Vgbhfive's Blog

SpringBoot集成Prometheus-自定义指标

Posted on Edited on

引入

这里首先说以下整体大概的思路，第一步当然是引入对应的 SDK，第二步则是添加配置信息、定义自定义指标，并进行注册，接下来的第三步则是指标根据具体业务的处理逻辑，那么最后一步就是在 prometheus 服务中增加 job 配置，最终在 grafana 中展示自定义指标。

实操

引入 SDK

1
2
3
4
5
6
7
8
9
10
11
<!-- Micrometer Prometheus registry  -->
<dependency>
	<groupId>io.micrometer</groupId>
	<artifactId>micrometer-registry-prometheus</artifactId>
</dependency>
<!-- https://mvnrepository.com/artifact/io.micrometer/micrometer-registry-prometheus -->
<dependency>
    <groupId>io.micrometer</groupId>
    <artifactId>micrometer-registry-prometheus</artifactId>
    <version>1.9.0</version>
</dependency>

配置信息

1
2
3
4
5
6
management.endpoints.enabled-by-default=false
management.endpoint.health.enabled=true
management.endpoint.metrics.enabled=true
management.endpoint.prometheus.enabled=true
management.endpoints.web.exposure.include=health,prometheus,metrics
management.metrics.tags.application=${spring.application.name}

注册自定义指标

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
@Component
public class ContentListener implements ApplicationListener<ContextRefreshedEvent> {

	private Logger logger = LoggerFactory.getLogger(ContentListener.class);

	private Counter counter;

	@Override
	public void onApplicationEvent(ContextRefreshedEvent event) {
		logger.info("start register prometheus counter");
		CollectorRegistry collectorRegistry = event.getApplicationContext().getBean(CollectorRegistry.class);
		counter = Counter.build().name("xx_counter").labelNames("label1", "label2").help("xx counter help").register(collectorRegistry);
		logger.info("end register prometheus counter");
   	}

   	private Counter getCounter() {
   		return this.counter;
   	}
}

指标业务逻辑

  1. 自定义指标注解

    1
    2
    3
    4
    5
    @Documented
    @Retention(RUNTIME)
    @Target(METHOD)
    public @interface Counter {
    }

  2. 注解实现

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    @Aspect
    @Component
    public class CounterAspect {

        private static final Logger logger = LoggerFactory.getLogger(CounterAspect.class);

        @Autowired
        private ContentListener contentListener;

        @Pointcut("execution(* com.example.controller.*.*(..)) && @annotation(com.example.annotation.Counter)")
        private void cut() { }

        @Around("cut()")
        public Object advice(ProceedingJoinPoint pjp) throws Throwable {
            Object retValue = pjp.proceed();

            try {
                // 记录结果指标
                contentListener.getCounter().labels("label1", "label2").inc();
            } catch (Exception e) {
                logger.error("error", e);
            }
            return retValue;
        }
    }

Prometheus Server 中增加 Job 配置

1
2
3
4
5
6
job_name: 'springboot-demo'
  # metrics_path defaults to '/metrics'
  # scheme defaults to 'http'.
  metrics_path: /actuator/prometheus
  static_configs:
    targets: ['localhost:8080']

接着重新启动 prometheus

1
./prometheus –config.file=prometheus.yml

grafana 展示自定义指标

  1. 检查 curl http://localhost:8080/actuator/prometheus 是否存在自定义指标。
  2. 在 grafana 中配置 prometheus server 数据源。
  3. 在面板中配置对应的指标和需要展示的图例。

总结

个人备注

此博客内容均为作者学习所做笔记，侵删！
若转作其他用途，请注明来源！