目录
- 1.创建项目
- 2.自定义指标
- 方式一
- 方式二
- 3. 测试
- 4.项目中的应用
1.创建项目
pom.xml引入相关依赖
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<groupId>com.olive</groupId>
<artifactId>prometheus-meter-demo</artifactId>
<version>0.0.1-SNAPSHOT</version>
<parent>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-parent</artifactId>
<version>2.3.7.RELEASE</version>
<relativePath />
</parent>
<properties>
<java.version>1.8</java.version>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
<spring-boot.version>2.3.7.RELEASE</spring-boot.version>
</properties>
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-aop</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-actuator</artifactId>
</dependency>
<!-- Micrometer Prometheus registry -->
<dependency>
<groupId>io.micrometer</groupId>
<artifactId>micrometer-registry-prometheus</artifactId>
</dependency>
</dependencies>
<dependencyManagement>
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-dependencies</artifactId>
<version>${spring-boot.version}</version>
<type>pom</type>
<scope>import</scope>
</dependency>
</dependencies>
</dependencyManagement>
</project>
2.自定义指标
方式一
直接使用micrometer
核心包的类进行指标定义和注册
package com.olive.monitor;
import javax.annotation.PostConstruct;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import io.micrometer.core.instrument.Counter;
import io.micrometer.core.instrument.DistributionSummary;
import io.micrometer.core.instrument.MeterRegistry;
@Component
public class NativeMetricsMontior {
/**
* 支付次数
*/
private Counter payCount;
/**
* 支付金额统计
*/
private DistributionSummary payAmountSum;
@Autowired
private MeterRegistry registry;
@PostConstruct
private void init() {
payCount = registry.counter("pay_request_count", "payCount", "pay-count");
payAmountSum = registry.summary("pay_amount_sum", "payAmountSum", "pay-amount-sum");
}
public Counter getPayCount() {
return payCount;
}
public DistributionSummary getPayAmountSum() {
return payAmountSum;
}
}
方式二
通过引入micrometer-registry-prometheus
包,该包结合prometheus,对micrometer进行了封装
<dependency>
<groupId>io.micrometer</groupId>
<artifactId>micrometer-registry-prometheus</artifactId>
</dependency>
同样定义两个metrics
package com.olive.monitor;
import javax.annotation.PostConstruct;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import io.prometheus.client.CollectorRegistry;
import io.prometheus.client.Counter;
@Component
public class PrometheusMetricsMonitor {
/**
* 订单发起次数
*/
private Counter orderCount;
/**
* 金额统计
*/
private Counter orderAmountSum;
@Autowired
private CollectorRegistry registry;
@PostConstruct
private void init() {
orderCount = Counter.build().name("order_request_count")
.help("order request count.")
.labelNames("orderCount")
.register();
orderAmountSum = Counter.build().name("order_amount_sum")
.help("order amount sum.")
.labelNames("orderAmountSum")
.register();
registry.register(orderCount);
registry.register(orderAmountSum);
}
public Counter getOrderCount() {
return orderCount;
}
public Counter getOrderAmountSum() {
return orderAmountSum;
}
}
prometheus 4种常用Metrics
Counter
连续增加不会减少的计数器,可以用于记录只增不减的类型,例如:网站访问人数,系统运行时间等。
对于Counter类型的指标,只包含一个inc()的方法,就是用于计数器+1.
一般而言,Counter类型的metric指标在冥冥中我们使用_total结束,如http_requests_total.
Gauge
可增可减的仪表盘,曲线图
对于这类可增可减的指标,用于反应应用的当前状态。
例如在监控主机时,主机当前空闲的内存大小,可用内存大小等等。
对于Gauge指标的对象则包含两个主要的方法inc()和dec(),用于增加和减少计数。
Histogram
主要用来统计数据的分布情况,这是一种特殊的metrics数据类型,代表的是一种近似的百分比估算数值,统计所有离散的指标数据在各个取值区段内的次数。例如:我们想统计一段时间内http请求响应小于0.005秒、小于0.01秒、小于0.025秒的数据分布情况。那么使用Histogram采集每一次http请求的时间,同时设置bucket。
Summary
Summary和Histogram非常相似,都可以统计事件发生的次数或者大小,以及其分布情况,他们都提供了对时间的计数_count以及值的汇总_sum,也都提供了可以计算统计样本分布情况的功能,不同之处在于Histogram可以通过histogram_quantile函数在服务器计算分位数。而Sumamry的分位数则是直接在客户端进行定义的。因此对于分位数的计算,Summary在通过PromQL进行查询的时候有更好的性能表现,而Histogram则会消耗更多的资源,但是相对于客户端而言Histogram消耗的资源就更少。用哪个都行,根据实际场景自由调整即可。
3. 测试
定义两个controller分别使用NativeMetricsMontior
和PrometheusMetricsMonitor
package com.olive.controller;
import java.util.Random;
import javax.annotation.Resource;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;
import com.olive.monitor.NativeMetricsMontior;
@RestController
public class PayController {
@Resource
private NativeMetricsMontior monitor;
@RequestMapping("/pay")
public String pay(@RequestParam("amount") Double amount) throws Exception {
// 统计支付次数
monitor.getPayCount().increment();
Random random = new Random();
//int amount = random.nextInt(100);
if(amount==null) {
amount = 0.0;
}
// 统计支付总金额
monitor.getPayAmountSum().record(amount);
return "支付成功, 支付金额: " + amount;
}
}
package com.olive.controller;
import java.util.Random;
import javax.annotation.Resource;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;
import com.olive.monitor.PrometheusMetricsMonitor;
@RestController
public class OrderController {
@Resource
private PrometheusMetricsMonitor monitor;
@RequestMapping("/order")
public String order(@RequestParam("amount") Double amount) throws Exception {
// 订单总数
monitor.getOrderCount()
.labels("orderCount")
.inc();
Random random = new Random();
//int amount = random.nextInt(100);
if(amount==null) {
amount = 0.0;
}
// 统计订单总金额
monitor.getOrderAmountSum()
.labels("orderAmountSum")
.inc(amount);
return "下单成功, 订单金额: " + amount;
}
}
启动服务
访问http://127.0.0.1:9595/actuator/prometheus
;正常看到监测数据
改变amount多次方式http://127.0.0.1:8080/order?amount=100
和http://127.0.0.1:8080/pay?amount=10
后;再访问http://127.0.0.1:9595/actuator/prometheus
。查看监控数据
4.项目中的应用
项目中按照上面说的方式进行数据埋点监控不太现实;在spring项目中基本通过AOP进行埋点监测。比如写一个切面Aspect
;这样的方式就非常友好。能在入口就做了数据埋点监测,无须在controller里进行代码编写。
package com.olive.aspect;
import java.time.LocalDate;
import java.util.concurrent.TimeUnit;
import javax.servlet.http.HttpServletRequest;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Pointcut;
import org.springframework.stereotype.Component;
import org.springframework.util.StringUtils;
import org.springframework.web.context.request.RequestContextHolder;
import org.springframework.web.context.request.ServletRequestAttributes;
import io.micrometer.core.instrument.Metrics;
@Aspect
@Component
public class PrometheusMetricsAspect {
// 切入所有controller包下的请求方法
@Pointcut("execution(* com.olive.controller..*.*(..))")
public void controllerPointcut() {
}
@Around("controllerPointcut()")
public Object MetricsCollector(ProceedingJoinPoint joinPoint) throws Throwable {
HttpServletRequest request = ((ServletRequestAttributes) RequestContextHolder.getRequestAttributes()).getRequest();
String userId = StringUtils.hasText(request.getParameter("userId")) ?
request.getParameter("userId") : "no userId";
// 获取api url
String api = request.getServletPath();
// 获取请求方法
String method = request.getMethod();
long startTs = System.currentTimeMillis();
LocalDate now = LocalDate.now();
String[] tags = new String[10];
tags[0] = "api";
tags[1] = api;
tags[2] = "method";
tags[3] = method;
tags[4] = "day";
tags[5] = now.toString();
tags[6] = "userId";
tags[7] = userId;
String amount = StringUtils.hasText(request.getParameter("amount")) ?
request.getParameter("amount") : "0.0";
tags[8] = "amount";
tags[9] = amount;
// 请求次数加1
//自定义的指标名称:custom_http_request_all,指标包含数据
Metrics.counter("custom_http_request_all", tags).increment();
Object object = null;
try {
object = joinPoint.proceed();
} catch (Exception e) {
//请求失败次数加1
Metrics.counter("custom_http_request_error", tags).increment();
throw e;
} finally {
long endTs = System.currentTimeMillis() - startTs;
//记录请求响应时间
Metrics.timer("custom_http_request_time", tags).record(endTs, TimeUnit.MILLISECONDS);
}
return object;
}
}
编写好切面后,重启服务;访问controller的接口,同样可以进行自定义监控指标埋点