SRS(simple-rtmp-server)流媒体服务器源码分析--启动

SRS(simple-rtmp-server)流媒体服务器源码分析--系统启动

一、前言

       小卒最近看SRS源码，随手写下博客，其一为了整理思路，其二也是为日后翻看方便。如果不足之处，请指教！

首先总结一下SRS源码的优点：

       1、轻量级，代码结构清楚，目前SRS3.0代码8万行左右，但几乎满足直播业务的所有要求。

       2、SRS采用State Threads，支持高并发量，高性能。

       3、SRS支持rtmp和hls，满足PC和移动直播要求。

       4、SRS支持集群部署。小集群Forward，大集群edge。

代码分析可分为两个阶段：
       一：分析代码框架，理清楚组织流程
       二：分析代码细节，熟悉SRS工作原理

二、代码分析

       相关SRS源码其他总结：

           SRS(simple-rtmp-server)流媒体服务器源码分析--系统启动
           SRS(simple-rtmp-server)流媒体服务器源码分析--RTMP消息play

           SRS(simple-rtmp-server)流媒体服务器源码分析--RTMP信息Publish

           SRS(simple-rtmp-server)流媒体服务器源码分析--HLS切片

现阶段，我主要以代码框架梳理为主。Srs源码框架如下图：

       

         系统在启动时，初始化相关类，监听相关端口，若来一个访问请求，则为该链接创建一个线程，专门处理与该链接的操作。

         main函数在srs_main_server.cpp这个文件中。在main函数中，启动参数在这里不做过多介绍。直接从run(）-> run_master()看起。

int run_master()
{
    int ret = ERROR_SUCCESS;
    
    if ((ret = _srs_server->initialize_st()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = _srs_server->initialize_signal()) != ERROR_SUCCESS) {
        return ret;
    }
    //将pid进程号写进文件
    if ((ret = _srs_server->acquire_pid_file()) != ERROR_SUCCESS) {
        return ret;
    }
    //客户端监听
    if ((ret = _srs_server->listen()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = _srs_server->register_signal()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = _srs_server->http_handle()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = _srs_server->ingest()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = _srs_server->cycle()) != ERROR_SUCCESS) {
        return ret;
    }
    
    return 0;
}

进入客户监听

 if ((ret = _srs_server->listen()) != ERROR_SUCCESS) {
        return ret;
    }

 监听内容：  不同的连接请求，有不同的监听。

int SrsServer::listen()
{
    int ret = ERROR_SUCCESS;
    // 创建一个rtmp的Streamlistener
    if ((ret = listen_rtmp()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = listen_http_api()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = listen_http_stream()) != ERROR_SUCCESS) {
        return ret;
    }
    
    if ((ret = listen_stream_caster()) != ERROR_SUCCESS) {
        return ret;
    }
    
    return ret;
}

1、首先分析RTMP连接 

int SrsServer::listen_rtmp()
{
    int ret = ERROR_SUCCESS;
    
    // stream service port.
    std::vector<std::string> ip_ports = _srs_config->get_listens();
    srs_assert((int)ip_ports.size() > 0);
    
    close_listeners(SrsListenerRtmpStream);
    
    for (int i = 0; i < (int)ip_ports.size(); i++) {
        SrsListener* listener = new SrsStreamListener(this, SrsListenerRtmpStream);
        listeners.push_back(listener);
        
        std::string ip;
        int port;
        srs_parse_endpoint(ip_ports[i], ip, port);
        
        if ((ret = listener->listen(ip, port)) != ERROR_SUCCESS) {
            srs_error("RTMP stream listen at %s:%d failed. ret=%d", ip.c_str(), port, ret);
            return ret;
        }
    }
    
    return ret;
}

        这里是listen_rtmp()函数，你也可以去看看listen_http_api()函数、listen_http_stream（）函数，其实结构都很相似，只是在创建SrsStreamListener对象时，传入了不同的参数SrsListenerRtmpStream、SrsListenerHttpApi、SrsListenerHttpStream，代表了不同类型的监听对象。

// listen_rtmp 中listen监听走这里了。
int SrsStreamListener::listen(string i, int p)
{
    int ret = ERROR_SUCCESS;
    
    ip = i;
    port = p;

    srs_freep(listener);
    listener = new SrsTcpListener(this, ip, port);

    if ((ret = listener->listen()) != ERROR_SUCCESS) {
        srs_error("tcp listen failed. ret=%d", ret);
        return ret;
    }
    
    srs_info("listen thread current_cid=%d, "
        "listen at port=%d, type=%d, fd=%d started success, ep=%s:%d",
        _srs_context->get_id(), p, type, listener->fd(), i.c_str(), p);

    srs_trace("%s listen at tcp://%s:%d, fd=%d", srs_listener_type2string(type).c_str(), ip.c_str(), port, listener->fd());

    return ret;
}

注意，这里有大量纯虚函数，不要走错路了。进入TCP监听代码

// rtmp tcp监听
int SrsTcpListener::listen()
{
//C++ Socket编程
    int ret = ERROR_SUCCESS;
    // 1、创建套接字，流式Socket(SOCK_STREAM)
    if ((_fd = socket(AF_INET, SOCK_STREAM, 0)) == -1) {
        ret = ERROR_SOCKET_CREATE;
        srs_error("create linux socket error. port=%d, ret=%d", port, ret);
        return ret;
    }
    srs_verbose("create linux socket success. port=%d, fd=%d", port, _fd);
    
    int reuse_socket = 1;
    if (setsockopt(_fd, SOL_SOCKET, SO_REUSEADDR, &reuse_socket, sizeof(int)) == -1) {
        ret = ERROR_SOCKET_SETREUSE;
        srs_error("setsockopt reuse-addr error. port=%d, ret=%d", port, ret);
        return ret;
    }
    srs_verbose("setsockopt reuse-addr success. port=%d, fd=%d", port, _fd);
    
    sockaddr_in addr;
    addr.sin_family = AF_INET;
    addr.sin_port = htons(port);
    addr.sin_addr.s_addr = inet_addr(ip.c_str());
	// 2、绑定套接字到一个IP地址和一个端口上
    if (bind(_fd, (const sockaddr*)&addr, sizeof(sockaddr_in)) == -1) {
        ret = ERROR_SOCKET_BIND;
        srs_error("bind socket error. ep=%s:%d, ret=%d", ip.c_str(), port, ret);
        return ret;
    }
    srs_verbose("bind socket success. ep=%s:%d, fd=%d", ip.c_str(), port, _fd);
    // 3、将套接字设置为监听模式等待连接请求
    if (::listen(_fd, SERVER_LISTEN_BACKLOG) == -1) {
        ret = ERROR_SOCKET_LISTEN;
        srs_error("listen socket error. ep=%s:%d, ret=%d", ip.c_str(), port, ret);
        return ret;
    }
    srs_verbose("listen socket success. ep=%s:%d, fd=%d", ip.c_str(), port, _fd);
    
    if ((_stfd = st_netfd_open_socket(_fd)) == NULL){
        ret = ERROR_ST_OPEN_SOCKET;
        srs_error("st_netfd_open_socket open socket failed. ep=%s:%d, ret=%d", ip.c_str(), port, ret);
        return ret;
    }
    srs_verbose("st open socket success. ep=%s:%d, fd=%d", ip.c_str(), port, _fd);
    // 4、等到连接一个客户之后，开启一个新的线程
    if ((ret = pthread->start()) != ERROR_SUCCESS) {
        srs_error("st_thread_create listen thread error. ep=%s:%d, ret=%d", ip.c_str(), port, ret);
        return ret;
    }
    srs_verbose("create st listen thread success, ep=%s:%d", ip.c_str(), port);
    
    return ret;
}

       此代码为C++ TCP  Socket代码，思路比较清晰，可以看到，每接受到一个rtmp访问请求，创建一个”线程“，这里暂时将其称为线程，后面再做具体介绍。创建线程代码如下：

int SrsReusableThread::start()
{
    return pthread->start();
}

int SrsThread::start()
{
    int ret = ERROR_SUCCESS;

    if(tid) {
        srs_info("thread %s already running.", _name);
        return ret;
    }
    if((tid = st_thread_create(thread_fun, this, (_joinable? 1:0), 0)) == NULL){
        ret = ERROR_ST_CREATE_CYCLE_THREAD;
	srs_error("st_thread_create failed. ret=%d", ret);
	return ret;
    }
    disposed = false;
    // we set to loop to true for thread to run.
    loop = true;
    // wait for cid to ready, for parent thread to get the cid.
    while (_cid < 0) {
        st_usleep(10 * 1000);
    }
    // now, cycle thread can run.
    can_run = true;
    return ret;
}

       来到了st_thread_create，这里要注意，这是SRS开源项目具有高并发，高性能的重要一步。这里创建的是协程，不是线程。协程是有别于进程和线程的一种组件，具有进程的独立性和线程的轻量级，听说微信能够支持8亿用户量，也是采用协程这种网络服务框架：http://www.infoq.com/cn/articles/CplusStyleCorourtine-At-Wechat。

从这里可以看出，srs是一个单线程的服务器，采用协程，主持高并发，高性能。

创建协程，协程函数为：thread_fun()

// 每连链接一个用户，创建一个协程程，该函数为协程函数
void* SrsThread::thread_fun(void* arg)
{
	SrsThread* obj = (SrsThread*)arg;
	srs_assert(obj);
	// 进入线程循环
	obj->thread_cycle();
	
	// for valgrind to detect.
	SrsThreadContext* ctx = dynamic_cast<SrsThreadContext*>(_srs_context);
	if (ctx) {
		ctx->clear_cid();
	}
	
	st_thread_exit(NULL);
	
	return NULL;
}

此时，真正进入了协程循环处理

void SrsThread::thread_cycle()
    {
        int ret = ERROR_SUCCESS;
        
        _srs_context->generate_id();
        srs_info("thread %s cycle start", _name);
        
        _cid = _srs_context->get_id();
        
        srs_assert(handler);
        handler->on_thread_start();
        
        // thread is running now.
        really_terminated = false;
        
        // wait for cid to ready, for parent thread to get the cid.
        while (!can_run && loop) {
            st_usleep(10 * 1000);
        }
        
        while (loop) {
            if ((ret = handler->on_before_cycle()) != ERROR_SUCCESS) {
                srs_warn("thread %s on before cycle failed, ignored and retry, ret=%d", _name, ret);
                goto failed;
            }
            srs_info("thread %s on before cycle success", _name);
            // 注意纯虚函数的应用
            if ((ret = handler->cycle()) != ERROR_SUCCESS) {
                if (!srs_is_client_gracefully_close(ret) && !srs_is_system_control_error(ret)) {
                    srs_warn("thread %s cycle failed, ignored and retry, ret=%d", _name, ret);
                }
                goto failed;
            }
            srs_info("thread %s cycle success", _name);
            
            if ((ret = handler->on_end_cycle()) != ERROR_SUCCESS) {
                srs_warn("thread %s on end cycle failed, ignored and retry, ret=%d", _name, ret);
                goto failed;
            }
            srs_info("thread %s on end cycle success", _name);
            
        failed:
            if (!loop) {
                break;
            }
            
            // to improve performance, donot sleep when interval is zero.
            // @see: https://github.com/ossrs/srs/issues/237
            if (cycle_interval_us != 0) {
                st_usleep(cycle_interval_us);
            }
        }
        
        // readly terminated now.
        really_terminated = true;
        
        handler->on_thread_stop();
        srs_info("thread %s cycle finished", _name);
    }

       至此，一定要熟悉C++纯虚函数的引用，本人刚学了几天C++，对虚函数和纯虚函数在SRS源码中的应用很不习惯！  好了，进入循环ret = handler->cycle()

int SrsConnection::cycle()
{
    int ret = ERROR_SUCCESS;
    
    _srs_context->generate_id();
    id = _srs_context->get_id();
    
    ip = srs_get_peer_ip(st_netfd_fileno(stfd));
    //srs_trace("ip:%s", ip);
    
    ret = do_cycle();
    
    // if socket io error, set to closed.
    if (srs_is_client_gracefully_close(ret)) {
        ret = ERROR_SOCKET_CLOSED;
    }
    
    // success.
    if (ret == ERROR_SUCCESS) {
        srs_trace("client finished.");
    }
    
    // client close peer.
    if (ret == ERROR_SOCKET_CLOSED) {
        srs_warn("client disconnect peer. ret=%d", ret);
    }

    return ERROR_SUCCESS;
}

进入ret=do_cycle()；

// TODO: return detail message when error for client.
int SrsRtmpConn::do_cycle()
{
    int ret = ERROR_SUCCESS;
    
    srs_trace("RTMP client ip=%s", ip.c_str());

    rtmp->set_recv_timeout(SRS_CONSTS_RTMP_RECV_TIMEOUT_US);
    rtmp->set_send_timeout(SRS_CONSTS_RTMP_SEND_TIMEOUT_US);
    
    if ((ret = rtmp->handshake()) != ERROR_SUCCESS) {
        srs_error("rtmp handshake failed. ret=%d", ret);
        return ret;
    }
    srs_verbose("rtmp handshake success");
    
    if ((ret = rtmp->connect_app(req)) != ERROR_SUCCESS) {
        srs_error("rtmp connect vhost/app failed. ret=%d", ret);
        return ret;
    }
    srs_verbose("rtmp connect app success");
    
    // set client ip to request.
    req->ip = ip;
    
    // discovery vhost, resolve the vhost from config
    SrsConfDirective* parsed_vhost = _srs_config->get_vhost(req->vhost);
    if (parsed_vhost) {
        req->vhost = parsed_vhost->arg0();
    }
    
    srs_info("discovery app success. schema=%s, vhost=%s, port=%s, app=%s",
        req->schema.c_str(), req->vhost.c_str(), req->port.c_str(), req->app.c_str());
    
    if (req->schema.empty() || req->vhost.empty() || req->port.empty() || req->app.empty()) {
        ret = ERROR_RTMP_REQ_TCURL;
        srs_error("discovery tcUrl failed. "
            "tcUrl=%s, schema=%s, vhost=%s, port=%s, app=%s, ret=%d",
            req->tcUrl.c_str(), req->schema.c_str(), req->vhost.c_str(), req->port.c_str(), req->app.c_str(), ret);
        return ret;
    }
    
    // check vhost
    if ((ret = check_vhost()) != ERROR_SUCCESS) {
        srs_error("check vhost failed. ret=%d", ret);
        return ret;
    }
    srs_verbose("check vhost success.");
    
    srs_trace("connect app, "
        "tcUrl=%s, pageUrl=%s, swfUrl=%s, schema=%s, vhost=%s, port=%s, app=%s, args=%s", 
        req->tcUrl.c_str(), req->pageUrl.c_str(), req->swfUrl.c_str(), 
        req->schema.c_str(), req->vhost.c_str(), req->port.c_str(),
        req->app.c_str(), (req->args? "(obj)":"null"));
    
    // show client identity
    if(req->args) {
        std::string srs_version;
        std::string srs_server_ip;
        int srs_pid = 0;
        int srs_id = 0;
        
        SrsAmf0Any* prop = NULL;
        if ((prop = req->args->ensure_property_string("srs_version")) != NULL) {
            srs_version = prop->to_str();
        }
        if ((prop = req->args->ensure_property_string("srs_server_ip")) != NULL) {
            srs_server_ip = prop->to_str();
        }
        if ((prop = req->args->ensure_property_number("srs_pid")) != NULL) {
            srs_pid = (int)prop->to_number();
        }
        if ((prop = req->args->ensure_property_number("srs_id")) != NULL) {
            srs_id = (int)prop->to_number();
        }
        
        srs_info("edge-srs ip=%s, version=%s, pid=%d, id=%d", 
            srs_server_ip.c_str(), srs_version.c_str(), srs_pid, srs_id);
        if (srs_pid > 0) {
            srs_trace("edge-srs ip=%s, version=%s, pid=%d, id=%d", 
                srs_server_ip.c_str(), srs_version.c_str(), srs_pid, srs_id);
        }
    }
    
    ret = service_cycle();
    
    http_hooks_on_close();

    return ret;
}

行了，rtmp连接就到这里，要不然都快到rtmp流接受代码了，和系统启动越走越远了，rtmp流接受后面再分析。

2、再分析http-api连接，回到int SrsServer::listen()函数中，梳理http-api链接

int SrsServer::listen_http_api()
{
    int ret = ERROR_SUCCESS;
    
#ifdef SRS_AUTO_HTTP_API
    close_listeners(SrsListenerHttpApi);
    if (_srs_config->get_http_api_enabled()) {
        SrsListener* listener = new SrsStreamListener(this, SrsListenerHttpApi);
        listeners.push_back(listener);
        
        std::string ep = _srs_config->get_http_api_listen();
        
        std::string ip;
        int port;
        srs_parse_endpoint(ep, ip, port);
        
        if ((ret = listener->listen(ip, port)) != ERROR_SUCCESS) {
            srs_error("HTTP api listen at %s:%d failed. ret=%d", ip.c_str(), port, ret);
            return ret;
        }
    }
#endif
    
    return ret;
}

listen_http_api（）函数和listen_rtmp（）函数内容非常像，再走到listener->listen()里面看看，结果来到了

int SrsStreamListener::listen(string i, int p)
{
    int ret = ERROR_SUCCESS;
    
    ip = i;
    port = p;

    srs_freep(listener);
    listener = new SrsTcpListener(this, ip, port);

    if ((ret = listener->listen()) != ERROR_SUCCESS) {
        srs_error("tcp listen failed. ret=%d", ret);
        return ret;
    }
    
    srs_info("listen thread current_cid=%d, "
        "listen at port=%d, type=%d, fd=%d started success, ep=%s:%d",
        _srs_context->get_id(), p, type, listener->fd(), i.c_str(), p);

    srs_trace("%s listen at tcp://%s:%d, fd=%d", srs_listener_type2string(type).c_str(), ip.c_str(), port, listener->fd());

    return ret;
}

和rtmp链接监听机制完全一样，只是type不同而已

enum SrsListenerType 
{
    // RTMP client,
    SrsListenerRtmpStream       = 0,
    // HTTP api,
    SrsListenerHttpApi          = 1,
    // HTTP stream, HDS/HLS/DASH
    SrsListenerHttpStream       = 2,
    // UDP stream, MPEG-TS over udp.
    SrsListenerMpegTsOverUdp    = 3,
    // TCP stream, RTSP stream.
    SrsListenerRtsp             = 4,
    // TCP stream, FLV stream over HTTP.
    SrsListenerFlv              = 5,
};

我就看了两个链接监听，监听到此为止。

 

3、http api回调注册

回到run_master（）函数中，从_srs_server->http_handle()看起。

int SrsServer::http_handle()
{
    int ret = ERROR_SUCCESS;
    
#ifdef SRS_AUTO_HTTP_API
    srs_assert(http_api_mux);
    if ((ret = http_api_mux->handle("/", new SrsHttpNotFoundHandler())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/", new SrsGoApiApi())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/", new SrsGoApiV1())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/versions", new SrsGoApiVersion())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/summaries", new SrsGoApiSummaries())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/rusages", new SrsGoApiRusages())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/self_proc_stats", new SrsGoApiSelfProcStats())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/system_proc_stats", new SrsGoApiSystemProcStats())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/meminfos", new SrsGoApiMemInfos())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/authors", new SrsGoApiAuthors())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/features", new SrsGoApiFeatures())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/vhosts/", new SrsGoApiVhosts())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/streams/", new SrsGoApiStreams())) != ERROR_SUCCESS) {
        return ret;
    }
    if ((ret = http_api_mux->handle("/api/v1/clients/", new SrsGoApiClients())) != ERROR_SUCCESS) {
        return ret;
    }
    
    // test the request info.
    if ((ret = http_api_mux->handle("/api/v1/tests/requests", new SrsGoApiRequests())) != ERROR_SUCCESS) {
        return ret;
    }
    // test the error code response.
    if ((ret = http_api_mux->handle("/api/v1/tests/errors", new SrsGoApiError())) != ERROR_SUCCESS) {
        return ret;
    }
    // test the redirect mechenism.
    if ((ret = http_api_mux->handle("/api/v1/tests/redirects", new SrsHttpRedirectHandler("/api/v1/tests/errors", SRS_CONSTS_HTTP_MovedPermanently))) != ERROR_SUCCESS) {
        return ret;
    }
    // test the http vhost.
    if ((ret = http_api_mux->handle("error.srs.com/api/v1/tests/errors", new SrsGoApiError())) != ERROR_SUCCESS) {
        return ret;
    }
    
    // TODO: FIXME: for console.
    // TODO: FIXME: support reload.
    std::string dir = _srs_config->get_http_stream_dir() + "/console";
    if ((ret = http_api_mux->handle("/console/", new SrsHttpFileServer(dir))) != ERROR_SUCCESS) {
        srs_error("http: mount console dir=%s failed. ret=%d", dir.c_str(), ret);
        return ret;
    }
    srs_trace("http: api mount /console to %s", dir.c_str());
#endif

    return ret;
}

该函数注册了http-api回调接口。可以参考：https://github.com/ossrs/srs/wiki/v2_CN_HTTPApi

比如我们可以访问http://ip:1985/api/v1  其中ip为SRS服务器地址，就可以看到从该接口返回srs服务器参数。

4、ingest（拉流，SRS主动去拉流，和推流相反）处理

注意：SRS对拉流的处理比较特殊， SRS拉流是通过ffmpeg工具去实现的，SRS代码只是实现简单的系统调用，这部分内容在后面的章节中详细说明。

int SrsIngester::start()
{
    int ret = ERROR_SUCCESS;
    
    if ((ret = parse()) != ERROR_SUCCESS) {
        clear_engines();
        ret = ERROR_SUCCESS;
        return ret;
    }
    
    // even no ingesters, we must also start it,
    // for the reload may add more ingesters.
    
    // start thread to run all encoding engines.
    if ((ret = pthread->start()) != ERROR_SUCCESS) {
        srs_error("st_thread_create failed. ret=%d", ret);
        return ret;
    }
    srs_trace("ingest thread cid=%d, current_cid=%d", pthread->cid(), _srs_context->get_id());
    
    return ret;
}

到此，可以看出，和监听过程一样，进入int SrsThread::start()函数，只是传入对象不一样而已。

 

5、SRS自服务

int SrsServer::cycle()
{
    int ret = ERROR_SUCCESS;
	srs_trace("SrsServer")

    ret = do_cycle();

#ifdef SRS_AUTO_GPERF_MC
    destroy();
    
    // remark, for gmc, never invoke the exit().
    srs_warn("sleep a long time for system st-threads to cleanup.");
    st_usleep(3 * 1000 * 1000);
    srs_warn("system quit");
#else
    // normally quit with neccessary cleanup by dispose().
    srs_warn("main cycle terminated, system quit normally.");
    dispose();
    srs_trace("srs terminated");
    
    // for valgrind to detect.
    srs_freep(_srs_config);
    srs_freep(_srs_log);
    
    exit(0);
#endif
    
    return ret;
}

int SrsServer::do_cycle()
{
    int ret = ERROR_SUCCESS;
    
    // find the max loop
    int max = srs_max(0, SRS_SYS_TIME_RESOLUTION_MS_TIMES);
    
#ifdef SRS_AUTO_STAT
    max = srs_max(max, SRS_SYS_RUSAGE_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_CPU_STAT_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_DISK_STAT_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_MEMINFO_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_PLATFORM_INFO_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_NETWORK_DEVICE_RESOLUTION_TIMES);
    max = srs_max(max, SRS_SYS_NETWORK_RTMP_SERVER_RESOLUTION_TIMES);
#endif
    
    // for asprocess.
    bool asprocess = _srs_config->get_asprocess();
    
    // the deamon thread, update the time cache
    while (true) {
        if(handler && (ret = handler->on_cycle((int)conns.size())) != ERROR_SUCCESS){
            srs_error("cycle handle failed. ret=%d", ret);
            return ret;
        }
            
        // the interval in config.
        int heartbeat_max_resolution = (int)(_srs_config->get_heartbeat_interval() / SRS_SYS_CYCLE_INTERVAL);
        
        // dynamic fetch the max.
        int temp_max = max;
        temp_max = srs_max(temp_max, heartbeat_max_resolution);
        
        for (int i = 0; i < temp_max; i++) {
            st_usleep(SRS_SYS_CYCLE_INTERVAL * 1000);
            
            // asprocess check.
            if (asprocess && ::getppid() != ppid) {
                srs_warn("asprocess ppid changed from %d to %d", ppid, ::getppid());
                return ret;
            }
            
            // gracefully quit for SIGINT or SIGTERM.
            if (signal_gracefully_quit) {
                srs_trace("cleanup for gracefully terminate.");
                return ret;
            }
        
            // for gperf heap checker,
            // @see: research/gperftools/heap-checker/heap_checker.cc
            // if user interrupt the program, exit to check mem leak.
            // but, if gperf, use reload to ensure main return normally,
            // because directly exit will cause core-dump.
#ifdef SRS_AUTO_GPERF_MC
            if (signal_gmc_stop) {
                srs_warn("gmc got singal to stop server.");
                return ret;
            }
#endif
        
            // do reload the config.
            if (signal_reload) {
                signal_reload = false;
                srs_info("get signal reload, to reload the config.");
                
                if ((ret = _srs_config->reload()) != ERROR_SUCCESS) {
                    srs_error("reload config failed. ret=%d", ret);
                    return ret;
                }
                srs_trace("reload config success.");
            }
            
            // notice the stream sources to cycle.
            if ((ret = SrsSource::cycle_all()) != ERROR_SUCCESS) {
                return ret;
            }
            
            // update the cache time
            if ((i % SRS_SYS_TIME_RESOLUTION_MS_TIMES) == 0) {
                srs_info("update current time cache.");
                srs_update_system_time_ms();
            }
            
#ifdef SRS_AUTO_STAT
            if ((i % SRS_SYS_RUSAGE_RESOLUTION_TIMES) == 0) {
                srs_info("update resource info, rss.");
                srs_update_system_rusage();
            }
            if ((i % SRS_SYS_CPU_STAT_RESOLUTION_TIMES) == 0) {
                srs_info("update cpu info, cpu usage.");
                srs_update_proc_stat();
            }
            if ((i % SRS_SYS_DISK_STAT_RESOLUTION_TIMES) == 0) {
                srs_info("update disk info, disk iops.");
                srs_update_disk_stat();
            }
            if ((i % SRS_SYS_MEMINFO_RESOLUTION_TIMES) == 0) {
                srs_info("update memory info, usage/free.");
                srs_update_meminfo();
            }
            if ((i % SRS_SYS_PLATFORM_INFO_RESOLUTION_TIMES) == 0) {
                srs_info("update platform info, uptime/load.");
                srs_update_platform_info();
            }
            if ((i % SRS_SYS_NETWORK_DEVICE_RESOLUTION_TIMES) == 0) {
                srs_info("update network devices info.");
                srs_update_network_devices();
            }
            if ((i % SRS_SYS_NETWORK_RTMP_SERVER_RESOLUTION_TIMES) == 0) {
                srs_info("update network server kbps info.");
                resample_kbps();
            }
    #ifdef SRS_AUTO_HTTP_CORE
            if (_srs_config->get_heartbeat_enabled()) {
                if ((i % heartbeat_max_resolution) == 0) {
                    srs_info("do http heartbeat, for internal server to report.");
                    http_heartbeat->heartbeat();
                }
            }
    #endif
#endif
            
            srs_info("server main thread loop");
        }
    }

    return ret;
}

主线程，更新srs时间和缓存！！至此，系统启动代码结构梳理完了。

三、总结

• 启动不同的业务。
• 监听不同的客户端类型。
• 每链接一个客户端，SRS为其创建一个协程，专门负责该路链接信息交互。
• SRS系统采用了协程网络服务框架，使得系统具有高并发，高性能等有点。