Table of Contents
写在前面 #
上一篇文章中kubernetes系列教程（七）深入玩转pod调度介绍了kubernetes中Pod的调度机制，通过实战演练介绍Pod调度到node的几种方法：1. 通过nodeName固定选择调度，2. 通过nodeSelector定向选择调度，3. 通过node Affinity亲和力调度，接下来介绍kubernetes系列教程pod的健康检查机制。
1. 健康检查 #
1.1 健康检查概述 #
应用在运行过程中难免会出现错误，如程序异常，软件异常，硬件故障，网络故障等，kubernetes提供Health Check健康检查机制，当发现应用异常时会自动重启容器，将应用从service服务中剔除，保障应用的高可用性。k8s定义了三种探针Probe：
- readiness probes 准备就绪检查，通过readiness是否准备接受流量，准备完毕加入到endpoint，否则剔除
- liveness probes 在线检查机制，检查应用是否可用，如死锁，无法响应，异常时会自动重启容器
- startup probes 启动检查机制，应用一些启动缓慢的业务，避免业务长时间启动而被前面的探针kill掉
每种探测机制支持三种健康检查方法，分别是命令行exec，httpGet和tcpSocket，其中exec通用性最强，适用与大部分场景，tcpSocket适用于TCP业务，httpGet适用于web业务。
- exec 提供命令或shell的检测，在容器中执行命令检查，返回码为0健康，非0异常
- httpGet http协议探测，在容器中发送http请求，根据http返回码判断业务健康情况
- tcpSocket tcp协议探测，向容器发送tcp建立连接，能建立则说明正常
每种探测方法能支持几个相同的检查参数，用于设置控制检查时间：
- initialDelaySeconds 初始第一次探测间隔，用于应用启动的时间，防止应用还没启动而健康检查失败
- periodSeconds 检查间隔，多久执行probe检查，默认为10s
- timeoutSeconds 检查超时时长，探测应用timeout后为失败
- successThreshold 成功探测阈值，表示探测多少次为健康正常，默认探测1次
1.2 exec命令行健康检查 #
许多应用程序运行过程中无法检测到内部故障，如死锁，出现故障时通过重启业务可以恢复，kubernetes提供liveness在线健康检查机制，我们以exec为例，创建一个容器启动过程中创建一个文件/tmp/liveness-probe.log，10s后将其删除，定义liveness健康检查机制在容器中执行命令ls -l /tmp/liveness-probe.log，通过文件的返回码判断健康状态，如果返回码非0，暂停20s后kubelet会自动将该容器重启。
- 定义一个容器，启动时创建一个文件，健康检查时ls -l /tmp/liveness-probe.log返回码为0，健康检查正常，10s后将其删除，返回码为非0，健康检查异常
1[root@node-1 demo]# cat centos-exec-liveness-probe.yaml
2apiVersion: v1
3kind: Pod
4metadata:
5 name: exec-liveness-probe
6 annotations:
7 kubernetes.io/description: "exec-liveness-probe"
8spec:
9 containers:
10 - name: exec-liveness-probe
11 image: centos:latest
12 imagePullPolicy: IfNotPresent
13 args: #容器启动命令，生命周期为30s
14 - /bin/sh
15 - -c
16 - touch /tmp/liveness-probe.log && sleep 10 && rm -f /tmp/liveness-probe.log && sleep 20
17 livenessProbe:
18 exec: #健康检查机制，通过ls -l /tmp/liveness-probe.log返回码判断容器的健康状态
19 command:
20 - ls
21 - l
22 - /tmp/liveness-probe.log
23 initialDelaySeconds: 1
24 periodSeconds: 5
25 timeoutSeconds: 1
- 应用配置生成容器
1[root@node-1 demo]# kubectl apply -f centos-exec-liveness-probe.yaml
2pod/exec-liveness-probe created
- 查看容器的event日志，容器启动后，10s以内容器状态正常，11s开始执行liveness健康检查，检查异常，触发容器重启
1[root@node-1 demo]# kubectl describe pods exec-liveness-probe | tail
2Events:
3 Type Reason Age From Message
4 ---- ------ ---- ---- -------
5 Normal Scheduled 28s default-scheduler Successfully assigned default/exec-liveness-probe to node-3
6 Normal Pulled 27s kubelet, node-3 Container image "centos:latest" already present on machine
7 Normal Created 27s kubelet, node-3 Created container exec-liveness-probe
8 Normal Started 27s kubelet, node-3 Started container exec-liveness-probe
9 #容器已启动
10 Warning Unhealthy 20s (x2 over 25s) kubelet, node-3 Liveness probe failed: /tmp/liveness-probe.log
11ls: cannot access l: No such file or directory #执行健康检查，检查异常
12 Warning Unhealthy 15s kubelet, node-3 Liveness probe failed: ls: cannot access l: No such file or directory
13ls: cannot access /tmp/liveness-probe.log: No such file or directory
14 Normal Killing 15s kubelet, node-3 Container exec-liveness-probe failed liveness probe, will be restarted
15 #重启容器
- 查看容器重启次数，容器不停的执行，重启次数会响应增加，可以看到RESTARTS的次数在持续增加
1[root@node-1 demo]# kubectl get pods exec-liveness-probe
2NAME READY STATUS RESTARTS AGE
3exec-liveness-probe 1/1 Running 6 5m19s
1.3 httpGet健康检查 #
- httpGet probe主要主要用于web场景，通过向容器发送http请求，根据返回码判断容器的健康状态，返回码小于4xx即表示健康，如下定义一个nginx应用，通过探测http://:port/index.html的方式判断健康状态
1[root@node-1 demo]# cat nginx-httpGet-liveness-readiness.yaml
2apiVersion: v1
3kind: Pod
4metadata:
5 name: nginx-httpget-livess-readiness-probe
6 annotations:
7 kubernetes.io/description: "nginx-httpGet-livess-readiness-probe"
8spec:
9 containers:
10 - name: nginx-httpget-livess-readiness-probe
11 image: nginx:latest
12 ports:
13 - name: http-80-port
14 protocol: TCP
15 containerPort: 80
16 livenessProbe: #健康检查机制，通过httpGet实现实现检查
17 httpGet:
18 port: 80
19 scheme: HTTP
20 path: /index.html
21 initialDelaySeconds: 3
22 periodSeconds: 10
23 timeoutSeconds: 3
- 生成pod并查看健康状态
1[root@node-1 demo]# kubectl apply -f nginx-httpGet-liveness-readiness.yaml
2pod/nginx-httpget-livess-readiness-probe created
3[root@node-1 demo]# kubectl get pods nginx-httpget-livess-readiness-probe
4NAME READY STATUS RESTARTS AGE
5nginx-httpget-livess-readiness-probe 1/1 Running 0 6s
- 模拟故障，将pod中的path文件所属文件删除，此时发送http请求时会健康检查异常，会触发容器自动重启
1查询pod所属的节点
2[root@node-1 demo]# kubectl get pods nginx-httpget-livess-readiness-probe -o wide
3NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
4nginx-httpget-livess-readiness-probe 1/1 Running 1 3m9s 10.244.2.19 node-3 <none> <none>
5
6登录到pod中将文件删除
7[root@node-1 demo]# kubectl exec -it nginx-httpget-livess-readiness-probe /bin/bash
8root@nginx-httpget-livess-readiness-probe:/# ls -l /usr/share/nginx/html/index.html
9-rw-r--r-- 1 root root 612 Sep 24 14:49 /usr/share/nginx/html/index.html
10root@nginx-httpget-livess-readiness-probe:/# rm -f /usr/share/nginx/html/index.html
- 再次查看pod的列表，此时会RESTART的次数会增加1，表示重启重启过一次，AGE则多久前重启的时间
1[root@node-1 demo]# kubectl get pods nginx-httpget-livess-readiness-probe
2NAME READY STATUS RESTARTS AGE
3nginx-httpget-livess-readiness-probe 1/1 Running 1 4m22s
- 查看pod的详情，观察容器重启的情况，通过Liveness 检查容器出现404错误，触发重启。
1[root@node-1 demo]# kubectl describe pods nginx-httpget-livess-readiness-probe | tail
2Events:
3 Type Reason Age From Message
4 ---- ------ ---- ---- -------
5 Normal Scheduled 5m45s default-scheduler Successfully assigned default/nginx-httpget-livess-readiness-probe to node-3
6 Normal Pulling 3m29s (x2 over 5m45s) kubelet, node-3 Pulling image "nginx:latest"
7 Warning Unhealthy 3m29s (x3 over 3m49s) kubelet, node-3 Liveness probe failed: HTTP probe failed with statuscode: 404
8 Normal Killing 3m29s kubelet, node-3 Container nginx-httpget-livess-readiness-probe failed liveness probe, will be restarted
9 Normal Pulled 3m25s (x2 over 5m41s) kubelet, node-3 Successfully pulled image "nginx:latest"
10 Normal Created 3m25s (x2 over 5m40s) kubelet, node-3 Created container nginx-httpget-livess-readiness-probe
11 Normal Started 3m25s (x2 over 5m40s) kubelet, node-3 Started container nginx-httpget-livess-readiness-probe
1.4 tcpSocket健康检查 #
- tcpsocket健康检查适用于TCP业务，通过向指定容器建立一个tcp连接，可以建立连接则健康检查正常，否则健康检查异常，依旧以nignx为例使用tcp健康检查机制，探测80端口的连通性
1[root@node-1 demo]# cat nginx-tcp-liveness.yaml
2apiVersion: v1
3kind: Pod
4metadata:
5 name: nginx-tcp-liveness-probe
6 annotations:
7 kubernetes.io/description: "nginx-tcp-liveness-probe"
8spec:
9 containers:
10 - name: nginx-tcp-liveness-probe
11 image: nginx:latest
12 ports:
13 - name: http-80-port
14 protocol: TCP
15 containerPort: 80
16 livenessProbe: #健康检查为tcpSocket，探测TCP 80端口
17 tcpSocket:
18 port: 80
19 initialDelaySeconds: 3
20 periodSeconds: 10
21 timeoutSeconds: 3
- 应用配置创建容器
1[root@node-1 demo]# kubectl apply -f nginx-tcp-liveness.yaml
2pod/nginx-tcp-liveness-probe created
3
4[root@node-1 demo]# kubectl get pods nginx-tcp-liveness-probe
5NAME READY STATUS RESTARTS AGE
6nginx-tcp-liveness-probe 1/1 Running 0 6s
- 模拟故障，获取pod所属节点，登录到pod中，安装查看进程工具htop
1获取pod所在node
2[root@node-1 demo]# kubectl get pods nginx-tcp-liveness-probe -o wide
3NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
4nginx-tcp-liveness-probe 1/1 Running 0 99s 10.244.2.20 node-3 <none> <none>
5
6登录到pod中
7[root@node-1 demo]# kubectl exec -it nginx-httpget-livess-readiness-probe /bin/bash
8
9#执行apt-get update更新和apt-get install htop安装工具
10root@nginx-httpget-livess-readiness-probe:/# apt-get update
11Get:1 http://cdn-fastly.deb.debian.org/debian buster InRelease [122 kB]
12Get:2 http://security-cdn.debian.org/debian-security buster/updates InRelease [39.1 kB]
13Get:3 http://cdn-fastly.deb.debian.org/debian buster-updates InRelease [49.3 kB]
14Get:4 http://security-cdn.debian.org/debian-security buster/updates/main amd64 Packages [95.7 kB]
15Get:5 http://cdn-fastly.deb.debian.org/debian buster/main amd64 Packages [7899 kB]
16Get:6 http://cdn-fastly.deb.debian.org/debian buster-updates/main amd64 Packages [5792 B]
17Fetched 8210 kB in 3s (3094 kB/s)
18Reading package lists... Done
19root@nginx-httpget-livess-readiness-probe:/# apt-get install htop
20Reading package lists... Done
21Building dependency tree
22Reading state information... Done
23Suggested packages:
24 lsof strace
25The following NEW packages will be installed:
26 htop
270 upgraded, 1 newly installed, 0 to remove and 5 not upgraded.
28Need to get 92.8 kB of archives.
29After this operation, 230 kB of additional disk space will be used.
30Get:1 http://cdn-fastly.deb.debian.org/debian buster/main amd64 htop amd64 2.2.0-1+b1 [92.8 kB]
31Fetched 92.8 kB in 0s (221 kB/s)
32debconf: delaying package configuration, since apt-utils is not installed
33Selecting previously unselected package htop.
34(Reading database ... 7203 files and directories currently installed.)
35Preparing to unpack .../htop_2.2.0-1+b1_amd64.deb ...
36Unpacking htop (2.2.0-1+b1) ...
37Setting up htop (2.2.0-1+b1) ...
- 运行htop查看进程，容器进程通常为1
- kill掉进程观察容器状态，观察RESTART次数重启次数增加
1root@nginx-httpget-livess-readiness-probe:/# kill 1
2root@nginx-httpget-livess-readiness-probe:/# command terminated with exit code 137
3
4查看pod情况
5[root@node-1 demo]# kubectl get pods nginx-tcp-liveness-probe
6NAME READY STATUS RESTARTS AGE
7nginx-tcp-liveness-probe 1/1 Running 1 13m
- 查看容器详情，发现容器有重启的记录
1[root@node-1 demo]# kubectl describe pods nginx-tcp-liveness-probe | tail
2Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
3 node.kubernetes.io/unreachable:NoExecute for 300s
4Events:
5 Type Reason Age From Message
6 ---- ------ ---- ---- -------
7 Normal Scheduled 14m default-scheduler Successfully assigned default/nginx-tcp-liveness-probe to node-3
8 Normal Pulling 44s (x2 over 14m) kubelet, node-3 Pulling image "nginx:latest"
9 Normal Pulled 40s (x2 over 14m) kubelet, node-3 Successfully pulled image "nginx:latest"
10 Normal Created 40s (x2 over 14m) kubelet, node-3 Created container nginx-tcp-liveness-probe
11 Normal Started 40s (x2 over 14m) kubelet, node-3 Started container nginx-tcp-liveness-probe
1.5 readiness健康就绪 #
就绪检查用于应用接入到service的场景，用于判断应用是否已经就绪完毕，即是否可以接受外部转发的流量，健康检查正常则将pod加入到service的endpoints中，健康检查异常则从service的endpoints中删除，避免影响业务的访问。
- 创建一个pod，使用httpGet的健康检查机制，定义readiness就绪检查探针检查路径/test.html
1[root@node-1 demo]# cat httpget-liveness-readiness-probe.yaml
2apiVersion: v1
3kind: Pod
4metadata:
5 name: nginx-tcp-liveness-probe
6 annotations:
7 kubernetes.io/description: "nginx-tcp-liveness-probe"
8 labels: #需要定义labels，后面定义的service需要调用
9 app: nginx
10spec:
11 containers:
12 - name: nginx-tcp-liveness-probe
13 image: nginx:latest
14 ports:
15 - name: http-80-port
16 protocol: TCP
17 containerPort: 80
18 livenessProbe: #存活检查探针
19 httpGet:
20 port: 80
21 path: /index.html
22 scheme: HTTP
23 initialDelaySeconds: 3
24 periodSeconds: 10
25 timeoutSeconds: 3
26 readinessProbe: #就绪检查探针
27 httpGet:
28 port: 80
29 path: /test.html
30 scheme: HTTP
31 initialDelaySeconds: 3
32 periodSeconds: 10
33 timeoutSeconds: 3
- 定义一个service，将上述的pod加入到service中，注意使用上述定义的labels，app=nginx
1[root@node-1 demo]# cat nginx-service.yaml
2apiVersion: v1
3kind: Service
4metadata:
5 labels:
6 app: nginx
7 name: nginx-service
8spec:
9 ports:
10 - name: http
11 port: 80
12 protocol: TCP
13 targetPort: 80
14 selector:
15 app: nginx
16 type: ClusterIP
- 生成配置
1[root@node-1 demo]# kubectl apply -f httpget-liveness-readiness-probe.yaml
2pod/nginx-tcp-liveness-probe created
3[root@node-1 demo]# kubectl apply -f nginx-service.yaml
4service/nginx-service created
- 此时pod状态正常，此时readiness健康检查异常
1[root@node-1 ~]# kubectl get pods nginx-httpget-livess-readiness-probe
2NAME READY STATUS RESTARTS AGE
3nginx-httpget-livess-readiness-probe 1/1 Running 2 153m
4
5#readiness健康检查异常，404报错（最后一行）
6[root@node-1 demo]# kubectl describe pods nginx-tcp-liveness-probe | tail
7 node.kubernetes.io/unreachable:NoExecute for 300s
8Events:
9 Type Reason Age From Message
10 ---- ------ ---- ---- -------
11 Normal Scheduled 2m6s default-scheduler Successfully assigned default/nginx-tcp-liveness-probe to node-3
12 Normal Pulling 2m5s kubelet, node-3 Pulling image "nginx:latest"
13 Normal Pulled 2m1s kubelet, node-3 Successfully pulled image "nginx:latest"
14 Normal Created 2m1s kubelet, node-3 Created container nginx-tcp-liveness-probe
15 Normal Started 2m1s kubelet, node-3 Started container nginx-tcp-liveness-probe
16 Warning Unhealthy 2s (x12 over 112s) kubelet, node-3 Readiness probe failed: HTTP probe failed with statuscode: 404
- 查看services的endpoints，发现此时endpoints为空,因为readiness就绪检查异常，kubelet认为此时pod并未就绪，因此并未将其加入到endpoints中。
1[root@node-1 ~]# kubectl describe services nginx-service
2Name: nginx-service
3Namespace: default
4Labels: app=nginx
5Annotations: kubectl.kubernetes.io/last-applied-configuration:
6 {"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"labels":{"app":"nginx"},"name":"nginx-service","namespace":"default"},"s...
7Selector: app=nginx
8Type: ClusterIP
9IP: 10.110.54.40
10Port: http 80/TCP
11TargetPort: 80/TCP
12Endpoints: <none> #Endpoints对象为空
13Session Affinity: None
14Events: <none>
15
16#endpoints状态
17[root@node-1 demo]# kubectl describe endpoints nginx-service
18Name: nginx-service
19Namespace: default
20Labels: app=nginx
21Annotations: endpoints.kubernetes.io/last-change-trigger-time: 2019-09-30T14:27:37Z
22Subsets:
23 Addresses: <none>
24 NotReadyAddresses: 10.244.2.22 #pod处于NotReady状态
25 Ports:
26 Name Port Protocol
27 ---- ---- --------
28 http 80 TCP
29
30Events: <none>
- 进入到pod中手动创建网站文件，使readiness健康检查正常
1[root@node-1 ~]# kubectl exec -it nginx-httpget-livess-readiness-probe /bin/bash
2root@nginx-httpget-livess-readiness-probe:/# echo "readiness probe demo" >/usr/share/nginx/html/test.html
- 此时readiness健康检查正常，kubelet检测到pod就绪会将其加入到endpoints中
1健康检查正常
2[root@node-1 demo]# curl http://10.244.2.22/test.html
3
4查看endpoints情况
5readines[root@node-1 demo]# kubectl describe endpoints nginx-service
6Name: nginx-service
7Namespace: default
8Labels: app=nginx
9Annotations: endpoints.kubernetes.io/last-change-trigger-time: 2019-09-30T14:33:01Z
10Subsets:
11 Addresses: 10.244.2.22 #就绪地址，已从NotReady中提出，加入到正常的Address列表中
12 NotReadyAddresses: <none>
13 Ports:
14 Name Port Protocol
15 ---- ---- --------
16 http 80 TCP
17
18查看service状态
19[root@node-1 demo]# kubectl describe services nginx-service
20Name: nginx-service
21Namespace: default
22Labels: app=nginx
23Annotations: kubectl.kubernetes.io/last-applied-configuration:
24 {"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"labels":{"app":"nginx"},"name":"nginx-service","namespace":"default"},"s...
25Selector: app=nginx
26Type: ClusterIP
27IP: 10.110.54.40
28Port: http 80/TCP
29TargetPort: 80/TCP
30Endpoints: 10.244.2.22:80 #已和endpoints关联
31Session Affinity: None
32Events: <none>
- 同理，如果此时容器的健康检查异常，kubelet会自动将其动endpoint中
1删除站点信息，使健康检查异常
2[root@node-1 demo]# kubectl exec -it nginx-tcp-liveness-probe /bin/bash
3root@nginx-tcp-liveness-probe:/# rm -f /usr/share/nginx/html/test.html
4
5查看pod健康检查event日志
6[root@node-1 demo]# kubectl get pods nginx-tcp-liveness-probe
7NAME READY STATUS RESTARTS AGE
8nginx-tcp-liveness-probe 0/1 Running 0 11m
9[root@node-1 demo]# kubectl describe pods nginx-tcp-liveness-probe | tail
10 node.kubernetes.io/unreachable:NoExecute for 300s
11Events:
12 Type Reason Age From Message
13 ---- ------ ---- ---- -------
14 Normal Scheduled 12m default-scheduler Successfully assigned default/nginx-tcp-liveness-probe to node-3
15 Normal Pulling 12m kubelet, node-3 Pulling image "nginx:latest"
16 Normal Pulled 11m kubelet, node-3 Successfully pulled image "nginx:latest"
17 Normal Created 11m kubelet, node-3 Created container nginx-tcp-liveness-probe
18 Normal Started 11m kubelet, node-3 Started container nginx-tcp-liveness-probe
19 Warning Unhealthy 119s (x32 over 11m) kubelet, node-3 Readiness probe failed: HTTP probe failed with statuscode: 404
20
21查看endpoints
22[root@node-1 demo]# kubectl describe endpoints nginx-service
23Name: nginx-service
24Namespace: default
25Labels: app=nginx
26Annotations: endpoints.kubernetes.io/last-change-trigger-time: 2019-09-30T14:38:01Z
27Subsets:
28 Addresses: <none>
29 NotReadyAddresses: 10.244.2.22 #健康检查异常，此时加入到NotReady状态
30 Ports:
31 Name Port Protocol
32 ---- ---- --------
33 http 80 TCP
34
35Events: <none>
36
37查看service状态，此时endpoints为空
38[root@node-1 demo]# kubectl describe services nginx-service
39Name: nginx-service
40Namespace: default
41Labels: app=nginx
42Annotations: kubectl.kubernetes.io/last-applied-configuration:
43 {"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"labels":{"app":"nginx"},"name":"nginx-service","namespace":"default"},"s...
44Selector: app=nginx
45Type: ClusterIP
46IP: 10.110.54.40
47Port: http 80/TCP
48TargetPort: 80/TCP
49Endpoints: #为空
50Session Affinity: None
51Events: <none>
1.6 TKE设置健康检查 #
TKE中可以设定应用的健康检查机制，健康检查机制包含在不同的Workload中，可以通过模板生成健康监测机制，定义过程中可以选择高级选项，默认健康检查机制是关闭状态，包含前面介绍的两种探针：存活探针livenessProbe和就绪探针readinessProbe，根据需要分别开启
TKE健康检查
开启探针之后进入设置健康检查，支持上述介绍的三种方法：执行命令检查、TCP端口检查，HTTP请求检查
TKE健康检查方法
选择不同的检查方法填写不同的参数即可，如启动间隔，检查间隔，响应超时，等参数，以HTTP请求检查方法为例：
TKE http健康检查方法
设置完成后创建workload时候会自动生成yaml文件，以刚创建的deployment为例，生成健康检查yaml文件内容如下：
1apiVersion: apps/v1beta2
2kind: Deployment
3metadata:
4 annotations:
5 deployment.kubernetes.io/revision: "1"
6 description: tke-health-check-demo
7 creationTimestamp: "2019-09-30T12:28:42Z"
8 generation: 1
9 labels:
10 k8s-app: tke-health-check-demo
11 qcloud-app: tke-health-check-demo
12 name: tke-health-check-demo
13 namespace: default
14 resourceVersion: "2060365354"
15 selfLink: /apis/apps/v1beta2/namespaces/default/deployments/tke-health-check-demo
16 uid: d6cf1f25-e37d-11e9-87fd-567eb17a3218
17spec:
18 minReadySeconds: 10
19 progressDeadlineSeconds: 600
20 replicas: 1
21 revisionHistoryLimit: 10
22 selector:
23 matchLabels:
24 k8s-app: tke-health-check-demo
25 qcloud-app: tke-health-check-demo
26 strategy:
27 rollingUpdate:
28 maxSurge: 1
29 maxUnavailable: 0
30 type: RollingUpdate
31 template:
32 metadata:
33 creationTimestamp: null
34 labels:
35 k8s-app: tke-health-check-demo
36 qcloud-app: tke-health-check-demo
37 spec:
38 containers:
39 - image: nginx:latest
40 imagePullPolicy: Always
41 livenessProbe: #通过模板生成的健康检查机制
42 failureThreshold: 1
43 httpGet:
44 path: /
45 port: 80
46 scheme: HTTP
47 periodSeconds: 3
48 successThreshold: 1
49 timeoutSeconds: 2
50 name: tke-health-check-demo
51 resources:
52 limits:
53 cpu: 500m
54 memory: 1Gi
55 requests:
56 cpu: 250m
57 memory: 256Mi
58 securityContext:
59 privileged: false
60 procMount: Default
61 terminationMessagePath: /dev/termination-log
62 terminationMessagePolicy: File
63 dnsPolicy: ClusterFirst
64 imagePullSecrets:
65 - name: qcloudregistrykey
66 - name: tencenthubkey
67 restartPolicy: Always
68 schedulerName: default-scheduler
69 securityContext: {}
70 terminationGracePeriodSeconds: 30
写在最后 #
本章介绍kubernetes中健康检查两种Probe：livenessProbe和readinessProbe，livenessProbe主要用于存活检查，检查容器内部运行状态，readiness主要用于就绪检查，是否可以接受流量，通常需要和service的endpoints结合，当就绪准备妥当时加入到endpoints中，当就绪异常时从endpoints中删除，从而实现了services的健康检查和服务探测机制。对于Probe机制提供了三种检测的方法，分别适用于不同的场景：1. exec命令行，通过命令或shell实现健康检查，2. tcpSocket通过TCP协议探测端口，建立tcp连接，3. httpGet通过建立http请求探测，读者可多实操掌握其用法。
附录 #
健康检查：https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
TKE健康检查设置方法：https://cloud.tencent.com/document/product/457/32815
『 转载 』该文章来源于网络，侵删。