.2.213:80//index.html": dial tcp 10.244.2.213:80: connect: connection refused, message: ""容器的钩子,不论是启动还是停止之前的命令,只能使用exec。
启动及停止exec指定命令写入文件
[root@master01 k8s-yaml]# vim test4.yamlapiVersion: v1
kind: Pod
metadata:name: centos1labels:app: centos1
spec:containers:- name: centosimage: centos:7command: ["/bin/bash","-c","sleep 30"]lifecycle:postStart:exec:command: ["/bin/bash","-c","echo start > /opt/123.txt && sleep 10"]preStop:exec:command: ["/bin/bash","-c","echo stop > /opt/123.txt && sleep 10"]
启动及停止exec指定命令写入文件,通过挂载目录查看
[root@master01 k8s-yaml]# vim test4.yamlapiVersion: v1
kind: Pod
metadata:name: centos1labels:app: centos1
spec:containers:- name: centosimage: centos:7command: ["/bin/bash","-c","sleep 30"]volumeMounts:- name: data-vmountPath: /opt/test1
#容器内的目录,挂载卷的名称lifecycle:postStart:exec:command: ["/bin/bash","-c","echo start >> /opt/test1/123.txt && sleep 10"]preStop:exec:command: ["/bin/bash","-c","echo stop >> /opt/test1/456.txt && sleep 10"]volumes:- name: data-vhostPath:path: /opt/testtype: DirectoryOrCreate[root@master01 k8s-yaml]# kubectl apply -f test4.yaml --force
pod/centos1 configured
[root@master01 k8s-yaml]# kubectl exec -it centos1 bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
[root@centos1 /]# cd /opt/test1/
[root@centos1 test1]# ls
123.txt 456.txt
[root@centos1 test1]# command terminated with exit code 137
[root@master01 k8s-yaml]# kubectl exec -it centos1 bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
[root@centos1 /]# cd /opt/test1/
[root@centos1 test1]# ls
123.txt 456.txt
[root@centos1 test1]# cat 456.txt
stop
[root@centos1 test1]# cat 123.txt
start
启动及停止exec指定命令写入文件
[root@master01 k8s-yaml]# vim test4.yamlapiVersion: v1
kind: Pod
metadata:name: centos1labels:app: centos1
spec:containers:- name: centosimage: centos:7command: ["/bin/bash","-c","sleep 30"]volumeMounts:- name: data-vmountPath: /opt/test1
#容器内的目录,挂载卷的名称lifecycle:postStart:exec:command: ["/bin/bash","-c","echo start >> /opt/test1/123.txt && sleep 10"]preStop:exec:command: ["/bin/bash","-c","echo stop >> /opt/test1/321.txt && sleep 10"]volumes:- name: data-vhostPath:path: /opt/testtype: DirectoryOrCreate[root@master01 k8s-yaml]# kubectl apply -f test4.yaml
pod/centos1 created
[root@master01 k8s-yaml]# kubectl exec -it centos1 bash
kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead.
[root@centos1 /]# cd /opt/test1/
[root@centos1 test1]# ls
123.txt
[root@centos1 test1]# command terminated with exit code 137
[root@master01 k8s-yaml]# kubectl delete pod centos1
pod "centos1" deleted[root@master01 k8s-yaml]# kubectl get pod -o wide
centos1 0/1 ContainerCreating 0 2s <none> node02 <none> <none>[root@node02 test]# cat 321.txt
stop
探针三种:
启动 : 启动执行完毕之后,后续不再执行。
存活 :
就绪
存活和就绪会伴随整个pod的生命周期
三种方法:
exec
tcpSocket
httpGet
作业:
启动钩子和退出钩子
和节点挂载:/usr/share/nginx/html 节点: /opt/node
exec执行 要能在目录中看到开始和打印的结果
包含探针:
1、启动探针:
方法:exec 检测 /usr/share/nginx/html/index.html 文件是否存在
2、存活探针
方法:httpGET
访问验证返回码是否正确
3、就绪探针
tcpSocket
监听容器的80端口是否正常
启动探针、存活探针、就绪探针、启动及停止钩子通过exec指定命令写入文件
[root@master01 k8s-yaml]# vim test5.yamlapiVersion: v1
kind: Pod
metadata:name: nginx1labels:app: nginx1
spec:containers:- name: nignx1image: nginx:1.22volumeMounts:- name: data-vmountPath: /usr/share/nginx/html
#设置容器钩子lifecycle:
#设置启动钩子postStart:exec:command: ["/bin/bash","-c","echo start >> /usr/share/nginx/html/index.html"]
#设置停止钩子preStop:exec:command: ["/bin/bash","-c","echo stop >> /usr/share/nginx/html/error.html"]startupProbe:exec:command: ["/usr/bin/test","-e","/usr/share/nginx/html/index.html"]livenessProbe:httpGet:port: 80scheme: HTTPpath: /index.htmlreadinessProbe:tcpSocket:port: 80volumes:- name: data-vhostPath:path: /opt/nodetype: DirectoryOrCreate[root@master01 k8s-yaml]# kubectl apply -f test5.yaml [root@master01 k8s-yaml]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx1 1/1 Running 0 51s 10.244.1.66 node01 <none> <none>[root@master01 k8s-yaml]# kubectl describe pod nginx1 Liveness: http-get http://:80/index.html delay=0s timeout=1s period=10s #success=1 #failure=3Readiness: tcp-socket :80 delay=0s timeout=1s period=10s #success=1 #failure=3Startup: exec [/usr/bin/test -e /usr/share/nginx/html/index.html] delay=0s timeout=1s period=10s #success=1 #failure=3Events:Type Reason Age From Message---- ------ ---- ---- -------Normal Scheduled 63s default-scheduler Successfully assigned default/nginx1 to node01Normal Pulled 63s kubelet Container image "nginx:1.22" already present on machineNormal Created 63s kubelet Created container nignx1Normal Started 63s kubelet Started container nignx1[root@node01 node]# cat /opt/node/index.html
start[root@master01 k8s-yaml]# kubectl delete pod nginx1
pod "nginx1" deleted[root@node01 node]# cat error.html
stop