深入理解Eureka-Client 发送心跳(三)

DiscoverClient

com.netflix.discovery.DiscoveryClient ,使用的@Inject //google guice 注入遵循 JSR-330规范

@Inject
DiscoveryClient(ApplicationInfoManager applicationInfoManager, EurekaClientConfig config, AbstractDiscoveryClientOptionalArgs args,
                Provider backupRegistryProvider) {
     
    // 省略N多代码
    // 初始化定时器信息
    initScheduledTasks();
    
}
private void initScheduledTasks() {
    // 省略N多代码。。。
    if (clientConfig.shouldRegisterWithEureka()) {
        int renewalIntervalInSecs = instanceInfo.getLeaseInfo().getRenewalIntervalInSecs();
        int expBackOffBound = clientConfig.getHeartbeatExecutorExponentialBackOffBound();
        logger.info("Starting heartbeat executor: " + "renew interval is: " + renewalIntervalInSecs);
 
        // 在这里,初始化一个定时器任务
        scheduler.schedule(
                new TimedSupervisorTask(
                        "heartbeat",
                        scheduler,
                        heartbeatExecutor,
                        renewalIntervalInSecs,
                        TimeUnit.SECONDS,
                        expBackOffBound,
                        new HeartbeatThread()
                ),
                renewalIntervalInSecs, TimeUnit.SECONDS);
 
        // 省略N多代码。。。
    } else {
        logger.info("Not registering with Eureka server per configuration");
    }
}

由上可以看出,在DiscoverClient这个类初始化的时候,会初始化定期任务,每30秒执行一次,用来发送心跳

HeartbeatThread

这个是用来续约的线程,主要看其run方法,

private class HeartbeatThread implements Runnable {
 
    public void run() {
        if (renew()) {
            // 更新最后一次心跳的时间
            lastSuccessfulHeartbeatTimestamp = System.currentTimeMillis();
        }
    }
}
// 续约的主方法
boolean renew() {
    EurekaHttpResponse httpResponse;
    try {
        httpResponse = eurekaTransport.registrationClient.sendHeartBeat(instanceInfo.getAppName(), instanceInfo.getId(), instanceInfo, null);
        logger.debug("{} - Heartbeat status: {}", PREFIX + appPathIdentifier, httpResponse.getStatusCode());
        if (httpResponse.getStatusCode() == 404) {
            REREGISTER_COUNTER.increment();
            logger.info("{} - Re-registering apps/{}", PREFIX + appPathIdentifier, instanceInfo.getAppName());
            return register();
        }
        return httpResponse.getStatusCode() == 200;
    } catch (Throwable e) {
        logger.error("{} - was unable to send heartbeat!", PREFIX + appPathIdentifier, e);
        return false;
    }
}

上面的代码很简单,主要就是启动一个线程,然后线程执行renew()方法, 最终发送心跳给Eureka-Server

接口地址: apps/ + appName + /' + id ,

如果接口返回值为404,就是说不存在,从来没有注册过,那么重新走注册流程

lastDirtyTimestamp

即该instance在client端最后被修改的时间戳

Eureka-Server接收心跳

InstanceResource

@PUT
public Response renewLease(
        @HeaderParam(PeerEurekaNode.HEADER_REPLICATION) String isReplication,
        @QueryParam("overriddenstatus") String overriddenStatus,
        @QueryParam("status") String status,
        @QueryParam("lastDirtyTimestamp") String lastDirtyTimestamp) {
    boolean isFromReplicaNode = "true".equals(isReplication);
    // 续约
    boolean isSuccess = registry.renew(app.getName(), id, isFromReplicaNode);
    // 续约失败
    // Not found in the registry, immediately ask for a register
    if (!isSuccess) {
        logger.warn("Not Found (Renew): {} - {}", app.getName(), id);
        return Response.status(Status.NOT_FOUND).build();
    }
    // Check if we need to sync based on dirty time stamp, the client
    // instance might have changed some value
    Response response = null;
    // 比较lastDirtyTimestamp 
    if (lastDirtyTimestamp != null && serverConfig.shouldSyncWhenTimestampDiffers()) {
        // 比较lastDirtyTimestamp的大小,这个还是比较重要的
        response = this.validateDirtyTimestamp(Long.valueOf(lastDirtyTimestamp), isFromReplicaNode);
        if (response.getStatus() == Response.Status.NOT_FOUND.getStatusCode()
                && (overriddenStatus != null)
                && !(InstanceStatus.UNKNOWN.name().equals(overriddenStatus))
                && isFromReplicaNode) {
            registry.storeOverriddenStatusIfRequired(app.getAppName(), id, InstanceStatus.valueOf(overriddenStatus));
        }
    } else {
        response = Response.ok().build();
    }
    logger.debug("Found (Renew): {} - {}; reply status={}" + app.getName(), id, response.getStatus());
    return response;
}
 
 
 
 
 
 
private Response validateDirtyTimestamp(Long lastDirtyTimestamp,
                                        boolean isReplication) {
    // 获取本机的instance实例信息
    InstanceInfo appInfo = registry.getInstanceByAppAndId(app.getName(), id, false);
    if (appInfo != null) {
        //如果lastDirtyTimestamp不为空,并且lastDirtyTimestamp和本地的不相等
        if ((lastDirtyTimestamp != null) && (!lastDirtyTimestamp.equals(appInfo.getLastDirtyTimestamp()))) {
            Object[] args = {id, appInfo.getLastDirtyTimestamp(), lastDirtyTimestamp, isReplication};
            // lastDirtyTimestamp>本地的时间,则认为当前实例是无效的,返回404错误,客户端重新发起注册
            if (lastDirtyTimestamp > appInfo.getLastDirtyTimestamp()) {
                logger.debug(
                        "Time to sync, since the last dirty timestamp differs -"
                                + " ReplicationInstance id : {},Registry : {} Incoming: {} Replication: {}",
                        args);
                return Response.status(Status.NOT_FOUND).build();
            } else if (appInfo.getLastDirtyTimestamp() > lastDirtyTimestamp) {
                // 如果是集群同步请求,本地的时间,大于客户端传过来的时间,则返回 “冲突” 这个状态回去,以本地的时间大的为准
                if (isReplication) {
                    logger.debug(
                            "Time to sync, since the last dirty timestamp differs -"
                                    + " ReplicationInstance id : {},Registry : {} Incoming: {} Replication: {}",
                            args);
                    return Response.status(Status.CONFLICT).entity(appInfo).build();
                } else {
                    return Response.ok().build();
                }
            }
        }
 
    }
    return Response.ok().build();
}

代码说明:

1.lastDirtyTimestamp 是客户端向服务端发请求的版本号 , 一切请求都以版本号大的为准。, 如: 注册

2.在调用续约的方法之后,Eureka Server 会对请求过来的lastDirtyTimestamp和本地的做对比, 如果

请求lastDirtyTimestamp>本地的时间,则认为当前实例是无效的,返回404错误,客户端重新发起注册。

3.如果是集群同步请求,本地的时间,大于其他Eureka Server传过来的时间,则返回 “冲突” 这个状态回去,

以本地的时间大的为准,注意是集群同步请求,如果是客户端传过的,是不会有这个规则的。

应用续约

//PeerAwareInstanceRegistryImpl.java
public boolean renew(final String appName, final String id, final boolean isReplication) {
    // 执行续约操作
    if (super.renew(appName, id, isReplication)) {
        // 同步Eureka-Server集群
        replicateToPeers(Action.Heartbeat, appName, id, null, null, isReplication);
        return true;
    }
    return false;
}


//AbstractInstanceRegistry.java
public boolean renew(String appName, String id, boolean isReplication) {
    // 增加续约次数到统计枚举
    RENEW.increment(isReplication);
    // 从Eureka-Server端本地的CurrentHashMap中,通过appName获取Lease信息
    Map> gMap = registry.get(appName);
    Lease leaseToRenew = null;
    if (gMap != null) {
        leaseToRenew = gMap.get(id);
    }
    // lease为空,lease在第一次注册的时候会创建,为空,则表示从来没有注册过,租约不存在
    if (leaseToRenew == null) {
        RENEW_NOT_FOUND.increment(isReplication);
        logger.warn("DS: Registry: lease doesn't exist, registering resource: {} - {}", appName, id);
        return false;
    } else {
        // 获取lease里面的instance信息
        InstanceInfo instanceInfo = leaseToRenew.getHolder();
        if (instanceInfo != null) {
            // touchASGCache(instanceInfo.getASGName());
            // 一系列状态判断,目前还不是很清楚,但是不影响主流程
            InstanceStatus overriddenInstanceStatus = this.getOverriddenInstanceStatus(
                    instanceInfo, leaseToRenew, isReplication);
            if (overriddenInstanceStatus == InstanceStatus.UNKNOWN) {
                logger.info("Instance status UNKNOWN possibly due to deleted override for instance {}"
                        + "; re-register required", instanceInfo.getId());
                RENEW_NOT_FOUND.increment(isReplication);
                return false;
            }
            if (!instanceInfo.getStatus().equals(overriddenInstanceStatus)) {
                Object[] args = {
                        instanceInfo.getStatus().name(),
                        instanceInfo.getOverriddenStatus().name(),
                        instanceInfo.getId()
                };
                logger.info(
                        "The instance status {} is different from overridden instance status {} for instance {}. "
                                + "Hence setting the status to overridden status", args);
                instanceInfo.setStatus(overriddenInstanceStatus);
            }
        }
        // 设置每分钟的续约次数
        renewsLastMin.increment();
        // 续约
        leaseToRenew.renew();
        return true;
    }
}

从上面可以看到整个续约过程,主要就是从本地的CurrentHashMap中获取租约信息, 获取到了之后,设置每分钟的续约次数以及续约时间。
renewsLastMin.increment(), 这个里面。主要是更新一个currentBucket的变量,类型为AtomicLong , 同时有个定时器一分钟去更新一次。一分钟之后,这个值会重新设置为0 。

leaseToRenew.renew() , 更新lastUpdateTimestamp, duration默认为90秒

//Lease.java
public void renew() {
    lastUpdateTimestamp = System.currentTimeMillis() + duration;
 
}

总结:

在一下三种情况,续约是返回404 , 需要客户端重新发起注册的。

1.当客户端的lastDirtyTimestamp> 大于服务端的instance的lastDirtyTimestamp时候,会认为服务端

的信息是无效的,因此无法续约,需要重新发起注册请求。

2.服务端的注册信息不存在

3.服务端的instance的status = UNKONW, 为什么会出现UNKONW这个状态呢,因为在deleteStatusOverride

的时候存在传入UNKONW的可能性。

你可能感兴趣的:(深入理解Eureka-Client 发送心跳(三))