springboot2.X整合prometheus监控的实例讲解

2023-07-04 22:06:03 348

springboot2.x暴露健康状况通过prometheus监控

加入依赖



org.springframework.boot
spring-boot-starter-actuator


io.micrometer
micrometer-registry-prometheus

application.yml加入相关配置

management:
security:
enabled:false
#prometheus+grafana+springboot2监控集成配置
metrics:
export:
prometheus:
enabled:true
jmx:
enabled:true
endpoints:
web:
exposure:
include:'*'
base-path:/metrics
#prometheus+grafana+springboot2监控集成配置

主启动类加入配置

packagecom.drore.saas;
importcom.drore.saas.services.service.StorageService;
importio.micrometer.core.instrument.MeterRegistry;
importorg.springframework.boot.CommandLineRunner;
importorg.springframework.boot.SpringApplication;
importorg.springframework.boot.actuate.autoconfigure.metrics.MeterRegistryCustomizer;
importorg.springframework.boot.autoconfigure.SpringBootApplication;
importorg.springframework.boot.web.servlet.ServletComponentScan;
importorg.springframework.context.annotation.Bean;
@SpringBootApplication
@ServletComponentScan
publicclassTenantappApplication{
publicstaticvoidmain(String[]args){
SpringApplication.run(TenantappApplication.class,args);
}
@Bean
CommandLineRunnerinit(StorageServicestorageService){
return(args)->{
storageService.init();
};
}
#prometheus+grafana+springboot2监控集成配置
@Bean
MeterRegistryCustomizermeterRegistryCustomizer(MeterRegistrymeterRegistry){
returnmeterRegistry1->{
meterRegistry.config()
.commonTags("application","Tenantapp");
};
}
#prometheus+grafana+springboot2监控集成配置
}

启动之后通过路径访问查看健康状况

http://xxxxxx/metrics/prometheus

查看到的数据如下：

[root@saas98/]$curl"http://10.98.94.220:80/ts/metrics/prometheus"
#HELPprocess_start_time_secondsStarttimeoftheprocesssinceunixepoch.
#TYPEprocess_start_time_secondsgauge
process_start_time_seconds{application="Tenantapp",}1.556068841226E9
#HELPtomcat_threads_busy_threads
#TYPEtomcat_threads_busy_threadsgauge
tomcat_threads_busy_threads{application="Tenantapp",name="http-nio-9081",}1.0
#HELPtomcat_sessions_expired_sessions_total
#TYPEtomcat_sessions_expired_sessions_totalcounter
tomcat_sessions_expired_sessions_total{application="Tenantapp",}0.0
#HELPtomcat_sessions_active_current_sessions
#TYPEtomcat_sessions_active_current_sessionsgauge
tomcat_sessions_active_current_sessions{application="Tenantapp",}0.0
#HELPjvm_gc_memory_promoted_bytes_totalCountofpositiveincreasesinthesizeoftheoldgenerationmemorypoolbeforeGCtoafterGC
#TYPEjvm_gc_memory_promoted_bytes_totalcounter
jvm_gc_memory_promoted_bytes_total{application="Tenantapp",}1.18894656E8
#HELPtomcat_global_request_max_seconds
#TYPEtomcat_global_request_max_secondsgauge
tomcat_global_request_max_seconds{application="Tenantapp",name="http-nio-9081",}3.366
#HELPjvm_memory_used_bytesTheamountofusedmemory
#TYPEjvm_memory_used_bytesgauge
jvm_memory_used_bytes{application="Tenantapp",area="heap",id="SurvivorSpace",}653880.0
jvm_memory_used_bytes{application="Tenantapp",area="nonheap",id="Metaspace",}1.36445248E8
jvm_memory_used_bytes{application="Tenantapp",area="heap",id="EdenSpace",}7511504.0
jvm_memory_used_bytes{application="Tenantapp",area="nonheap",id="CodeCache",}3.8031424E7
jvm_memory_used_bytes{application="Tenantapp",area="heap",id="TenuredGen",}1.3880212E8
jvm_memory_used_bytes{application="Tenantapp",area="nonheap",id="CompressedClassSpace",}1.7220968E7
#HELPtomcat_sessions_created_sessions_total
#TYPEtomcat_sessions_created_sessions_totalcounter
tomcat_sessions_created_sessions_total{application="Tenantapp",}0.0
#HELPsystem_cpu_countThenumberofprocessorsavailabletotheJavavirtualmachine
#TYPEsystem_cpu_countgauge
system_cpu_count{application="Tenantapp",}1.0
#HELPtomcat_global_sent_bytes_total
#TYPEtomcat_global_sent_bytes_totalcounter
tomcat_global_sent_bytes_total{application="Tenantapp",name="http-nio-9081",}8168269.0
#HELPjvm_threads_daemon_threadsThecurrentnumberoflivedaemonthreads
#TYPEjvm_threads_daemon_threadsgauge
jvm_threads_daemon_threads{application="Tenantapp",}34.0
#HELPlogback_events_totalNumberoferrorleveleventsthatmadeittothelogs
#TYPElogback_events_totalcounter
logback_events_total{application="Tenantapp",level="debug",}0.0
logback_events_total{application="Tenantapp",level="error",}965.0
logback_events_total{application="Tenantapp",level="warn",}4.0
logback_events_total{application="Tenantapp",level="info",}1047.0
logback_events_total{application="Tenantapp",level="trace",}0.0
#HELPtomcat_cache_access_total
#TYPEtomcat_cache_access_totalcounter
tomcat_cache_access_total{application="Tenantapp",}0.0
#HELPtomcat_servlet_error_total
#TYPEtomcat_servlet_error_totalcounter
tomcat_servlet_error_total{application="Tenantapp",name="dispatcherServlet",}0.0
tomcat_servlet_error_total{application="Tenantapp",name="statViewServlet",}0.0
tomcat_servlet_error_total{application="Tenantapp",name="default",}0.0
#HELPjvm_buffer_total_capacity_bytesAnestimateofthetotalcapacityofthebuffersinthispool

prometheus-operator监控java应用整合

目前的prometheus-operator目录下的相关文件

https://github.com/coreos/prometheus-operator.git

已经移植到

https://github.com/coreos/kube-prometheus.git

具体原因参照官网。

下面也有自己总结的文档的地址，大家可以参照下：

https://github.com/hkj123/kube-prometheus-manifests.git

具体的文件如下图展示，这里有很多文件，我们只介绍关注的几个文件，其他的可以参照官网了解：

构建springboot项目的时候需要注意的点：

---
kind:Service
apiVersion:v1
metadata:
labels:
app:ms#prometheus配置监控需要通过app去监控，这个k8s-app和app都可以
namespace:drore-saas
name:ms
spec:
ports:
-name:http-ms#prometheus配置监控需要通过portname去识别，当然也可以直接配置端口，不过这样看起来更专业
nodePort:31945
protocol:TCP
port:80
targetPort:80
selector:
app:ms
type:NodePort
---

部署prometheus需要注意的点

https://github.com/coreos/kube-prometheus/manifests

提供了一套整合的例子：

需要修改的文件:prometheus-clusterRole.yaml

默认的文件内容：

apiVersion:rbac.authorization.k8s.io/v1
kind:ClusterRole
metadata:
name:prometheus-k8s
rules:
-apiGroups:
-""
resources:
-nodes/metrics
verbs:
-get
-nonResourceURLs:
-/metrics
verbs:
-get

可以发现这里的监控的resources：-nodes/metrics

我们需要监控自己构建的java应用需要扩大权限

apiVersion:rbac.authorization.k8s.io/v1
kind:ClusterRole
metadata:
name:prometheus-k8s
rules:
-apiGroups:
-""
resources:
-nodes
-services
-endpoints
-pods
-nodes/proxy
verbs:
-get
-list
-watch
-apiGroups:
-""
resources:
-configmaps
-nodes/metrics
verbs:
-get
-nonResourceURLs:
-/metrics
verbs:
-get

书写monitor监控文件，提供了一下参考的文件：prometheus-serviceMonitor.yaml

自己编写的：prometheus-serviceMonitorSelfServiceMs.yaml

apiVersion:monitoring.coreos.com/v1
kind:ServiceMonitor
metadata:
name:drore-saas-ms
namespace:monitoring
spec:
endpoints:
-interval:15s
port:http-ms#prometheus配置注意这个port的名称
path:/ms/metrics/prometheus#prometheus配置注意这个路径
jobLabel:app#prometheus配置注意我们用的是app在springboot构建的时候
namespaceSelector:#表示去匹配某一命名空间中的service，如果想从所有的namespace中匹配用any:true
matchNames:
-drore-saas
selector:
matchLabels:
app:ms#prometheus配置注意我们用的是app在springboot构建的时候

配置完毕之后使用

kubectlapply-fprometheus-clusterRole.yaml-fprometheus-serviceMonitorSelfServiceMs.yaml

就可以登录管理界面去看了

http://118.31.17.205:31144/targets

报警规则可以通过prometheus-rules.yaml去配置

参考

https://github.com/coreos/prometheus-operator

补充：springboot2.x暴露监控endpoint并配置prometheus及grafana对多个targets进行监控

1.build.gradle中添加依赖

compilegroup:'org.springframework.boot',name:'spring-boot-starter-actuator'
compilegroup:'org.springframework.boot',name:'spring-boot-starter-security'
compile('io.micrometer:micrometer-registry-prometheus')

2.springboot2.x暴露信息

不同于之前的Actuator1.x，Actuator2.x的大多数端点默认被禁掉。Actuator2.x中的默认端点增加了/actuator前缀。

3.springboot服务端改造

配置信息

server.port=8085
management.endpoints.web.exposure.include=prometheus
management.endpoint.metrics.enabled=true
spring.security.user.name=xxx
spring.security.user.password=123456
spring.security.user.roles=ACTUATOR_ADMIN
management.server.port=8090
management.endpoints.web.base-path=/admin

指定权限

@Configuration
publicclassActuatorSecurityConfigextendsWebSecurityConfigurerAdapter{
@Override
protectedvoidconfigure(HttpSecurityhttp)throwsException{
http.requestMatcher(EndpointRequest.toAnyEndpoint()).authorizeRequests().antMatchers("/actuator/*").hasRole(
"ACTUATOR_ADMIN").and().httpBasic();
}
}

这样我们可以看到：

这个springboot服务启动了两个端口：

8085端口，进行正常的业务处理

8090端口，admin，暴露出监控信息，并通过springsecurity进行权限控制

打开浏览器访问http://192.168.211.2:8090/admin/prometheus

输入用户名密码

4.Prometheus监控多个springboot服务

下载prometheus-2.11.1.linux-amd64.tar.gz

编写配置文件prometheus.yml

global:
scrape_interval:10s
scrape_timeout:10s
evaluation_interval:10m
scrape_configs:
-job_name:app-gateway
scrape_interval:5s
scrape_timeout:5s
metrics_path:/admin/prometheus
scheme:http
basic_auth:
username:root
password:1230456
static_configs:
-targets:
-192.168.211.2:8099
-job_name:search
scrape_interval:5s
scrape_timeout:5s
metrics_path:/admin/prometheus
scheme:http
basic_auth:
username:root
password:1230456
static_configs:
-targets:
-192.168.211.2:8090

其中，我的springboot服务部署在192.168.211.2上，而prometheus和grafana都部署在192.168.211.101上

配置了两个服务

执行以下命令启动./prometheus--config.file=prometheus.yml&

访问http://192.168.211.101:9090/targets

可以看到

5.grafana可视化监控多个prometheustargets

下载grafana-6.2.5.linux-amd64.tar.gz

启动grafana

./grafana-server&

http://192.168.211.101:3000

在dashboard中配置时，要采用这样的方式

jvm_gc_pause_seconds_count{action="endofminorGC",cause="MetadataGCThreshold",instance="192.168.211.2:8090",job="search"}

通过instance和job来进行区分prometheus的targets

可以看到dashboard中有两个prometheus的targets

以上为个人经验，希望能给大家一个参考，也希望大家多多支持毛票票。如有错误或未考虑完全的地方，望不吝赐教。

声明：本文内容来源于网络，版权归原作者所有，内容由互联网用户自发贡献自行上传，本网站不拥有所有权，未作人工编辑处理，也不承担相关法律责任。如果您发现有涉嫌版权的内容，欢迎发送邮件至：czq8825#qq.com（发邮件时，请将#更换为@）进行举报，并提供相关证据，一经查实，本站将立刻删除涉嫌侵权内容。