为什么用kibana:当然运维可以直接在dashboard里exec进去,然后命令行看情况,但是开发或者测试不行,那是机密的,我们得要一个页面供他们使用,使用需要kibana。
~]# docker pull kibana:6.8.6
~]# docker images|grep kibana
~]# docker tag adfab5632ef4 harbor.od.com/infra/kibana:v6.8.6
~]# docker push harbor.od.com/infra/kibana:v6.8.6
~]# mkdir /data/k8s-yaml/kibana
~]# cd /data/k8s-yaml/kibana/
kibana]# vi dp.yaml
kind: Deployment
apiVersion: extensions/v1beta1
metadata:
name: kibana
namespace: infra
labels:
name: kibana
spec:
replicas: 1
selector:
matchLabels:
name: kibana
template:
metadata:
labels:
app: kibana
name: kibana
spec:
containers:
- name: kibana
image: harbor.od.com/infra/kibana:v6.8.6
imagePullPolicy: IfNotPresent
ports:
- containerPort: 5601
protocol: TCP
env:
- name: ELASTICSEARCH_URL
value: http://10.4.7.12:9200
imagePullSecrets:
- name: harbor
securityContext:
runAsUser: 0
strategy:
type: RollingUpdate
rollingUpdate:
maxUnavailable: 1
maxSurge: 1
revisionHistoryLimit: 7
progressDeadlineSeconds: 600
kibana]# vi svc.yaml
kind: Service
apiVersion: v1
metadata:
name: kibana
namespace: infra
spec:
ports:
- protocol: TCP
port: 5601
targetPort: 5601
selector:
app: kibana
kibana]# vi ingress.yaml
kind: Ingress
apiVersion: extensions/v1beta1
metadata:
name: kibana
namespace: infra
spec:
rules:
- host: kibana.od.com
http:
paths:
- path: /
backend:
serviceName: kibana
servicePort: 5601
# 11机器,解析域名:
~]# vi /var/named/od.com.zone
serial 前滚一位
kibana A 10.4.7.10
~]# systemctl restart named
~]# dig -t A kibana.od.com @10.4.7.11 +short
# 22机器(21机器还夯着log),应用资源清单:
~]# kubectl apply -f http://k8s-yaml.od.com/kibana/dp.yaml
~]# kubectl apply -f http://k8s-yaml.od.com/kibana/svc.yaml
~]# kubectl apply -f http://k8s-yaml.od.com/kibana/ingress.yaml
~]# kubectl get pods -n infra
我用的低配8C32G,机器快跑不动了,还会显示service not yet
点完可能会转圈转很久
去创建
创建后,你就能看到日志
把prod里的configservice和admin依次起来
# 200机器:
cd /data/k8s-yaml/prod/dubbo-demo-consumer/
dubbo-demo-consumer]# cp ../../test/dubbo-demo-consumer/dp.yaml .
# y
# 修改namespace为prod,fat改成pro,http地址也改了
dubbo-demo-consumer]# vi dp.yaml