springcloud 服务健康状态监控

SpringBoot自带监控功能Actuator，可以帮助实现对程序内部运行情况监控，比如监控状况、Bean加载情况、环境变量、日志信息、线程信息等，但是无法监控eureka注册中心下各个微服务的健康状态，只能监控每一个微服务的启停状态，如果想要监控到这个状态，则需要手动进行埋点并把相关指标数据暴露到prometheus即可

对于整个监控流程图大致如下：

 grafana和prometheus在我的其他文章中有相关说明，在此主要把注册中心这边通过actuator暴露指标数据并由prometheus采集的这个过程进行详细编码说明

首先创建一个PrometheusMetricsService类，此类主要负责对接监控数据采集服务，将指标通过此类设置到actuator/prometheus插件环境里

@Service
public class PrometheusMetricsService {


    
    private static final String EUREKA_INSTANCE_STATUS = "eureka_instance_status";


    
    private static final String EUREKA_INSTANCE_COUNT = "eureka_instance_count";


    private static final String LABEL_CLIENT = "client";


    private final Gauge instanceStatusGauge;
    private final Gauge instanceCountGauge;




    public PrometheusMetricsService(CollectorRegistry registry) {
        instanceStatusGauge = Gauge
                .build(EUREKA_INSTANCE_STATUS, "instance status")
                .labelNames(LABEL_CLIENT)
                .register(registry);


        instanceCountGauge = Gauge
                .build(EUREKA_INSTANCE_COUNT, "instance count")
                .labelNames(LABEL_CLIENT)
                .register(registry);
    }


    
    void metricInstanceStatus(String client, Integer statusValue) {
        instanceStatusGauge.labels(client).set(statusValue);
    }


    
    void metricInstanceCount(String client, Integer count) {
        instanceCountGauge.labels(client).set(count);
    }


}

然后需要创建一个InstanceStateCollector类，此类定时通过eureka注册实例类PeerAwareInstanceRegistry获取注册到eureka的服务的健康状态，并通过调用上一步的PrometheusMetricsService类将各个服务的信息暴露出来

@Component
public class InstanceStateCollector {


    @Autowired
    PeerAwareInstanceRegistry registry;


    @Autowired
    PrometheusMetricsService prometheusMetricsService;


    private static final Logger log = LoggerFactory.getLogger(InstanceStateCollector.class);


    @Scheduled(cron = "0/5 * * * * ?")
    public void collect() {
        try {
            Applications applications = registry.getApplications();


            applications.getRegisteredApplications().forEach((registeredApplication) -> {
                Integer count = registeredApplication.size();
                String client = registeredApplication.getName();


                prometheusMetricsService.metricInstanceCount(client, count);


                registeredApplication.getInstances().forEach((instance) -> {
                    prometheusMetricsService.metricInstanceStatus(client,  getStatus(instance.getStatus()));


                });
            });
        } catch (Exception e) {
            log.warn("eureka下的客户端的健康状态监控信息获取失败..." + e);
        }
    }


    private Integer getStatus(InstanceInfo.InstanceStatus status){
        if(status == InstanceInfo.InstanceStatus.UP){
            return 1;
        }else if(status == InstanceInfo.InstanceStatus.DOWN){
            return 5;
        }else if(status == InstanceInfo.InstanceStatus.STARTING){
            return 2;
        }else if(status == InstanceInfo.InstanceStatus.OUT_OF_SERVICE){
            return 3;
        }else{
            return 4;
        }
    }
}

这样子prometheus监控到的各个微服务的状态status将不只是0和1的状态了，将会出现如下状态