K8s部署单节点Zookeeper并进行监控
0、写在前面
1> K8s监控Zookeeper,这里并没有使用zookeeper-exporter的方式进行监控,主要是由于zookeeper-exporter提供的相关指标不太全,zookeeper官网提供的监控指标信息可参看如下地址:https://github.com/apache/zookeeper/blob/master/zookeeper-server/src/main/java/org/apache/zookeeper/server/ServerMetrics.java,同时参看zookeeper官网发现,在zookeeper 3.6版本之后,官网也给出了相对应的监控方式(zookeeper官网地址:https://zookeeper.apache.org/doc/r3.6.4/zookeeperMonitor.html,zookeeper监控相关文档地址:https://github.com/apache/zookeeper/blob/master/zookeeper-docs/src/main/resources/markdown/zookeeperMonitor.md),所以本文采用部署serviceMonitor方式监控zookeeper。
2> zookeeper部署文件由于zookeeper官方镜像对于zookeeper部署时创建的用户名和用户组为zookeeper,但是博主所部属的K8s环境对应的用户名、用户组均为root,所以如果挂载zookeeper配置文件覆盖原有配置文件时,会报只读文件没有操作权限的提示信息,因此这部分在部署时,需要指定zookeeper所使用的用户名、用户组信息,对于zookeeper所使用的用户名、用户组信息,可以在先不指定用户名、用户组部署成功之后,进入shell控制台,通过使用id zookeeper命令查看
3> prometheus监控zookeeper时,需要开放7000端口,prometheus通过暴露出来的端口获取到对应的指标数据
4> zookeeper部署以及serviceMonitor配置,都是在Kuboard中执行,如果使用命令行或者其他可视化操作平台请自行按照相关操作执行
1.1、部署Deployment
---
kind: Deployment
apiVersion: apps/v1
metadata:
labels:
app: zk-deployment
name: zookeeper
namespace: k8s-middleware
spec:
replicas: 1
selector:
matchLabels:
app: zk
template:
metadata:
name: zk
labels:
app: zk
spec:
# 由于zookeeper部署时定义了用户组为zookeeper,此处使用zookeeper的用户组覆盖当前的用户组
securityContext:
fsGroup: 1000
runAsGroup: 1000
runAsUser: 1000
volumes:
- name: localtime
hostPath:
path: /usr/share/zoneinfo/Asia/Shanghai
- configMap:
defaultMode: 493
name: zookeeper-configmap
name: zkconf
containers:
- name: zookeeper
image: zookeeper:3.6.2
imagePullPolicy: IfNotPresent
volumeMounts:
- name: localtime
mountPath: /etc/localtime
readOnly: true
- mountPath: /conf/zoo.cfg
name: zkconf
subPath: zoo.cfg
---
kind: Service
apiVersion: v1
metadata:
name: zookeeper
namespace: k8s-middleware
labels:
app: zk
spec:
ports:
- port: 2181
name: client
protocol: TCP
targetPort: 2181
- name: metrics
port: 7000
protocol: TCP
targetPort: 7000
clusterIP: None
selector:
app: zk
---
apiVersion: v1
data:
zoo.cfg: >-
dataDir=/data
dataLogDir=/datalog
clientPort=2181
tickTime=2000
initLimit=10
syncLimit=5
autopurge.snapRetainCount=10
autopurge.purgeInterval=24
maxClientCnxns=600
standaloneEnabled=true
admin.enableServer=true
server.1=localhost:2888:3888
## Metrics Providers
# https://prometheus.io Metrics Exporter
metricsProvider.className=org.apache.zookeeper.metrics.prometheus.PrometheusMetricsProvider
metricsProvider.httpPort=7000
metricsProvider.exportJvmInfo=true
kind: ConfigMap
metadata:
name: zookeeper-configmap
namespace: k8s-middleware
---
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
name: zookeeper-prometheus
namespace: k8s-middleware
spec:
endpoints:
- interval: 1m
port: metrics
namespaceSelector:
matchNames:
- k8s-middleware
selector:
matchLabels:
app: zk
3、配置Grafana监控大盘
关于Grafana监控大盘,这里使用的大盘ID是10465,相关介绍可参看如下地址:https://grafana.com/grafana/dashboards/10465-zookeeper-by-prometheus/
使用此大盘时,注意修改变量中数据源信息。