【SpringBoot 从入门到架构师】第19章：SpringBoot监控与运维

1. SpringBoot Actuator监控端点配置

Spring Boot Actuator 提供了丰富的生产就绪特性来监控和管理应用。以下是 Actuator 监控端点的详细配置指南：

基本依赖

<dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-actuator</artifactId> </dependency>

端点配置

2.1 启用/禁用端点

management: endpoints: web: exposure: # 暴露所有端点（生产环境谨慎使用） include: "*" # 或指定特定端点 # include: health,info,metrics,prometheus # 禁用所有端点 # exposure: # exclude: "*" # 启用/禁用特定端点 enabled-by-default: true jmx: exposure: include: "*"

2.2 常用端点列表

management: endpoint: # 健康检查 health: enabled: true show-details: always # always, when-authorized, never show-components: always # 应用信息 info: enabled: true # 指标数据 metrics: enabled: true # Prometheus 格式指标 prometheus: enabled: true # 线程转储 threaddump: enabled: true # 堆转储 heapdump: enabled: true # 配置属性 configprops: enabled: true # 环境变量 env: enabled: true # 日志级别管理 loggers: enabled: true # HTTP 请求追踪 httptrace: enabled: true # 审计事件 auditevents: enabled: true # 调度任务 scheduledtasks: enabled: true # Spring Beans beans: enabled: true # 条件评估报告 conditions: enabled: true # 映射信息 mappings: enabled: true # 关闭应用（生产环境禁用！） shutdown: enabled: false

安全配置

3.1 基本安全

spring: security: user: name: admin password: ${ACTUATOR_PASSWORD:admin123} roles: ACTUATOR_ADMIN management: endpoints: web: base-path: /manage # 自定义基础路径 path-mapping: health: healthcheck exposure: include: health,info,metrics

3.2 Spring Security 集成

@Configuration @EnableWebSecurity public class ActuatorSecurityConfig extends WebSecurityConfigurerAdapter { @Override protected void configure(HttpSecurity http) throws Exception { http .requestMatcher(EndpointRequest.toAnyEndpoint()) .authorizeRequests() .requestMatchers(EndpointRequest.to("health", "info")) .permitAll() .requestMatchers(EndpointRequest.toAnyEndpoint()) .hasRole("ACTUATOR_ADMIN") .and() .httpBasic(); } }

健康检查配置

4.1 自定义健康指示器

@Component public class CustomHealthIndicator implements HealthIndicator { @Override public Health health() { boolean isHealthy = checkService(); if (isHealthy) { return Health.up() .withDetail("service", "available") .withDetail("responseTime", "100ms") .build(); } else { return Health.down() .withDetail("service", "unavailable") .withDetail("error", "Connection timeout") .build(); } } private boolean checkService() { // 实现健康检查逻辑 return true; } }

4.2 健康检查组

management: endpoint: health: show-details: when-authorized group: custom: include: diskSpace,customHealthIndicator readiness: include: db,redis liveness: include: ping # 健康检查状态顺序 status: order: "DOWN,OUT_OF_SERVICE,UP,UNKNOWN"

指标配置

5.1 Micrometer 配置

management: metrics: export: # Prometheus 配置 prometheus: enabled: true step: 1m descriptions: true # InfluxDB 配置 influx: enabled: false db: springboot uri: http://localhost:8086 step: 1m # StatsD 配置 statsd: enabled: false host: localhost port: 8125 flavor: datadog # 自定义指标标签 tags: application: ${spring.application.name} environment: ${spring.profiles.active:default} # 启用/禁用特定指标 enable: jvm: true system: true logback: true tomcat: true http: true process: true # 指标分布统计 distribution: percentiles-histogram: http.server.requests: true percentiles: http.server.requests: 0.5,0.95,0.99 sla: http.server.requests: 10ms,50ms,100ms,200ms,500ms,1s,2s

5.2 自定义指标

@Service public class OrderService { private final MeterRegistry meterRegistry; private final Counter orderCounter; private final Timer orderTimer; public OrderService(MeterRegistry meterRegistry) { this.meterRegistry = meterRegistry; // 创建计数器 this.orderCounter = Counter.builder("orders.total") .description("Total number of orders") .tag("service", "order-service") .register(meterRegistry); // 创建计时器 this.orderTimer = Timer.builder("orders.process.time") .description("Time taken to process orders") .register(meterRegistry); } public void processOrder() { orderCounter.increment(); orderTimer.record(() -> { // 处理订单逻辑 try { Thread.sleep(100); } catch (InterruptedException e) { Thread.currentThread().interrupt(); } }); } }

信息端点配置

management: info: env: enabled: true java: enabled: true os: enabled: true info: app: name: "@project.name@" version: "@project.version@" description: "@project.description@" environment: ${spring.profiles.active} contact: email: admin@example.com phone: 123-456-7890 custom: feature: "New payment system" status: "Active"

日志管理

management: endpoint: loggers: enabled: true logging: level: root: INFO com.example: DEBUG org.springframework.web: INFO

通过 ​​/actuator/loggers​​ 端点动态调整日志级别：

# 查看所有日志级别 GET /actuator/loggers # 查看特定包日志级别 GET /actuator/loggers/com.example # 修改日志级别 POST /actuator/loggers/com.example Content-Type: application/json { "configuredLevel": "DEBUG" }

HTTP 追踪

management: tracing: sampling: probability: 1.0 # 采样率 endpoint: httptrace: enabled: true

生产环境建议配置

# application-prod.yml management: endpoints: web: exposure: include: health,info,metrics,prometheus base-path: /internal endpoint: health: show-details: when-authorized probes: enabled: true group: readiness: include: "*" liveness: include: ping,diskSpace metrics: export: prometheus: enabled: true tags: application: ${spring.application.name} environment: prod region: ${REGION:unknown} server: port: 9090 # 与主应用端口分离 # 添加延迟，防止启动时立即标记为就绪 health: readiness-state: enabled: true liveness-state: enabled: true # 健康检查路径 spring: application: name: my-service # Kubernetes 探针配置 cloud: kubernetes: probes: liveness: path: /internal/health/liveness initial-delay: 60 period: 10 readiness: path: /internal/health/readiness initial-delay: 30 period: 10

自定义端点

@Component @Endpoint(id = "custom") public class CustomEndpoint { @ReadOperation public Map<String, Object> getInfo() { Map<String, Object> info = new HashMap<>(); info.put("status", "OK"); info.put("timestamp", Instant.now()); info.put("version", "1.0.0"); return info; } @WriteOperation public String executeOperation(@Selector String operation) { return "Executed: " + operation; } @DeleteOperation public String reset() { return "Reset completed"; } }

监控集成

Prometheus + Grafana

# prometheus.yml scrape_configs: - job_name: 'spring-boot' metrics_path: '/actuator/prometheus' static_configs: - targets: ['localhost:8080'] scrape_interval: 15s

监控面板建议

1. JVM 监控 ：堆内存、GC、线程数
2. HTTP 监控 ：请求率、延迟、错误率
3. 系统监控 ：CPU、内存、磁盘
4. 业务监控 ：自定义指标

注意事项

1. 生产安全 ：确保敏感端点（如 shutdown, env）受到保护
2. 性能影响 ：某些端点（如 heapdump）可能影响性能
3. 网络隔离 ：将监控端点限制在内网访问
4. 版本兼容 ：不同 Spring Boot 版本端点可能有所差异
5. 资源限制 ：合理设置端点响应数据大小

这个配置指南涵盖了 Spring Boot Actuator 的主要功能，可以根据实际需求进行调整。

2. SpringBoot Admin可视化监控平台

一、什么是 Spring Boot Admin？

Spring Boot Admin 是一个用于管理和监控 Spring Boot 应用程序的开源工具，它提供了：

• 统一的 Web UI 管理界面
• 实时监控应用状态
• 详细的健康检查
• 日志查看和管理
• JMX beans 管理
• 线程和内存监控
• 环境属性查看

二、核心架构

两个主要角色

• Admin Server ：监控中心服务器
• Admin Client ：被监控的 Spring Boot 应用

通信方式

• HTTP（通过 Actuator 端点）
• Spring Cloud Discovery（Eureka, Consul, Nacos）
• 直接注册

三、快速开始

Admin Server 搭建

<!-- pom.xml --> <dependency> <groupId>de.codecentric</groupId> <artifactId>spring-boot-admin-starter-server</artifactId> <version>2.7.10</version> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-web</artifactId> </dependency>

# application.yml server: port: 8080 spring: application: name: admin-server

Client 应用配置

<!-- client pom.xml --> <dependency> <groupId>de.codecentric</groupId> <artifactId>spring-boot-admin-starter-client</artifactId> <version>2.7.10</version> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-actuator</artifactId> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-web</artifactId> </dependency>

# client application.yml spring: application: name: user-service boot: admin: client: url: http://localhost:8080 # Admin Server地址 instance: service-url: http://localhost:8081 server: port: 8081 management: endpoints: web: exposure: include: "*" # 暴露所有监控端点 endpoint: health: show-details: always

四、核心功能详解

应用监控面板

• 应用列表 ：显示所有注册的应用
• 状态概览 ：UP/DOWN/OFFLINE
• 详细信息 ：版本、运行时间、IP地址

健康检查

• 数据库健康状态
• 磁盘空间
• 自定义健康指示器

性能监控

# 开启性能监控 management: metrics: export: prometheus: enabled: true enable: jvm: true system: true logback: true

日志管理

// 动态调整日志级别 @RestController public class LoggingController { @PostMapping("/actuator/loggers/{name}") public void configureLogLevel(@PathVariable String name, @RequestBody Map<String, String> data) { // 动态修改日志级别 } }

五、高级配置

安全配置

@Configuration public class SecurityConfig extends WebSecurityConfigurerAdapter { @Override protected void configure(HttpSecurity http) throws Exception { http .authorizeRequests() .antMatchers("/assets/**").permitAll() .antMatchers("/login").permitAll() .anyRequest().authenticated() .and() .formLogin().loginPage("/login") .and() .logout().logoutUrl("/logout") .and() .httpBasic() .and() .csrf().disable(); } }

邮件/钉钉通知

spring: boot: admin: notify: mail: enabled: true to: admin@example.com from: monitor@example.com dingtalk: enabled: true webhook-url: https://oapi.dingtalk.com/robot/send secret: your-secret

集成 Prometheus + Grafana

management: endpoints: web: exposure: include: health,info,prometheus metrics: export: prometheus: enabled: true

六、集群部署方案

使用 Nacos 服务发现

# Admin Server spring: cloud: nacos: discovery: server-addr: localhost:8848 boot: admin: context-path: /admin # Client 应用 spring: cloud: nacos: discovery: server-addr: localhost:8848 boot: admin: client: url: http://localhost:8080/admin

多 Admin Server 负载均衡

[Nginx/LB] | --------------------------------- | | [Admin Server 1] [Admin Server 2] | | --------------------------------- | [Nacos/Eureka] | --------------------------------- | | | [Service A] [Service B] [Service C]

七、最佳实践

生产环境配置建议

spring: boot: admin: # 启用状态缓存，减少频繁查询 monitor: status-interval: 10s status-lifetime: 10s # 通知配置 notify: reminder: enabled: true period: 60s

自定义监控指标

@Component public class CustomHealthIndicator implements HealthIndicator { @Override public Health health() { // 检查外部依赖 boolean externalServiceUp = checkExternalService(); if (externalServiceUp) { return Health.up() .withDetail("externalService", "available") .build(); } else { return Health.down() .withDetail("externalService", "unavailable") .build(); } } }

监控大屏展示

// 自定义监控面板 dashboard: views: custom-view: title: "业务监控" order: 0 widgets: - title: "今日订单量" type: "counter" url: "/actuator/metrics/orders.today"

八、常见问题解决

客户端无法注册

• 检查网络连通性
• 确认 Actuator 端点已暴露
• 验证安全配置

监控数据不更新

# 调整轮询间隔 spring: boot: admin: monitor: default-timeout: 10000

内存泄漏监控

@EventListener public void handleOOMEvent(HeapDumpWebRequestEvent event) { // 处理堆转储事件 logger.warn("Potential memory leak detected!"); }

九、替代方案比较

十、总结

Spring Boot Admin 是中小型 Spring Boot 项目监控的理想选择，它：

1. 开箱即用 ：简单配置即可获得完整监控能力
2. 轻量级 ：资源消耗小，部署简单
3. 功能全面 ：覆盖了应用监控的主要需求
4. 可扩展 ：支持自定义指标和通知
5. 社区活跃 ：持续更新和维护

对于大型分布式系统，建议结合 Prometheus + Grafana 使用，Spring Boot Admin 作为补充，提供应用级别的管理和控制功能。

3. 项目健康检查、指标监控、日志监控

一、健康检查 (Health Checks)

内置健康指示器

# 访问端点 GET /actuator/health # 响应示例 { "status": "UP", "components": { "db": { "status": "UP", "details": { "database": "MySQL", "validationQuery": "isValid()" } }, "diskSpace": { "status": "UP", "details": { "total": 500107862016, "free": 300107862016, "threshold": 10485760 } }, "ping": { "status": "UP" } } }

自定义健康检查

@Component public class CustomHealthIndicator implements HealthIndicator { @Override public Health health() { // 检查自定义逻辑 boolean isServiceHealthy = checkService(); if (isServiceHealthy) { return Health.up() .withDetail("service", "available") .withDetail("responseTime", "50ms") .build(); } return Health.down() .withDetail("service", "unavailable") .withDetail("error", "Connection timeout") .build(); } private boolean checkService() { // 实现检查逻辑 return true; } }

数据库健康检查

spring: datasource: hikari: health-check-properties: connectTimeout: 1000 connectionTestQuery: "SELECT 1"

二、指标监控 (Metrics)

Micrometer 集成

<!-- 添加 Micrometer 依赖 --> <dependency> <groupId>io.micrometer</groupId> <artifactId>micrometer-registry-prometheus</artifactId> </dependency>

# 配置 Prometheus management: metrics: export: prometheus: enabled: true tags: application: ${spring.application.name}

自定义指标

@Component public class CustomMetrics { private final MeterRegistry meterRegistry; private final Counter requestCounter; private final Timer responseTimer; private final DistributionSummary payloadSize; public CustomMetrics(MeterRegistry meterRegistry) { this.meterRegistry = meterRegistry; // 创建计数器 this.requestCounter = Counter.builder("custom.requests") .description("Total number of requests") .tag("type", "api") .register(meterRegistry); // 创建计时器 this.responseTimer = Timer.builder("custom.response.time") .description("Response time of API") .register(meterRegistry); // 创建分布摘要 this.payloadSize = DistributionSummary.builder("custom.payload.size") .baseUnit("bytes") .register(meterRegistry); } public void recordRequest() { requestCounter.increment(); } public void recordResponseTime(long duration, TimeUnit unit) { responseTimer.record(duration, unit); } public void recordPayloadSize(long size) { payloadSize.record(size); } }

JVM 和系统指标

@RestController public class MetricsController { @Autowired private MeterRegistry meterRegistry; @GetMapping("/metrics/jvm") public Map<String, Object> getJvmMetrics() { Map<String, Object> metrics = new HashMap<>(); // 获取 JVM 内存使用情况 metrics.put("heap.used", meterRegistry.get("jvm.memory.used").gauge().value()); metrics.put("heap.max", meterRegistry.get("jvm.memory.max").gauge().value()); // GC 信息 metrics.put("gc.count", meterRegistry.get("jvm.gc.pause").timer().count()); metrics.put("gc.time", meterRegistry.get("jvm.gc.pause").timer().totalTime(TimeUnit.MILLISECONDS)); // 线程信息 metrics.put("threads.live", meterRegistry.get("jvm.threads.live").gauge().value()); metrics.put("threads.daemon", meterRegistry.get("jvm.threads.daemon").gauge().value()); return metrics; } }

三、日志监控 (Loggers)

动态日志级别调整

# 访问端点 GET /actuator/loggers # 查看所有日志级别 POST /actuator/loggers/{loggerName} # 修改日志级别 # 请求体示例 { "configuredLevel": "DEBUG" }

日志配置类

@Configuration @Slf4j public class LoggingConfig { @PostConstruct public void init() { // 初始化日志配置 LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory(); // 动态添加日志器 ch.qos.logback.classic.Logger logger = loggerContext.getLogger("com.example"); logger.setLevel(Level.DEBUG); } @RestController @RequestMapping("/api/logging") public class LoggingController { @PostMapping("/level") public ResponseEntity<?> changeLogLevel( @RequestParam String logger, @RequestParam String level) { // 获取 LoggerContext LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory(); // 设置日志级别 ch.qos.logback.classic.Logger logbackLogger = loggerContext.getLogger(logger); logbackLogger.setLevel(Level.valueOf(level)); log.info("Changed log level for {} to {}", logger, level); return ResponseEntity.ok().build(); } @GetMapping("/levels") public Map<String, String> getLogLevels() { Map<String, String> levels = new HashMap<>(); LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory(); for (ch.qos.logback.classic.Logger logger : loggerContext.getLoggerList()) { if (logger.getLevel() != null) { levels.put(logger.getName(), logger.getLevel().toString()); } } return levels; } } }

日志文件监控

management: endpoint: logfile: enabled: true external-file: logs/application.log # 指定日志文件路径

@Component public class LogFileMonitor { private static final Logger log = LoggerFactory.getLogger(LogFileMonitor.class); private final String logFilePath = "logs/application.log"; private long lastFileSize = 0; @Scheduled(fixedDelay = 30000) // 每30秒检查一次 public void monitorLogFile() { try { File logFile = new File(logFilePath); if (!logFile.exists()) { log.warn("Log file does not exist: {}", logFilePath); return; } long currentSize = logFile.length(); if (currentSize > lastFileSize) { long newBytes = currentSize - lastFileSize; log.info("Log file has grown by {} bytes", newBytes); // 可以在这里添加报警逻辑 if (newBytes > 1024 * 1024) { // 1MB log.warn("Log file growing too fast!"); } } lastFileSize = currentSize; } catch (Exception e) { log.error("Error monitoring log file", e); } } @EventListener public void handleLogEvent(ApplicationEvent event) { // 监听特定日志事件 if (event instanceof LoggingEvent) { // 处理日志事件 } } }

四、集成监控系统

集成 Prometheus + Grafana

# application-prometheus.yml management: metrics: export: prometheus: enabled: true step: 1m distribution: percentiles-histogram: http.server.requests: true percentiles: http.server.requests: 0.5, 0.95, 0.99 web: server: request: autotime: enabled: true

集成 ELK Stack

@Configuration public class ElkConfig { @Bean public LogstashTcpSocketAppender logstashAppender() { LogstashTcpSocketAppender appender = new LogstashTcpSocketAppender(); appender.setName("LOGSTASH"); appender.setContext(loggerContext()); appender.addDestination("localhost:5000"); LogstashEncoder encoder = new LogstashEncoder(); encoder.setCustomFields("{\"app\":\"my-spring-app\"}"); appender.setEncoder(encoder); return appender; } @Bean public LoggerContext loggerContext() { return (LoggerContext) LoggerFactory.getILoggerFactory(); } }

五、安全配置

@Configuration @EnableWebSecurity public class ActuatorSecurityConfig extends WebSecurityConfigurerAdapter { @Override protected void configure(HttpSecurity http) throws Exception { http .authorizeRequests() .antMatchers("/actuator/health", "/actuator/info").permitAll() .antMatchers("/actuator/**").hasRole("ADMIN") .anyRequest().authenticated() .and() .httpBasic() .and() .csrf().disable(); } @Bean @Override public UserDetailsService userDetailsService() { UserDetails user = User.withDefaultPasswordEncoder() .username("admin") .password("password") .roles("ADMIN") .build(); return new InMemoryUserDetailsManager(user); } }

六、自定义端点

@Component @Endpoint(id = "custom") public class CustomEndpoint { private final List<String> messages = new ArrayList<>(); @ReadOperation public List<String> getMessages() { return messages; } @WriteOperation public void addMessage(@Selector String message) { messages.add(message); } @DeleteOperation public void deleteMessage(@Selector String message) { messages.remove(message); } }

七、最佳实践建议

1. 生产环境配置 ：

management: endpoints: web: exposure: include: health,info,metrics,prometheus base-path: /internal/actuator endpoint: health: show-details: when_authorized probes: enabled: true

1. 健康检查分组 ：

management: endpoint: health: group: readiness: include: db,diskSpace liveness: include: ping

1. 指标标签 ：

@Bean MeterRegistryCustomizer<MeterRegistry> metricsCommonTags() { return registry -> registry.config() .commonTags("region", "us-east-1", "instance", InetAddress.getLocalHost().getHostName()); }

这些配置和代码示例可以帮助你构建完整的 Spring Boot 应用监控体系。