prometheus应用demo（一）接口监控

目录

完整代码（纯Cursor生成）

1、pom

2、配置和启动类

 3、自定义指标bean

4、上报

5、业务代码

一、统计API请求（次数、响应码等）

1、统计总数

关键代码：

（1）自定义指标DTO

（2）filter拦截HTTP请求并上报数据 

后台监控查看

结合Grafana

2、监控峰值

关键代码

（1）计算

（2）上报 

 测试

二、统计Repository接口的入参出参数量

1、简单统计

关键代码

（1）bean定义

（2）AOP 

测试

​编辑

​编辑

2、监控峰值

关键代码

（1）bean定义

（2）APO上报

测试

完整代码（纯Cursor生成）

1、pom
<?xml version="1.0" encoding="UTF-8"?>
<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 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><groupId>com.demo</groupId><artifactId>prometheus-demo</artifactId><version>1.0-SNAPSHOT</version><properties><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target><project.build.sourceEncoding>UTF-8</project.build.sourceEncoding></properties><parent><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-parent</artifactId><version>3.2.4</version><relativePath/></parent><dependencies><!-- Web --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><!-- Actuator --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-actuator</artifactId></dependency><!-- Prometheus Registry --><dependency><groupId>io.micrometer</groupId><artifactId>micrometer-registry-prometheus</artifactId></dependency><dependency><groupId>org.projectlombok</groupId><artifactId>lombok</artifactId></dependency><!-- AOP --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-aop</artifactId></dependency><!--尽量不要同时导入mybatis 和 mybatis_plus，避免版本差异--><dependency><groupId>com.baomidou</groupId><artifactId>mybatis-plus-spring-boot3-starter</artifactId><version>3.5.5</version></dependency><dependency><groupId>mysql</groupId><artifactId>mysql-connector-java</artifactId><version>8.0.13</version></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-data-redis</artifactId></dependency></dependencies><build><plugins><plugin><groupId>org.springframework.boot</groupId><artifactId>spring-boot-maven-plugin</artifactId></plugin></plugins></build></project>
2、配置和启动类
server.port=3333
server.servlet.context-path=/prometheusDemomanagement.endpoints.web.exposure.include=health,info,prometheus
management.endpoint.prometheus.enabled=true#mysql
spring.datasource.driver-class-name=com.mysql.cj.jdbc.Driver
spring.datasource.url=jdbc:mysql://localhost:3308/demo?useUnicode=true&characterEncoding=utf-8&useSSL=false&allowMultiQueries=true
spring.datasource.username=root
spring.datasource.password=wtyy
#mybatis
mybatis.mapper-locations=classpath*:mapper/*Mapper.xml
#??SQL
mybatis-plus.configuration.log-impl=org.apache.ibatis.logging.stdout.StdOutImpl

@SpringBootApplication
public class PrometheusApplication {public static void main(String[] args) {SpringApplication.run(PrometheusApplication.class, args);}
}
 3、自定义指标bean

（）API的

package com.demo.metrics;import lombok.extern.slf4j.Slf4j;import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;/*** 时间窗口指标统计（仅统计当前窗口调用量）*/
@Slf4j
public class TimeWindowMetrics {// 时间窗口大小（秒）private final int windowSizeSeconds;// 存储每个时间窗口的统计数据private final ConcurrentHashMap<String, WindowStats> windowStatsMap = new ConcurrentHashMap<>();public TimeWindowMetrics(int windowSizeSeconds) {this.windowSizeSeconds = windowSizeSeconds;}/*** 记录一次API调用*/public void recordApiCall(String uri, String method, int status) {String currentWindow = getCurrentTimeWindow();String key = uri + "#" + method + "#" + status;String windowKey = currentWindow + "#" + key;// 记录当前时间窗口的统计WindowStats windowStats = windowStatsMap.computeIfAbsent(windowKey, k -> new WindowStats());int currentCount = windowStats.increment();// 清理过期的时间窗口数据cleanExpiredWindows();log.debug("📊 时间窗口统计: {} 在窗口 {} 中调用次数: {}", key, currentWindow, currentCount);}/*** 获取指定URI的当前时间窗口调用量*/public int getCurrentWindowCount(String uri, String method, int status) {String currentWindow = getCurrentTimeWindow();String key = uri + "#" + method + "#" + status;String windowKey = currentWindow + "#" + key;WindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getCount() : 0;}/*** 获取当前时间窗口标识*/private String getCurrentTimeWindow() {LocalDateTime now = LocalDateTime.now();// 根据窗口大小计算时间窗口int windowIndex = now.getSecond() / windowSizeSeconds;return now.format(DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:")) + String.format("%02d", windowIndex * windowSizeSeconds);}/*** 清理过期的时间窗口数据*/private void cleanExpiredWindows() {if (windowStatsMap.size() > 1000) { // 避免内存泄漏String currentWindow = getCurrentTimeWindow();windowStatsMap.entrySet().removeIf(entry -> !entry.getKey().contains(currentWindow));log.debug("🧹 清理过期时间窗口数据，当前窗口: {}", currentWindow);}}/*** 时间窗口内的统计数据*/private static class WindowStats {private final AtomicInteger count = new AtomicInteger(0);public int increment() {return count.incrementAndGet();}public int getCount() {return count.get();}}
}

（2）Repo的

package com.demo.metrics;import lombok.extern.slf4j.Slf4j;import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;/*** Repository层时间窗口指标统计*/
@Slf4j
public class RepositoryTimeWindowMetrics {// 时间窗口大小（秒）private final int windowSizeSeconds;// 存储每个时间窗口的Repository统计数据private final ConcurrentHashMap<String, RepositoryWindowStats> windowStatsMap = new ConcurrentHashMap<>();public RepositoryTimeWindowMetrics(int windowSizeSeconds) {this.windowSizeSeconds = windowSizeSeconds;}/*** 记录一次Repository方法调用*/public void recordMethodCall(String className, String methodName, int listElementCount) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;// 记录当前时间窗口的统计RepositoryWindowStats windowStats = windowStatsMap.computeIfAbsent(windowKey, k -> new RepositoryWindowStats());windowStats.record(listElementCount);// 清理过期的时间窗口数据cleanExpiredWindows();log.debug("📊 Repository时间窗口统计: {} 在窗口 {} 中调用次数: {}, 元素总数: {}, 最大值: {}", key, currentWindow, windowStats.getCallCount(), windowStats.getTotalElements(), windowStats.getMaxElements());}/*** 获取指定方法的当前时间窗口调用次数*/public int getCurrentWindowCallCount(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getCallCount() : 0;}/*** 获取指定方法的当前时间窗口元素总数*/public int getCurrentWindowTotalElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getTotalElements() : 0;}/*** 获取指定方法的当前时间窗口最大元素个数*/public int getCurrentWindowMaxElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getMaxElements() : 0;}/*** 获取指定方法的当前时间窗口平均元素个数*/public double getCurrentWindowAvgElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);if (stats != null && stats.getCallCount() > 0) {return (double) stats.getTotalElements() / stats.getCallCount();}return 0.0;}/*** 获取当前时间窗口标识*/private String getCurrentTimeWindow() {LocalDateTime now = LocalDateTime.now();// 根据窗口大小计算时间窗口int windowIndex = now.getSecond() / windowSizeSeconds;return now.format(DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:")) + String.format("%02d", windowIndex * windowSizeSeconds);}/*** 清理过期的时间窗口数据*/private void cleanExpiredWindows() {if (windowStatsMap.size() > 1000) { // 避免内存泄漏String currentWindow = getCurrentTimeWindow();windowStatsMap.entrySet().removeIf(entry -> !entry.getKey().contains(currentWindow));log.debug("🧹 清理过期Repository时间窗口数据，当前窗口: {}", currentWindow);}}/*** Repository时间窗口内的统计数据*/private static class RepositoryWindowStats {private final AtomicInteger callCount = new AtomicInteger(0);private final AtomicInteger totalElements = new AtomicInteger(0);private final AtomicInteger maxElements = new AtomicInteger(0);public void record(int elementCount) {callCount.incrementAndGet();totalElements.addAndGet(elementCount);// 更新最大值maxElements.updateAndGet(current -> Math.max(current, elementCount));}public int getCallCount() {return callCount.get();}public int getTotalElements() {return totalElements.get();}public int getMaxElements() {return maxElements.get();}}
}

4、上报

(1)API的

package com.demo.filter;import com.demo.metrics.TimeWindowMetrics;
import io.micrometer.core.instrument.Gauge;
import io.micrometer.core.instrument.MeterRegistry;
import jakarta.servlet.FilterChain;
import jakarta.servlet.ServletException;
import jakarta.servlet.http.HttpServletRequest;
import jakarta.servlet.http.HttpServletResponse;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
import org.springframework.web.filter.OncePerRequestFilter;import java.io.IOException;@Component
@Slf4j
@Order(1)
public class UrlMetricsFilter extends OncePerRequestFilter {@Autowiredprivate MeterRegistry meterRegistry;// 时间窗口统计（60秒窗口）private final TimeWindowMetrics timeWindowMetrics = new TimeWindowMetrics(60);@Overrideprotected void doFilterInternal(HttpServletRequest request, HttpServletResponse response, FilterChain filterChain) throws ServletException, IOException {// 过滤掉 actuator 相关的监控端点if (shouldSkipFiltering(request)) {log.debug("忽略监控: {}", request.getRequestURI());filterChain.doFilter(request, response);return;}log.debug("进入UrlMetricsFilter - API: {} {}", request.getMethod(), request.getRequestURI());try {// 执行请求filterChain.doFilter(request, response);// 记录到时间窗口统计timeWindowMetrics.recordApiCall(request.getRequestURI(), request.getMethod(), response.getStatus());// 注册当前窗口指标registerCurrentWindowMetric(request.getRequestURI(), request.getMethod(), response.getStatus());} catch (Exception e) {log.error("UrlMetricsFilter处理异常", e);throw e;}}/*** 判断是否应该跳过过滤* @param request HTTP请求* @return true表示跳过，false表示需要过滤*/private boolean shouldSkipFiltering(HttpServletRequest request) {String requestURI = request.getRequestURI();// 过滤掉 actuator 相关的监控端点if (requestURI.contains("/actuator/")) {log.debug("跳过监控端点: {}", requestURI);return true;}// 可以添加更多需要跳过的路径// 例如：静态资源、健康检查等String[] skipPaths = {"/favicon.ico","/static/","/css/","/js/","/images/"};for (String skipPath : skipPaths) {if (requestURI.contains(skipPath)) {log.debug("跳过静态资源: {}", requestURI);return true;}}return false;}/*** 注册当前时间窗口监控指标*/private void registerCurrentWindowMetric(String uri, String method, int status) {String metricName = "api_requests_current_window";// 检查指标是否已存在if (meterRegistry.find(metricName).tag("uri", uri).tag("method", method)  .tag("status", String.valueOf(status)).gauge() == null) {try {Gauge.builder(metricName, () -> (double) timeWindowMetrics.getCurrentWindowCount(uri, method, status)).description("Current time window API request count").tag("uri", uri).tag("method", method).tag("status", String.valueOf(status)).register(meterRegistry);log.debug("✅ 注册当前窗口指标: {}#{} status:{}", uri, method, status);} catch (Exception e) {log.debug("当前窗口指标已存在: {}#{} status:{}", uri, method, status);}}}
}

（2）Repo的AOP

package com.demo.aop;import io.micrometer.core.instrument.Gauge;
import io.micrometer.core.instrument.MeterRegistry;
import lombok.extern.slf4j.Slf4j;
import org.aspectj.lang.JoinPoint;
import org.aspectj.lang.annotation.After;
import org.aspectj.lang.annotation.Aspect;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;import com.demo.metrics.RepositoryTimeWindowMetrics;import java.util.Collection;@Aspect
@Component
@Slf4j
public class RepositoryMonitoringAspect {@Autowiredprivate MeterRegistry meterRegistry;// Repository时间窗口统计（60秒窗口）private final RepositoryTimeWindowMetrics timeWindowMetrics = new RepositoryTimeWindowMetrics(60);// ========== 注释掉原有累计统计功能 ==========// 使用ConcurrentHashMap存储每个方法的统计Bean// private final ConcurrentHashMap<String, RepositoryMetricsBean> methodMetricsBeans = new ConcurrentHashMap<>();/** * Repository时间窗口统计指标:* - repository_method_calls_current_window (当前窗口调用量)* - repository_method_list_elements_max_current_window (当前窗口最大值)* - repository_method_list_elements_avg_current_window (当前窗口平均值)* - repository_method_list_elements_sum_current_window (当前窗口元素总数)** @param joinPoint*/@After("execution(* com.demo.repository.*.*(..))")public void monitorRepositoryMethod(JoinPoint joinPoint) {String className = joinPoint.getTarget().getClass().getSimpleName();String methodName = joinPoint.getSignature().getName();log.info("🔍 AOP拦截到Repository方法: {}#{}", className, methodName);// 获取List参数中的元素个数Object[] args = joinPoint.getArgs();int listElementCount = extractListElementCount(args);// ========== 注释掉原有累计统计功能 ==========/*// 1. 统计调用量 - 先查找已存在的，不存在才创建Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();if (callsCounter == null) {try {callsCounter = Counter.builder("repository_method_calls_total").description("Repository method call count").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();}}if (callsCounter != null) {callsCounter.increment();}// 2. 记录List元素总数 - 先查找已存在的，不存在才创建Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (sumCounter == null) {try {sumCounter = Counter.builder("repository_method_list_elements_sum").description("Total number of list elements").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();}}if (sumCounter != null) {sumCounter.increment(listElementCount);}// 3. 记录List元素统计 - 使用独立的MetricsBean处理recordListElementMetrics(className, methodName, listElementCount);// 4. 从Prometheus获取数据并计算平均值，repository_method_list_elements_avgcalculateAndReportAverage(className, methodName);*/// ========== 时间窗口统计 (保留功能) ==========// 记录到时间窗口统计timeWindowMetrics.recordMethodCall(className, methodName, listElementCount);// 注册时间窗口监控指标registerTimeWindowMetrics(className, methodName);log.info("📊 Repository时间窗口监控统计: {}#{} 调用，List元素个数: {}", className, methodName, listElementCount);}/*** 提取方法参数中List的元素个数* @param args 方法参数数组* @return List中元素的总个数*/private int extractListElementCount(Object[] args) {if (args == null || args.length == 0) {return 0;}int totalElements = 0;for (Object arg : args) {if (arg instanceof Collection) {Collection<?> collection = (Collection<?>) arg;totalElements += collection.size();log.debug("发现Collection参数，元素个数: {}", collection.size());}}return totalElements;}// ========== 注释掉原有累计统计相关方法 ==========/*/*** 记录List元素统计 - 使用独立的MetricsBean处理* @param className 类名* @param methodName 方法名  * @param listElementCount 当前List元素个数*//*private void recordListElementMetrics(String className, String methodName, int listElementCount) {// 1. 查找或创建对应方法的MetricsBeanRepositoryMetricsBean metricsBean = findOrCreateMetricsBean(className, methodName);// 2. 记录当前调用数据metricsBean.record(listElementCount);// 3. 聚合计算平均值metricsBean.aggregate();// 4. 确保Prometheus指标已注册ensurePrometheusMetricsRegistered(metricsBean);}/*** 查找或创建方法的MetricsBean*//*private RepositoryMetricsBean findOrCreateMetricsBean(String className, String methodName) {String methodKey = className + "#" + methodName;return methodMetricsBeans.computeIfAbsent(methodKey, key -> {log.debug("✅ 创建新的MetricsBean: {}", key);return new RepositoryMetricsBean(className, methodName);});}/*** 确保Prometheus指标已注册*//*private void ensurePrometheusMetricsRegistered(RepositoryMetricsBean metricsBean) {String className = metricsBean.getClassName();String methodName = metricsBean.getMethodName();// 注册最大值GaugeGauge maxGauge = meterRegistry.find("repository_method_list_elements_max").tag("class", className).tag("method", methodName).gauge();if (maxGauge == null) {try {Gauge.builder("repository_method_list_elements_max", () -> (double) metricsBean.getMaxElements()).description("Maximum number of list elements from MetricsBean").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册最大值Gauge: {}#{}", className, methodName);} catch (IllegalArgumentException e) {log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}}/*** 从Prometheus获取指标数据，计算平均值并上报* @param className 类名* @param methodName 方法名*//*private void calculateAndReportAverage(String className, String methodName) {// 使用安全的获取或创建方式try {// 捕获当前的className和methodName，避免闭包问题final String finalClassName = className;final String finalMethodName = methodName;Gauge.builder("repository_method_list_elements_avg", () -> calculateAverage(finalClassName, finalMethodName)).description("Average number of list elements (calculated from Prometheus data)").tag("class", finalClassName).tag("method", finalMethodName).register(meterRegistry);log.debug("🎯 注册平均值Gauge: {}#{}", finalClassName, finalMethodName);} catch (IllegalArgumentException e) {// 如果已存在，什么都不做，Gauge会继续工作log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}/*** 从MeterRegistry中获取指标值并计算平均值* @param className 类名* @param methodName 方法名* @return 平均值*//*private double calculateAverage(String className, String methodName) {try {// 从MeterRegistry获取调用次数Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();// 从MeterRegistry获取元素总数Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (callsCounter != null && sumCounter != null) {double calls = callsCounter.count();double sum = sumCounter.count();if (calls > 0) {double average = sum / calls;log.debug("📊 计算平均值: {}#{} = {} / {} = {}", className, methodName, sum, calls, average);return average;}}log.debug("📊 暂无足够数据计算平均值: {}#{}", className, methodName);return 0.0;} catch (Exception e) {log.error("❌ 计算平均值失败: {}#{}, 错误: {}", className, methodName, e.getMessage());return 0.0;}}*//*** 注册时间窗口监控指标*/private void registerTimeWindowMetrics(String className, String methodName) {// 1. 当前时间窗口调用量String callsMetricName = "repository_method_calls_current_window";if (meterRegistry.find(callsMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(callsMetricName, () -> (double) timeWindowMetrics.getCurrentWindowCallCount(className, methodName)).description("Current time window repository method call count").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口调用量指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口调用量指标已存在: {}#{}", className, methodName);}}// 2. 当前时间窗口元素总数String sumMetricName = "repository_method_list_elements_sum_current_window";if (meterRegistry.find(sumMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(sumMetricName, () -> (double) timeWindowMetrics.getCurrentWindowTotalElements(className, methodName)).description("Current time window total list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口元素总数指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口元素总数指标已存在: {}#{}", className, methodName);}}// 3. 当前时间窗口最大值String maxMetricName = "repository_method_list_elements_max_current_window";if (meterRegistry.find(maxMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(maxMetricName, () -> (double) timeWindowMetrics.getCurrentWindowMaxElements(className, methodName)).description("Current time window maximum list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口最大值指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口最大值指标已存在: {}#{}", className, methodName);}}// 4. 当前时间窗口平均值String avgMetricName = "repository_method_list_elements_avg_current_window";if (meterRegistry.find(avgMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(avgMetricName, () -> timeWindowMetrics.getCurrentWindowAvgElements(className, methodName)).description("Current time window average list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口平均值指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口平均值指标已存在: {}#{}", className, methodName);}}}
}

5、业务代码

controller--》service-->repo->mapper，随便举几个

UserController

@RestController
@RequestMapping("/user")
public class UserController {@Autowiredprivate UserService userService;@GetMapping("/queryByUserId")public User queryByUserId(@RequestParam("userId") Long userId) {return userService.queryByUserId(userId);}@PostMapping("/batchQueryByUserIds")public List<User> batchQueryByUserIdsPost(@RequestBody BatchQueryRequest request) {return userService.queryByUserIds(request.getUserIds());}
}
RoleRepository：
    public Role findByRoleId(Long roleId) {return roleMapper.selectById(roleId);}public List<Role> findByRoleIds(List<Long> roleIds) {if (roleIds == null || roleIds.isEmpty()) {return new ArrayList<>();}return roleMapper.selectBatchIds(roleIds);}

一、统计API请求（次数、响应码等）

监控API接口的调用次数、错误码等。在filter中实现。

1、统计总数

使用Cursor自动生成代码

关键代码：
（1）自定义指标DTO
package com.demo.metrics;import lombok.Data;
import java.util.Map;@Data
public class APIMetricsMetricsBean {private String url;private String httpMethod;private String apiName;private Map<String, String[]> parameters;private String responseBody;private int statusCode;private long responseTime;private String time;private String clientIp;
}
（2）filter拦截HTTP请求并上报数据 
package com.demo.filter;import com.demo.metrics.APIMetricsMetricsBean;
import com.fasterxml.jackson.databind.ObjectMapper;
import io.micrometer.core.instrument.Counter;
import io.micrometer.core.instrument.MeterRegistry;
import io.micrometer.core.instrument.Timer;
import jakarta.servlet.FilterChain;
import jakarta.servlet.ServletException;
import jakarta.servlet.http.HttpServletRequest;
import jakarta.servlet.http.HttpServletResponse;
import lombok.Data;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
import org.springframework.web.filter.OncePerRequestFilter;
import org.springframework.web.util.ContentCachingResponseWrapper;import java.io.IOException;
import java.text.SimpleDateFormat;
import java.time.LocalDateTime;
import java.util.Date;@Component
@Slf4j
@Order(1)
public class UrlMetricsFilter extends OncePerRequestFilter {@Autowiredprivate MeterRegistry meterRegistry;private final ObjectMapper objectMapper = new ObjectMapper();@Overrideprotected void doFilterInternal(HttpServletRequest request, HttpServletResponse response, FilterChain filterChain) throws ServletException, IOException {// 过滤掉 actuator 相关的监控端点if (shouldSkipFiltering(request)) {log.info("忽略上报");filterChain.doFilter(request, response);return;}log.info("进入UrlMetricsFilter - API: {} {}", request.getMethod(), request.getRequestURI());long startTime = System.currentTimeMillis();Timer.Sample sample = Timer.start(meterRegistry);// 使用ContentCachingResponseWrapper来捕获响应内容ContentCachingResponseWrapper responseWrapper = new ContentCachingResponseWrapper(response);try {// 执行请求filterChain.doFilter(request, responseWrapper);// 计算响应时间long responseTime = System.currentTimeMillis() - startTime;// 获取或注册TimerTimer durationTimer = meterRegistry.find("api_requests_duration_seconds").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).tag("status", String.valueOf(responseWrapper.getStatus())).timer();if (durationTimer == null) {durationTimer = Timer.builder("api_requests_duration_seconds").description("API request duration in seconds").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).tag("status", String.valueOf(responseWrapper.getStatus())).register(meterRegistry);}sample.stop(durationTimer);// 收集API指标数据APIMetricsMetricsBean metricsDTO = collectMetrics(request, responseWrapper, responseTime);// 上报指标reportMetrics(metricsDTO);// 记录API调用次数Counter requestCounter = meterRegistry.find("api_request_total").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).tag("status", String.valueOf(responseWrapper.getStatus())).counter();if (requestCounter == null) {requestCounter = Counter.builder("api_request_total").description("Total number of API requests").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).tag("status", String.valueOf(responseWrapper.getStatus())).register(meterRegistry);}requestCounter.increment();} catch (Exception e) {log.error("上报指标error", e);// 记录错误次数Counter errorCounter = meterRegistry.find("api_requests_errors_total").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).counter();if (errorCounter == null) {errorCounter = Counter.builder("api_requests_errors_total").description("Total number of API request errors").tag("uri", request.getRequestURI()).tag("method", request.getMethod()).register(meterRegistry);}errorCounter.increment();} finally {// 确保响应内容被写回responseWrapper.copyBodyToResponse();}}private APIMetricsMetricsBean collectMetrics(HttpServletRequest request, ContentCachingResponseWrapper response, long responseTime) {APIMetricsMetricsBean metricsDTO = new APIMetricsMetricsBean();// 基本信息metricsDTO.setUrl(request.getRequestURL().toString());metricsDTO.setHttpMethod(request.getMethod());metricsDTO.setApiName(request.getRequestURI());SimpleDateFormat simpleDateFormat = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");metricsDTO.setTime(simpleDateFormat.format(new Date()));metricsDTO.setResponseTime(responseTime);metricsDTO.setStatusCode(response.getStatus());// 客户端IPString clientIp = getClientIpAddress(request);metricsDTO.setClientIp(clientIp);// 请求参数metricsDTO.setParameters(request.getParameterMap());// 响应内容try {byte[] responseBody = response.getContentAsByteArray();if (responseBody.length > 0) {String responseContent = new String(responseBody, response.getCharacterEncoding());// 限制响应内容长度，避免过大的日志if (responseContent.length() > 1000) {responseContent = responseContent.substring(0, 1000) + "...";}metricsDTO.setResponseBody(responseContent);}} catch (Exception e) {log.warn("获取响应内容失败");metricsDTO.setResponseBody("获取响应内容失败: " + e.getMessage());}return metricsDTO;}private void reportMetrics(APIMetricsMetricsBean metricsDTO) {try {// 记录详细的API调用日志log.info("API指标上报: {}", objectMapper.writeValueAsString(metricsDTO));// 可以在这里添加其他上报逻辑，比如发送到监控系统// 例如：发送到Kafka、写入数据库、发送到外部监控服务等} catch (Exception e) {log.error("上报API指标失败", e);}}private String getClientIpAddress(HttpServletRequest request) {String clientIp = request.getHeader("X-Forwarded-For");if (clientIp == null || clientIp.isEmpty() || "unknown".equalsIgnoreCase(clientIp)) {clientIp = request.getHeader("X-Real-IP");}if (clientIp == null || clientIp.isEmpty() || "unknown".equalsIgnoreCase(clientIp)) {clientIp = request.getHeader("Proxy-Client-IP");}if (clientIp == null || clientIp.isEmpty() || "unknown".equalsIgnoreCase(clientIp)) {clientIp = request.getHeader("WL-Proxy-Client-IP");}if (clientIp == null || clientIp.isEmpty() || "unknown".equalsIgnoreCase(clientIp)) {clientIp = request.getRemoteAddr();}// 处理多个IP的情况，取第一个if (clientIp != null && clientIp.contains(",")) {clientIp = clientIp.split(",")[0].trim();}return clientIp;}/*** 判断是否应该跳过过滤* @param request HTTP请求* @return true表示跳过，false表示需要过滤*/private boolean shouldSkipFiltering(HttpServletRequest request) {String requestURI = request.getRequestURI();// 过滤掉 actuator 相关的监控端点if (requestURI.contains("/actuator/")) {log.debug("跳过监控端点: {}", requestURI);return true;}// 可以添加更多需要跳过的路径// 例如：静态资源、健康检查等String[] skipPaths = {"/favicon.ico","/static/","/css/","/js/","/images/"};for (String skipPath : skipPaths) {if (requestURI.contains(skipPath)) {log.debug("跳过静态资源: {}", requestURI);return true;}}return false;}
}

启动项目发现在自动上报

2025-08-04T11:24:15.461+08:00  INFO 33060 --- [           main] o.s.b.a.e.web.EndpointLinksResolver      : Exposing 3 endpoint(s) beneath base path '/actuator'
2025-08-04T11:24:15.500+08:00  INFO 33060 --- [           main] o.s.b.w.embedded.tomcat.TomcatWebServer  : Tomcat started on port 3333 (http) with context path '/prometheusDemo'
2025-08-04T11:24:15.507+08:00  INFO 33060 --- [           main] com.demo.PrometheusApplication           : Started PrometheusApplication in 1.362 seconds (process running for 1.908)
2025-08-04T11:24:15.755+08:00  INFO 33060 --- [-10.100.168.183] o.a.c.c.C.[.[.[/prometheusDemo]          : Initializing Spring DispatcherServlet 'dispatcherServlet'
2025-08-04T11:24:15.755+08:00  INFO 33060 --- [-10.100.168.183] o.s.web.servlet.DispatcherServlet        : Initializing Servlet 'dispatcherServlet'
2025-08-04T11:24:15.756+08:00  INFO 33060 --- [-10.100.168.183] o.s.web.servlet.DispatcherServlet        : Completed initialization in 1 ms
2025-08-04T11:24:17.861+08:00  INFO 33060 --- [nio-3333-exec-1] com.demo.filter.UrlMetricsFilter         : 忽略上报
2025-08-04T11:24:22.775+08:00  INFO 33060 --- [nio-3333-exec-2] com.demo.filter.UrlMetricsFilter         : 忽略上报
2025-08-04T11:24:27.775+08:00  INFO 33060 --- [nio-3333-exec-3] com.demo.filter.UrlMetricsFilter         : 忽略上报
后台监控查看

启动本地Prometheus ，访问 http://localhost:9090/ 进入Prometheus  UI页面。

① 输入api_request_total可以看到总览

 ② 输入sort_desc(sum by (uri) (api_request_total))可以看到按接口访问次数倒序排列情况：

③ 输入topk(2, sum by (uri) (api_request_total)) # 按URI分组，显示访问次数最多的前2个

更多预览方法参考官方说明。

结合Grafana

对于项目上报的数据，如果需要监控多个维度，可以在Prometheus UI新建多个query，但是很不直观。下面结合Grafana展示自定义报表。

2、监控峰值

上面统计的是API的调用总数，那在服务正常运行期间肯定会一直增长，对我们意义不大。如果能监控每个时间段的调用量，这样就能看出来低峰期、高峰期，预防恶意接口攻击。

这种统计本身就随着时间变化，所以前面提到的重启服务会重新计数也不用care了。

关键代码
（1）计算
package com.demo.metrics;import lombok.extern.slf4j.Slf4j;import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;/*** 时间窗口指标统计（仅统计当前窗口调用量）*/
@Slf4j
public class TimeWindowMetrics {// 时间窗口大小（秒）private final int windowSizeSeconds;// 存储每个时间窗口的统计数据private final ConcurrentHashMap<String, WindowStats> windowStatsMap = new ConcurrentHashMap<>();public TimeWindowMetrics(int windowSizeSeconds) {this.windowSizeSeconds = windowSizeSeconds;}/*** 记录一次API调用*/public void recordApiCall(String uri, String method, int status) {String currentWindow = getCurrentTimeWindow();String key = uri + "#" + method + "#" + status;String windowKey = currentWindow + "#" + key;// 记录当前时间窗口的统计WindowStats windowStats = windowStatsMap.computeIfAbsent(windowKey, k -> new WindowStats());int currentCount = windowStats.increment();// 清理过期的时间窗口数据cleanExpiredWindows();log.debug("📊 时间窗口统计: {} 在窗口 {} 中调用次数: {}", key, currentWindow, currentCount);}/*** 获取指定URI的当前时间窗口调用量*/public int getCurrentWindowCount(String uri, String method, int status) {String currentWindow = getCurrentTimeWindow();String key = uri + "#" + method + "#" + status;String windowKey = currentWindow + "#" + key;WindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getCount() : 0;}/*** 获取当前时间窗口标识*/private String getCurrentTimeWindow() {LocalDateTime now = LocalDateTime.now();// 根据窗口大小计算时间窗口int windowIndex = now.getSecond() / windowSizeSeconds;return now.format(DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:")) + String.format("%02d", windowIndex * windowSizeSeconds);}/*** 清理过期的时间窗口数据*/private void cleanExpiredWindows() {if (windowStatsMap.size() > 1000) { // 避免内存泄漏String currentWindow = getCurrentTimeWindow();windowStatsMap.entrySet().removeIf(entry -> !entry.getKey().contains(currentWindow));log.debug("🧹 清理过期时间窗口数据，当前窗口: {}", currentWindow);}}/*** 时间窗口内的统计数据*/private static class WindowStats {private final AtomicInteger count = new AtomicInteger(0);public int increment() {return count.incrementAndGet();}public int getCount() {return count.get();}}
}
（2）上报 
package com.demo.filter;import com.demo.metrics.TimeWindowMetrics;
import io.micrometer.core.instrument.Gauge;
import io.micrometer.core.instrument.MeterRegistry;
import jakarta.servlet.FilterChain;
import jakarta.servlet.ServletException;
import jakarta.servlet.http.HttpServletRequest;
import jakarta.servlet.http.HttpServletResponse;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.annotation.Order;
import org.springframework.stereotype.Component;
import org.springframework.web.filter.OncePerRequestFilter;import java.io.IOException;@Component
@Slf4j
@Order(1)
public class UrlMetricsFilter extends OncePerRequestFilter {@Autowiredprivate MeterRegistry meterRegistry;// 时间窗口统计（60秒窗口）private final TimeWindowMetrics timeWindowMetrics = new TimeWindowMetrics(60);@Overrideprotected void doFilterInternal(HttpServletRequest request, HttpServletResponse response, FilterChain filterChain) throws ServletException, IOException {// 过滤掉 actuator 相关的监控端点if (shouldSkipFiltering(request)) {log.debug("忽略监控: {}", request.getRequestURI());filterChain.doFilter(request, response);return;}log.debug("进入UrlMetricsFilter - API: {} {}", request.getMethod(), request.getRequestURI());try {// 执行请求filterChain.doFilter(request, response);// 记录到时间窗口统计timeWindowMetrics.recordApiCall(request.getRequestURI(), request.getMethod(), response.getStatus());// 注册当前窗口指标registerCurrentWindowMetric(request.getRequestURI(), request.getMethod(), response.getStatus());} catch (Exception e) {log.error("UrlMetricsFilter处理异常", e);throw e;}}/*** 判断是否应该跳过过滤* @param request HTTP请求* @return true表示跳过，false表示需要过滤*/private boolean shouldSkipFiltering(HttpServletRequest request) {String requestURI = request.getRequestURI();// 过滤掉 actuator 相关的监控端点if (requestURI.contains("/actuator/")) {log.debug("跳过监控端点: {}", requestURI);return true;}// 可以添加更多需要跳过的路径// 例如：静态资源、健康检查等String[] skipPaths = {"/favicon.ico","/static/","/css/","/js/","/images/"};for (String skipPath : skipPaths) {if (requestURI.contains(skipPath)) {log.debug("跳过静态资源: {}", requestURI);return true;}}return false;}/*** 注册当前时间窗口监控指标*/private void registerCurrentWindowMetric(String uri, String method, int status) {String metricName = "api_requests_current_window";// 检查指标是否已存在if (meterRegistry.find(metricName).tag("uri", uri).tag("method", method)  .tag("status", String.valueOf(status)).gauge() == null) {try {Gauge.builder(metricName, () -> (double) timeWindowMetrics.getCurrentWindowCount(uri, method, status)).description("Current time window API request count").tag("uri", uri).tag("method", method).tag("status", String.valueOf(status)).register(meterRegistry);log.debug("✅ 注册当前窗口指标: {}#{} status:{}", uri, method, status);} catch (Exception e) {log.debug("当前窗口指标已存在: {}#{} status:{}", uri, method, status);}}}
}
 测试

前面也提到了这个自带的UI不太好用，可以看到中间有没有连线的部分（服务重启造成的），

如果能像下面这样连接起来会好看一点。 

二、统计Repository接口的入参出参数量

监控Repository接口的调用总量、入参List个数平均值、入参List个数最大值，防止慢查询。用AOP实现。

1、简单统计

关键代码
（1）bean定义
package com.demo.metrics;import lombok.Data;
import lombok.extern.slf4j.Slf4j;import java.util.concurrent.atomic.LongAdder;/*** Repository方法统计指标Bean*/
@Data
@Slf4j
public class RepositoryMetricsBean {// 方法标识private String className;private String methodName;// 统计字段private int callCount;private long totalElements;private long averageElements;private long maxElements = -1; // 初始值为-1，表示还没有数据// 内部累加器private transient final LongAdder callCountAdder = new LongAdder();private transient final LongAdder totalElementsAdder = new LongAdder();public RepositoryMetricsBean(String className, String methodName) {this.className = className;this.methodName = methodName;}/*** 记录一次方法调用* @param elementCount List参数的元素个数*/public void record(int elementCount) {// 1. 增加调用次数callCountAdder.add(1);callCount = callCountAdder.intValue();// 2. 累加元素总数totalElementsAdder.add(elementCount);totalElements = totalElementsAdder.longValue();// 3. 比较并更新最大值if (elementCount > maxElements) {maxElements = elementCount;log.debug("📈 更新最大值: {}#{} -> {}", className, methodName, maxElements);} else {log.debug("📊 当前值不大于最大值: {}#{} {} <= {}", className, methodName, elementCount, maxElements);}}/*** 聚合计算平均值*/public void aggregate() {if (callCount > 0) {averageElements = totalElements / callCount;}}/*** 获取方法的唯一标识*/public String getMethodKey() {return className + "#" + methodName;}
}
（2）AOP 
package com.demo.aop;import io.micrometer.core.instrument.Counter;
import io.micrometer.core.instrument.Gauge;
import io.micrometer.core.instrument.MeterRegistry;
import lombok.extern.slf4j.Slf4j;
import org.aspectj.lang.JoinPoint;
import org.aspectj.lang.annotation.After;
import org.aspectj.lang.annotation.Aspect;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;import com.demo.metrics.RepositoryMetricsBean;import java.util.Collection;
import java.util.concurrent.ConcurrentHashMap;@Aspect
@Component
@Slf4j
public class RepositoryMonitoringAspect {@Autowiredprivate MeterRegistry meterRegistry;// 使用ConcurrentHashMap存储每个方法的统计Beanprivate final ConcurrentHashMap<String, RepositoryMetricsBean> methodMetricsBeans = new ConcurrentHashMap<>();/** 接口调用量 repository_method_calls_total* List元素个数最大值  repository_method_list_elements_max* List元素个数平均值  repository_method_list_elements_avg** @param joinPoint*/@After("execution(* com.demo.repository.*.*(..))")public void monitorRepositoryMethod(JoinPoint joinPoint) {String className = joinPoint.getTarget().getClass().getSimpleName();String methodName = joinPoint.getSignature().getName();log.info("🔍 AOP拦截到Repository方法: {}#{}", className, methodName);// 获取List参数中的元素个数Object[] args = joinPoint.getArgs();int listElementCount = extractListElementCount(args);// 1. 统计调用量 - 先查找已存在的，不存在才创建Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();if (callsCounter == null) {try {callsCounter = Counter.builder("repository_method_calls_total").description("Repository method call count").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();}}if (callsCounter != null) {callsCounter.increment();}// 2. 记录List元素总数 - 先查找已存在的，不存在才创建Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (sumCounter == null) {try {sumCounter = Counter.builder("repository_method_list_elements_sum").description("Total number of list elements").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();}}if (sumCounter != null) {sumCounter.increment(listElementCount);}// 3. 记录List元素统计 - 使用独立的MetricsBean处理recordListElementMetrics(className, methodName, listElementCount);// 4. 从Prometheus获取数据并计算平均值，repository_method_list_elements_avgcalculateAndReportAverage(className, methodName);log.info("📊 Repository监控统计: {}#{} 调用，List元素个数: {}", className, methodName, listElementCount);}/*** 提取方法参数中List的元素个数* @param args 方法参数数组* @return List中元素的总个数*/private int extractListElementCount(Object[] args) {if (args == null || args.length == 0) {return 0;}int totalElements = 0;for (Object arg : args) {if (arg instanceof Collection) {Collection<?> collection = (Collection<?>) arg;totalElements += collection.size();log.debug("发现Collection参数，元素个数: {}", collection.size());}}return totalElements;}/*** 记录List元素统计 - 使用独立的MetricsBean处理* @param className 类名* @param methodName 方法名  * @param listElementCount 当前List元素个数*/private void recordListElementMetrics(String className, String methodName, int listElementCount) {// 1. 查找或创建对应方法的MetricsBeanRepositoryMetricsBean metricsBean = findOrCreateMetricsBean(className, methodName);// 2. 记录当前调用数据metricsBean.record(listElementCount);// 3. 聚合计算平均值metricsBean.aggregate();// 4. 确保Prometheus指标已注册ensurePrometheusMetricsRegistered(metricsBean);}/*** 查找或创建方法的MetricsBean*/private RepositoryMetricsBean findOrCreateMetricsBean(String className, String methodName) {String methodKey = className + "#" + methodName;return methodMetricsBeans.computeIfAbsent(methodKey, key -> {log.debug("✅ 创建新的MetricsBean: {}", key);return new RepositoryMetricsBean(className, methodName);});}/*** 确保Prometheus指标已注册*/private void ensurePrometheusMetricsRegistered(RepositoryMetricsBean metricsBean) {String className = metricsBean.getClassName();String methodName = metricsBean.getMethodName();// 注册最大值GaugeGauge maxGauge = meterRegistry.find("repository_method_list_elements_max").tag("class", className).tag("method", methodName).gauge();if (maxGauge == null) {try {Gauge.builder("repository_method_list_elements_max", () -> (double) metricsBean.getMaxElements()).description("Maximum number of list elements from MetricsBean").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册最大值Gauge: {}#{}", className, methodName);} catch (IllegalArgumentException e) {log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}}/*** 从Prometheus获取指标数据，计算平均值并上报* @param className 类名* @param methodName 方法名*/private void calculateAndReportAverage(String className, String methodName) {// 使用安全的获取或创建方式try {// 捕获当前的className和methodName，避免闭包问题final String finalClassName = className;final String finalMethodName = methodName;Gauge.builder("repository_method_list_elements_avg", () -> calculateAverage(finalClassName, finalMethodName)).description("Average number of list elements (calculated from Prometheus data)").tag("class", finalClassName).tag("method", finalMethodName).register(meterRegistry);log.debug("🎯 注册平均值Gauge: {}#{}", finalClassName, finalMethodName);} catch (IllegalArgumentException e) {// 如果已存在，什么都不做，Gauge会继续工作log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}/*** 从MeterRegistry中获取指标值并计算平均值* @param className 类名* @param methodName 方法名* @return 平均值*/private double calculateAverage(String className, String methodName) {try {// 从MeterRegistry获取调用次数Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();// 从MeterRegistry获取元素总数Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (callsCounter != null && sumCounter != null) {double calls = callsCounter.count();double sum = sumCounter.count();if (calls > 0) {double average = sum / calls;log.debug("📊 计算平均值: {}#{} = {} / {} = {}", className, methodName, sum, calls, average);return average;}}log.debug("📊 暂无足够数据计算平均值: {}#{}", className, methodName);return 0.0;} catch (Exception e) {log.error("❌ 计算平均值失败: {}#{}, 错误: {}", className, methodName, e.getMessage());return 0.0;}}
}
测试

调用一次

调用两次localhost:3333/prometheusDemo/user/batchQueryByUserIds

一次传16个参数，一次传17个参数。看下UI统计

① 调用量

② 最大个数

③ 平均个数

重启项目测试

访问一次localhost:3333/prometheusDemo/user/batchQueryByUserIds带18个参数，重新计算了

2、监控峰值

和上面API的监控一样，如果只是单纯的累加，那么总调用量一直在累加、max始终只有一个数字，没有意义，我们需要监控峰值。

查询1: repository_method_calls_current_window                    (调用量曲线)
查询2: repository_method_list_elements_max_current_window       (最大值曲线)
查询3: repository_method_list_elements_avg_current_window       (平均值曲线)

关键代码
（1）bean定义
package com.demo.metrics;import lombok.extern.slf4j.Slf4j;import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;/*** Repository层时间窗口指标统计*/
@Slf4j
public class RepositoryTimeWindowMetrics {// 时间窗口大小（秒）private final int windowSizeSeconds;// 存储每个时间窗口的Repository统计数据private final ConcurrentHashMap<String, RepositoryWindowStats> windowStatsMap = new ConcurrentHashMap<>();public RepositoryTimeWindowMetrics(int windowSizeSeconds) {this.windowSizeSeconds = windowSizeSeconds;}/*** 记录一次Repository方法调用*/public void recordMethodCall(String className, String methodName, int listElementCount) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;// 记录当前时间窗口的统计RepositoryWindowStats windowStats = windowStatsMap.computeIfAbsent(windowKey, k -> new RepositoryWindowStats());windowStats.record(listElementCount);// 清理过期的时间窗口数据cleanExpiredWindows();log.debug("📊 Repository时间窗口统计: {} 在窗口 {} 中调用次数: {}, 元素总数: {}, 最大值: {}", key, currentWindow, windowStats.getCallCount(), windowStats.getTotalElements(), windowStats.getMaxElements());}/*** 获取指定方法的当前时间窗口调用次数*/public int getCurrentWindowCallCount(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getCallCount() : 0;}/*** 获取指定方法的当前时间窗口元素总数*/public int getCurrentWindowTotalElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getTotalElements() : 0;}/*** 获取指定方法的当前时间窗口最大元素个数*/public int getCurrentWindowMaxElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);return stats != null ? stats.getMaxElements() : 0;}/*** 获取指定方法的当前时间窗口平均元素个数*/public double getCurrentWindowAvgElements(String className, String methodName) {String currentWindow = getCurrentTimeWindow();String key = className + "#" + methodName;String windowKey = currentWindow + "#" + key;RepositoryWindowStats stats = windowStatsMap.get(windowKey);if (stats != null && stats.getCallCount() > 0) {return (double) stats.getTotalElements() / stats.getCallCount();}return 0.0;}/*** 获取当前时间窗口标识*/private String getCurrentTimeWindow() {LocalDateTime now = LocalDateTime.now();// 根据窗口大小计算时间窗口int windowIndex = now.getSecond() / windowSizeSeconds;return now.format(DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:")) + String.format("%02d", windowIndex * windowSizeSeconds);}/*** 清理过期的时间窗口数据*/private void cleanExpiredWindows() {if (windowStatsMap.size() > 1000) { // 避免内存泄漏String currentWindow = getCurrentTimeWindow();windowStatsMap.entrySet().removeIf(entry -> !entry.getKey().contains(currentWindow));log.debug("🧹 清理过期Repository时间窗口数据，当前窗口: {}", currentWindow);}}/*** Repository时间窗口内的统计数据*/private static class RepositoryWindowStats {private final AtomicInteger callCount = new AtomicInteger(0);private final AtomicInteger totalElements = new AtomicInteger(0);private final AtomicInteger maxElements = new AtomicInteger(0);public void record(int elementCount) {callCount.incrementAndGet();totalElements.addAndGet(elementCount);// 更新最大值maxElements.updateAndGet(current -> Math.max(current, elementCount));}public int getCallCount() {return callCount.get();}public int getTotalElements() {return totalElements.get();}public int getMaxElements() {return maxElements.get();}}
}
（2）APO上报
package com.demo.aop;import io.micrometer.core.instrument.Gauge;
import io.micrometer.core.instrument.MeterRegistry;
import lombok.extern.slf4j.Slf4j;
import org.aspectj.lang.JoinPoint;
import org.aspectj.lang.annotation.After;
import org.aspectj.lang.annotation.Aspect;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;import com.demo.metrics.RepositoryTimeWindowMetrics;import java.util.Collection;@Aspect
@Component
@Slf4j
public class RepositoryMonitoringAspect {@Autowiredprivate MeterRegistry meterRegistry;// Repository时间窗口统计（60秒窗口）private final RepositoryTimeWindowMetrics timeWindowMetrics = new RepositoryTimeWindowMetrics(60);// ========== 注释掉原有累计统计功能 ==========// 使用ConcurrentHashMap存储每个方法的统计Bean// private final ConcurrentHashMap<String, RepositoryMetricsBean> methodMetricsBeans = new ConcurrentHashMap<>();/** * Repository时间窗口统计指标:* - repository_method_calls_current_window (当前窗口调用量)* - repository_method_list_elements_max_current_window (当前窗口最大值)* - repository_method_list_elements_avg_current_window (当前窗口平均值)* - repository_method_list_elements_sum_current_window (当前窗口元素总数)** @param joinPoint*/@After("execution(* com.demo.repository.*.*(..))")public void monitorRepositoryMethod(JoinPoint joinPoint) {String className = joinPoint.getTarget().getClass().getSimpleName();String methodName = joinPoint.getSignature().getName();log.info("🔍 AOP拦截到Repository方法: {}#{}", className, methodName);// 获取List参数中的元素个数Object[] args = joinPoint.getArgs();int listElementCount = extractListElementCount(args);// ========== 注释掉原有累计统计功能 ==========/*// 1. 统计调用量 - 先查找已存在的，不存在才创建Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();if (callsCounter == null) {try {callsCounter = Counter.builder("repository_method_calls_total").description("Repository method call count").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();}}if (callsCounter != null) {callsCounter.increment();}// 2. 记录List元素总数 - 先查找已存在的，不存在才创建Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (sumCounter == null) {try {sumCounter = Counter.builder("repository_method_list_elements_sum").description("Total number of list elements").tag("class", className).tag("method", methodName).register(meterRegistry);} catch (IllegalArgumentException e) {sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();}}if (sumCounter != null) {sumCounter.increment(listElementCount);}// 3. 记录List元素统计 - 使用独立的MetricsBean处理recordListElementMetrics(className, methodName, listElementCount);// 4. 从Prometheus获取数据并计算平均值，repository_method_list_elements_avgcalculateAndReportAverage(className, methodName);*/// ========== 时间窗口统计 (保留功能) ==========// 记录到时间窗口统计timeWindowMetrics.recordMethodCall(className, methodName, listElementCount);// 注册时间窗口监控指标registerTimeWindowMetrics(className, methodName);log.info("📊 Repository时间窗口监控统计: {}#{} 调用，List元素个数: {}", className, methodName, listElementCount);}/*** 提取方法参数中List的元素个数* @param args 方法参数数组* @return List中元素的总个数*/private int extractListElementCount(Object[] args) {if (args == null || args.length == 0) {return 0;}int totalElements = 0;for (Object arg : args) {if (arg instanceof Collection) {Collection<?> collection = (Collection<?>) arg;totalElements += collection.size();log.debug("发现Collection参数，元素个数: {}", collection.size());}}return totalElements;}// ========== 注释掉原有累计统计相关方法 ==========/*/*** 记录List元素统计 - 使用独立的MetricsBean处理* @param className 类名* @param methodName 方法名  * @param listElementCount 当前List元素个数*//*private void recordListElementMetrics(String className, String methodName, int listElementCount) {// 1. 查找或创建对应方法的MetricsBeanRepositoryMetricsBean metricsBean = findOrCreateMetricsBean(className, methodName);// 2. 记录当前调用数据metricsBean.record(listElementCount);// 3. 聚合计算平均值metricsBean.aggregate();// 4. 确保Prometheus指标已注册ensurePrometheusMetricsRegistered(metricsBean);}/*** 查找或创建方法的MetricsBean*//*private RepositoryMetricsBean findOrCreateMetricsBean(String className, String methodName) {String methodKey = className + "#" + methodName;return methodMetricsBeans.computeIfAbsent(methodKey, key -> {log.debug("✅ 创建新的MetricsBean: {}", key);return new RepositoryMetricsBean(className, methodName);});}/*** 确保Prometheus指标已注册*//*private void ensurePrometheusMetricsRegistered(RepositoryMetricsBean metricsBean) {String className = metricsBean.getClassName();String methodName = metricsBean.getMethodName();// 注册最大值GaugeGauge maxGauge = meterRegistry.find("repository_method_list_elements_max").tag("class", className).tag("method", methodName).gauge();if (maxGauge == null) {try {Gauge.builder("repository_method_list_elements_max", () -> (double) metricsBean.getMaxElements()).description("Maximum number of list elements from MetricsBean").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册最大值Gauge: {}#{}", className, methodName);} catch (IllegalArgumentException e) {log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}}/*** 从Prometheus获取指标数据，计算平均值并上报* @param className 类名* @param methodName 方法名*//*private void calculateAndReportAverage(String className, String methodName) {// 使用安全的获取或创建方式try {// 捕获当前的className和methodName，避免闭包问题final String finalClassName = className;final String finalMethodName = methodName;Gauge.builder("repository_method_list_elements_avg", () -> calculateAverage(finalClassName, finalMethodName)).description("Average number of list elements (calculated from Prometheus data)").tag("class", finalClassName).tag("method", finalMethodName).register(meterRegistry);log.debug("🎯 注册平均值Gauge: {}#{}", finalClassName, finalMethodName);} catch (IllegalArgumentException e) {// 如果已存在，什么都不做，Gauge会继续工作log.debug("Gauge已存在，跳过注册: {}#{}", className, methodName);}}/*** 从MeterRegistry中获取指标值并计算平均值* @param className 类名* @param methodName 方法名* @return 平均值*//*private double calculateAverage(String className, String methodName) {try {// 从MeterRegistry获取调用次数Counter callsCounter = meterRegistry.find("repository_method_calls_total").tag("class", className).tag("method", methodName).counter();// 从MeterRegistry获取元素总数Counter sumCounter = meterRegistry.find("repository_method_list_elements_sum").tag("class", className).tag("method", methodName).counter();if (callsCounter != null && sumCounter != null) {double calls = callsCounter.count();double sum = sumCounter.count();if (calls > 0) {double average = sum / calls;log.debug("📊 计算平均值: {}#{} = {} / {} = {}", className, methodName, sum, calls, average);return average;}}log.debug("📊 暂无足够数据计算平均值: {}#{}", className, methodName);return 0.0;} catch (Exception e) {log.error("❌ 计算平均值失败: {}#{}, 错误: {}", className, methodName, e.getMessage());return 0.0;}}*//*** 注册时间窗口监控指标*/private void registerTimeWindowMetrics(String className, String methodName) {// 1. 当前时间窗口调用量String callsMetricName = "repository_method_calls_current_window";if (meterRegistry.find(callsMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(callsMetricName, () -> (double) timeWindowMetrics.getCurrentWindowCallCount(className, methodName)).description("Current time window repository method call count").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口调用量指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口调用量指标已存在: {}#{}", className, methodName);}}// 2. 当前时间窗口元素总数String sumMetricName = "repository_method_list_elements_sum_current_window";if (meterRegistry.find(sumMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(sumMetricName, () -> (double) timeWindowMetrics.getCurrentWindowTotalElements(className, methodName)).description("Current time window total list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口元素总数指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口元素总数指标已存在: {}#{}", className, methodName);}}// 3. 当前时间窗口最大值String maxMetricName = "repository_method_list_elements_max_current_window";if (meterRegistry.find(maxMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(maxMetricName, () -> (double) timeWindowMetrics.getCurrentWindowMaxElements(className, methodName)).description("Current time window maximum list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口最大值指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口最大值指标已存在: {}#{}", className, methodName);}}// 4. 当前时间窗口平均值String avgMetricName = "repository_method_list_elements_avg_current_window";if (meterRegistry.find(avgMetricName).tag("class", className).tag("method", methodName).gauge() == null) {try {Gauge.builder(avgMetricName, () -> timeWindowMetrics.getCurrentWindowAvgElements(className, methodName)).description("Current time window average list elements").tag("class", className).tag("method", methodName).register(meterRegistry);log.debug("✅ 注册时间窗口平均值指标: {}#{}", className, methodName);} catch (Exception e) {log.debug("时间窗口平均值指标已存在: {}#{}", className, methodName);}}}
}
测试

平均值，期间重启服务的同样断开了连线

最大值，期间重启服务的同样断开了连线

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。
如若转载，请注明出处：http://www.pswp.cn/news/917626.shtml
繁体地址，请注明出处：http://hk.pswp.cn/news/917626.shtml

如若内容造成侵权/违法违规/事实不符，请联系多彩编程网进行投诉反馈email:809451989@qq.com，一经查实，立即删除！

相关文章

逃离智能家居“孤岛”！用 Home Assistant 打造你的全屋互联自由王国

逃离智能家居“孤岛”！用 Home Assistant 打造你的全屋互联自由王国

文章目录&#x1f914; 痛点暴击&#xff1a;智能家居的“巴别塔困境”&#x1f6e0;️ Home Assistant 是个啥&#xff1f;简单粗暴版定义&#x1f50d; 硬核拆解&#xff1a;Home Assistant 的魅力之源&#x1f680; 上车指南&#xff1a;如何开始你的 HA 之旅&#xff1f;第…
阅读更多...
数据结构：如何判断一个链表中是否存在环（Check for LOOP in Linked List）

数据结构：如何判断一个链表中是否存在环（Check for LOOP in Linked List）

目录 初始思考&#xff1a;什么叫“链表有环”&#xff1f; ❌ 第一种直接想法&#xff08;失败&#xff09;&#xff1a;我们是不是能“记住走过的节点”&#xff1f; 那我们换一个思路&#xff1a;我们能否只用两个指针来检测环&#xff1f; 第一步&#xff1a;定义两个指…
阅读更多...
深入理解Java的SPI机制，使用auto-service库优化SPI

深入理解Java的SPI机制，使用auto-service库优化SPI

文章目录一、简介二、使用1、服务提供者&#xff08;或者第三方公共&#xff09;&#xff1a;定义接口2、服务提供者&#xff1a;定义实现类3、服务提供者&#xff1a;注册服务4、构建服务提供者jar包5、客户端&#xff1a;使用 ServiceLoader 来加载服务三、源码分析1、源码2、…
阅读更多...
PPT自动化 python-pptx - 10 : 表格（tables）

PPT自动化 python-pptx - 10 : 表格（tables）

在日常工作中&#xff0c;我们经常需要制作包含表格的 PowerPoint 演示文稿&#xff0c;以此清晰展示数据或文本信息。手动制作不仅耗时&#xff0c;当数据更新时还需重复操作&#xff0c;效率低下。而 python-pptx 库为我们提供了自动化操作 PowerPoint 表格的可能。本文将详细…
阅读更多...
在安卓中使用 FFmpegKit 剪切视频并添加文字水印

在安卓中使用 FFmpegKit 剪切视频并添加文字水印

在安卓中用到的三方库&#xff1a;https://github.com/arthenica/ffmpeg-kit 这个库很强大&#xff0c;支持很多平台&#xff0c;每个平台都有各自的分支代码&#xff0c;用了一段时间&#xff0c;稳定性挺好的&#xff0c; 找到安卓下的分支&#xff1a;FFmpegKit for Andro…
阅读更多...
Flask + HTML 项目开发思路

Flask + HTML 项目开发思路

Flask HTML 项目开发思路&#xff1a;以公共资源交易信息展示为例 一、开篇明义——为什么选 Flask 框架 在众多 Python Web 框架&#xff08;如 Django、Tornado 等&#xff09;里&#xff0c;本次项目坚定选择 Flask&#xff0c;背后有清晰的技术考量&#xff1a; 1. 轻量…
阅读更多...
Vue中:deep()和 ::v-deep选择器的区别

Vue中:deep()和 ::v-deep选择器的区别

在 Vue.js 中&#xff0c;:deep()和 ::v-deep都是用于穿透组件作用域的深度选择器&#xff0c;但它们在语法、适用场景和版本支持上存在区别。以下是两者的核心差异&#xff1a;一、​​语法与用法​ &#xff1a;Vue2中用 ::v-deep&#xff0c;Vue2中不支持:deep()&#xff0c…
阅读更多...
Deep learning based descriptor

Deep learning based descriptor

1、DH3D: Deep Hierarchical 3D Descriptors for Robust Large-Scale 6DoF Relocalization 论文链接 代码链接 这是一篇训练点云的文章&#xff0c;在训练出local descriptor之后&#xff0c;通过聚类的方法得出global descriptor&#xff0c;并且提出了hierarchical network&…
阅读更多...
PandasAI连接LLM对MySQL数据库进行数据分析

PandasAI连接LLM对MySQL数据库进行数据分析

1. 引言 在之前的文章《PandasAI连接LLM进行智能数据分析》中实现了使用PandasAI连接与DeepSeek模型通过自然语言进行数据分析。不过那个例子中使用的是PandasAI 2.X&#xff0c;并且使用的是本地.csv文件来作为数据。在实际应用的系统中&#xff0c;使用.csv作为库表的情况比…
阅读更多...
FloodFill算法——DFS

FloodFill算法——DFS

FloodFill算法就是用来寻找性质相同的连通快的算法&#xff0c;这篇博客都是用dfs来实现FloodFill算法 1.图像渲染 题目链接&#xff1a;733. 图像渲染 - 力扣&#xff08;LeetCode&#xff09; 题目解析&#xff1a;将和&#xff08;sr,sc&#xff09;相连的所有像素相同的…
阅读更多...
【BUUCTF系列】[极客大挑战 2019]LoveSQL 1

【BUUCTF系列】[极客大挑战 2019]LoveSQL 1

本文仅用于技术研究&#xff0c;禁止用于非法用途。 Author:枷锁 文章目录一、题目核心漏洞分析二、关键解题步骤与技术解析1. 确定列数&#xff08;ORDER BY&#xff09;2. 联合查询获取表名3. 爆破字段名4. 提取Flag三、漏洞根源与防御方案1. 漏洞成因2. 防御措施四、CTF技巧…
阅读更多...
AI时代，童装销售的“指路明灯”

AI时代，童装销售的“指路明灯”

别看现在AI、大数据这些词眼花缭乱的&#xff0c;当年我刚入行那会儿&#xff0c;也跟你一样&#xff0c;对着一堆库存和销量数据发愁&#xff0c;不知道劲儿该往哪使。童装销售这行&#xff0c;看着简单&#xff0c;其实水挺深。不过呢&#xff0c;这二十多年摸爬滚打下来&…
阅读更多...
Swin-Transformer从浅入深详解

Swin-Transformer从浅入深详解

第一部分&#xff1a;出现背景在 Swin Transformer 出现之前&#xff0c;计算机视觉&#xff08;Computer Vision, CV&#xff09;领域主要由 CNN (卷积神经网络) 主导。后来&#xff0c;NLP&#xff08;自然语言处理&#xff09;领域的 Transformer 模型被引入 CV&#xff0c;…
阅读更多...
如何手动打包 Linux（麒麟系统）的 Qt 程序

如何手动打包 Linux（麒麟系统）的 Qt 程序

gcc版本 gcc版本确保目标系统&#xff08;运行环境&#xff09;的 GCC 版本 高于或等于开发环境的版本&#xff0c;否则程序无法在目标平台运行。通过 gcc -v 可查看当前版本。cmake生成可执行文件 强烈建议在cmakelists添加设置运行时 rpath 为 $ORIGIN/…/lib&#xff08;相对…
阅读更多...
解决 “crypto.hash is not a function”：Vite 从 6.x 升级至 7.x 后 `pnpm run dev` 报错问题

解决 “crypto.hash is not a function”：Vite 从 6.x 升级至 7.x 后 `pnpm run dev` 报错问题

&#x1f680; 作者主页&#xff1a; 有来技术 &#x1f525; 开源项目&#xff1a; youlai-mall ︱vue3-element-admin︱youlai-boot︱vue-uniapp-template &#x1f33a; 仓库主页&#xff1a; GitCode︱ Gitee ︱ Github &#x1f496; 欢迎点赞 &#x1f44d; 收藏 ⭐评论 …
阅读更多...
我的创作纪念日____在 CSDN一年来的成长历程和收获

我的创作纪念日____在 CSDN一年来的成长历程和收获

365 天创作札记&#xff1a;在代码与文字的褶皱里&#xff0c;遇见 1300 束光一年来。点开csdn网站后台粉丝数的那一刻&#xff0c;1327 这个数字在屏幕上微微发烫。原来那些在深夜敲下的字符、调试到凌晨的代码示例、反复修改的技术拆解&#xff0c;真的在时光里悄悄织成了一张…
阅读更多...
VirtualBox 的 HOST 键（主机键）是 右Ctrl 键（即键盘右侧的 Ctrl 键）笔记250802

VirtualBox 的 HOST 键（主机键）是 右Ctrl 键（即键盘右侧的 Ctrl 键）笔记250802

VirtualBox 的 HOST 键&#xff08;主机键&#xff09;是 右Ctrl 键&#xff08;即键盘右侧的 Ctrl 键&#xff09;笔记250802 VirtualBox 的 HOST 键&#xff08;主机键&#xff09;是什么?HOST键 是 右Ctrl 键VirtualBox 的 主机键&#xff08;Host Key&#xff09; 是一个…
阅读更多...
Zama的使命

Zama的使命

全同态加密&#xff08;Fully Homomorphic Encryption&#xff0c;FHE&#xff09;实现互联网端到端加密的使命的重要里程碑。(FHE) 是一种无需解密即可处理数据的技术。它可用于在公共、无需许可的区块链上创建私人智能合约&#xff0c;只有特定用户才能看到交易数据和合约状态…
阅读更多...
Go语言流式输出技术实现-服务器推送事件（Server-Sent Events, SSE）

Go语言流式输出技术实现-服务器推送事件（Server-Sent Events, SSE）

目录引言背景与技术概述实现技术细节1. HTTP 头部配置2. 事件格式与发送3. 保持连接与刷新4. 处理连接关闭4.1 使用上下文管理连接生命周期4.2 使用通道管理客户端连接5. 客户端交互6.demo7.Go转发大模型流式输出demo引言 服务器推送事件&#xff08;Server-Sent Events, SSE&…
阅读更多...
高端房产管理小程序

高端房产管理小程序

系统介绍1、用户端地图找房&#xff1a;对接地图API&#xff0c;地图形式显示周边房源,支持新盘和租房两种模式查询房价走势&#xff1a;城市房价走势&#xff0c;由后台每月录入房源搜索&#xff1a;搜索房源&#xff0c;支持多维度筛选房源类型&#xff1a;新盘销售、房屋租赁…
阅读更多...
最新文章

Copyright @ 2022~2023