準(zhǔn)備工作
- Prometheus
v2.14.0 - Grafana
v6.5.0
使用 Docker 部署上面的基礎(chǔ)環(huán)境，配置文件可見 docker-compose.yml，輸入以下命令進(jìn)行一鍵部署：
# 檢查配置
docker-compose config
# 啟動(dòng)服務(wù)（-d 后臺(tái)啟動(dòng)）
docker-compose up -d
# 停止并清除服務(wù)
docker-compose down
添加依賴項(xiàng)
pom.xml
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-actuator</artifactId>
</dependency>
<!-- 借助 Micrometer 對(duì)接 Prometheus 監(jiān)控系統(tǒng) -->
<dependency>
<groupId>io.micrometer</groupId>
<artifactId>micrometer-registry-prometheus</artifactId>
</dependency>
</dependencies>
application.yml
spring:
application:
name: monitoring-prometheus-grafana
management:
endpoints:
web:
exposure:
# 將 Actuator 的 /actuator/prometheus 端點(diǎn)暴露出來
include: 'prometheus'
metrics:
tags:
application: ${spring.application.name}
logging:
level:
io.github.y0ngb1n.*: debug
使用方式
Step 1: 啟動(dòng)服務(wù)
$ mvn spring-boot:run
...
2019-12-08 22:28:11.916 INFO 36157 --- [ main] o.s.b.a.e.web.EndpointLinksResolver : Exposing 1 endpoint(s) beneath base path '/actuator'
2019-12-08 22:28:12.045 INFO 36157 --- [ main] o.s.b.w.embedded.tomcat.TomcatWebServer : Tomcat started on port(s): 8080 (http) with context path ''
2019-12-08 22:28:12.050 INFO 36157 --- [ main] i.g.y.m.p.PrometheusGrafanaApplication : Started PrometheusGrafanaApplication in 20.638 seconds (JVM running for 27.154)
Step 2: 訪問 /actuator/prometheus 端口
$ curl -sS http://127.0.0.1:8080/actuator/prometheus
# HELP jvm_gc_memory_promoted_bytes_total Count of positive increases in the size of the old generation memory pool before GC to after GC
# TYPE jvm_gc_memory_promoted_bytes_total counter
jvm_gc_memory_promoted_bytes_total{application="monitoring-prometheus-grafana",} 9986992.0
# HELP jvm_threads_daemon_threads The current number of live daemon threads
# TYPE jvm_threads_daemon_threads gauge
jvm_threads_daemon_threads{application="monitoring-prometheus-grafana",} 30.0
# HELP process_uptime_seconds The uptime of the Java virtual machine
# TYPE process_uptime_seconds gauge
process_uptime_seconds{application="monitoring-prometheus-grafana",} 75.133
...
將應(yīng)用與 Prometheus 對(duì)接
添加 Prometheus 的配置文件：prometheus.yml
scrape_configs:
# 任意寫，建議英文，不要包含特殊字符
- job_name: 'spring-boot-app'
# 多久采集一次數(shù)據(jù)
scrape_interval: 15s
# 采集時(shí)的超時(shí)時(shí)間
scrape_timeout: 10s
# 采集的路徑是啥
metrics_path: '/actuator/prometheus'
# 采集服務(wù)的地址，設(shè)置成 Spring Boot 應(yīng)用所在服務(wù)器的具體地址
static_configs:
- targets: ['192.168.31.44:8080']
可使用 ifconfig 查看宿主機(jī)的 IP 地址，Prometheus 服務(wù)器就會(huì)自動(dòng)每隔15秒請(qǐng)求 http://your-ip:8080/actuator/prometheus。更多配置參見 ??官方文檔
測(cè)試訪問 Prometheus
STEP 1：在瀏覽器中訪問 http://localhost:9090
STEP 2：查看監(jiān)控?cái)?shù)據(jù)
- 點(diǎn)擊
Insert metric at cursor，即可選擇監(jiān)控指標(biāo)； - 點(diǎn)擊
Graph，即可讓指標(biāo)以圖表方式展示； - 點(diǎn)擊
Execute按鈕，即可看到類似下圖的結(jié)果。
集成 Grafana 可視化
前面已經(jīng)使用 docker-compose 啟動(dòng)了 Grafana，下面進(jìn)行配置。
STEP 1：在瀏覽器中訪問 http://localhost:3000，并使用初始帳號(hào) admin:admin 進(jìn)行登錄
STEP 2：點(diǎn)擊 Add Data Source 進(jìn)行添加 Prometheus 數(shù)據(jù)源
STEP 3：創(chuàng)建 Dashboard
- 設(shè)置監(jiān)控指標(biāo)
- 設(shè)置 Dashboard 信息
- 保存 Dashboard
可以添加不同的指標(biāo)，指標(biāo)的取值詳見 Spring Boot 應(yīng)用的 /actuator/prometheus 端點(diǎn)。重復(fù)以上步驟操作，即可為 Dashboard 添加新的圖表。
如果你覺得自己定制一個(gè) Dashboard 過于復(fù)雜了，那你可以逛一下 Grafana 的 Dashboard 市場(chǎng)（https://grafana.com/grafana/dashboards），這樣就可以方便快捷地復(fù)用別人精美的 Dashboard 面板了。
參考資料
- https://www.callicoder.com/spring-boot-actuator-metrics-monitoring-dashboard-prometheus-grafana/
- http://www.itmuch.com/spring-boot/actuator-prometheus-grafana/，by 周立
- https://www.baeldung.com/spring-boot-self-hosted-monitoring
- https://yunlzheng.gitbook.io/prometheus-book/
- https://micrometer.io/docs/registry/prometheus
- https://prometheus.io/docs/introduction/overview/
- https://hub.docker.com/r/prom/prometheus/
- https://grafana.com/docs/grafana/latest/
感謝您的閱讀，本文由 楊斌的博客 版權(quán)所有。
如若轉(zhuǎn)載，請(qǐng)注明出處：楊斌的博客（https://y0ngb1n.github.io?utm_source=jianshu）
項(xiàng)目已托管于 GitHub：y0ngb1n/spring-boot-samples，歡迎 Star, Fork :kissing_heart:
