介绍

这是一个基于Netty框架二次封装的高性能Http接口服务，增加了对http请求路由的功能，并简化了操作，目的是将接收到的日志经过简单处理后快速推送到kafka ，服务于易企秀数据埋点业务，春节期间日处理10亿+ 特点：简单、高效 （在最普通的机器环境压测QPS最高可以达到2.5w/s） 项目地址：github.com/eqxiu/log-s…

设计

依赖

• Netty4
• logback kafka appender
• ipip

快速启动

1、编译

git clone https://github.com/yanchaoguo/log-server.git

mvn assembly:assembly

复制代码

2、配置 将编译好的工程放到 /data/work/log_server，目录结构如下：

----log_server
          |
          ----bin           #依赖的jar包
          |
          ----classes       #编译后生成的classes目录
          |
          ----bin           #启动脚本
          |
          ----logs          #日志文件  

复制代码

cd /data/work/log_server/bin 编辑 ls.sh 脚本，配置端口号和kafka地址

port=9001
kafka=hadoop006:9092,hadoop007:9092,hadoop008:9092

复制代码

3、启动

./ls.sh start

复制代码

快速开发

1、目录说明

2、在action目录下新增业务处理类，比如创建FastPushAction并实现Action接口的doAction方法；该类不对日志进行处理，没有太多业务逻辑，负责将接收到的日志发送到kafka ，可以作为demo参考，具体实现如下

/**
 * /fast_push为该业务逻辑的请求路径 ，如http://localhost:port/log-server/fast_push
 */
@Route(value = "/fast_push")
public class FastPushAction implements Action {

    private static final Logger logger = LoggerFactory.getLogger(PushLogAction.class);

    @Override
    public void doAction(Request request, Response response) {

            String logs = request.getContent();  //从body中获取日志内容
            //从参数中获取 logger 名称，为空则取默认值
            String loger = Utils.isNulDefault(request.getParam("loger"),LogConfigManager.trashTopic);

            int responseStatus = CodeManager.RESPONSE_CODE_NORMAL;

            //如果body中没有内容 尝试从url参数中获取
            if (StrUtil.isEmpty(logs) && request.getParams().size()>0) {
                logs = Utils.toJson(request.getParams());
            }
            if (StrUtil.isEmpty(logs)) {
                response.setContent("<h2>not find log content</h2>");
                responseStatus = CodeManager.RESPONSE_CODE_NOT_CONTENT;
                logger.warn("not find log content");
            }else {
                //将日志推送到kafka，其中 logger 的名称要和logback.xml中的配置一致
                KafkaLoggerFactory.getLogger(loger).info(logs);
            }
            // 设置返回信息：
            response.setStatus(responseStatus);
            response.setHeader("Access-Control-Allow-Origin", "*");
            response.setHeader(CookieHeaderNames.EXPIRES , -1);
            // 返回：
            response.send();

    }

}

复制代码

3、配置logback.xml

<?xml version="1.0" encoding="UTF-8" ?>
<configuration>
    <!--trash-->
    <logger name="trash" level="INFO" additivity="false">
        <appender-ref ref="trash_kafka"/>
    </logger>

  <appender name="trash_kafka" class="com.github.danielwegener.logback.kafka.KafkaAppender">
        <encoder class="com.github.danielwegener.logback.kafka.encoding.LayoutKafkaMessageEncoder">
            <layout class="ch.qos.logback.classic.PatternLayout">
                <pattern>%msg</pattern>
            </layout>
            <charset>UTF-8</charset>
        </encoder>
        <topic>trash</topic>
        <keyingStrategy class="com.github.danielwegener.logback.kafka.keying.RoundRobinKeyingStrategy" />
        <deliveryStrategy class="com.github.danielwegener.logback.kafka.delivery.AsynchronousDeliveryStrategy" />
        <producerConfig>bootstrap.servers=${kafka.servers}
        </producerConfig>
        <producerConfig>linger.ms=1000</producerConfig>
        <producerConfig>compression.type=none</producerConfig>
        <producerConfig>buffer.memory=1073741842</producerConfig>
        <producerConfig>acks=1</producerConfig>
   </appender>
</configuration>

复制代码

4、在 LogServer.java 中注册FastPushAction

 public static void start() {
        //注册action
        ServerSetting.setAction(PushLogAction.class);
        ServerSetting.setAction(FastPushAction.class);
        try {
            new LogServer().start(ServerSetting.getPort());
        } catch (InterruptedException e) {
            log.error("LoServer start error!", e);
        }
    }

复制代码

性能

4核2G单实例 数据大小2k，压测命令如下：

ab -n2000000 -c500  "http://****:9001/log-server/fast_push?debugMode=0&sdk=tracker-view.js&ver=1.1.1&d_i=2020021955e4d920&url=https%3A%2F%2Fb.scene.eprezi.cn%2Fs%2FDPawp3qi%3Fshare_level%3D10%26from_user%3D20200211285b0f8f%26from_id%3Db9509c4e-7%26share_time%3D1581400182382%26from%3Dsinglemessage%26isappinstalled%3D0%26adpop%3D1&tit=%E6%B5%B7%E8%89%BA-%E5%8C%97%E4%BA%AC%E8%88%9E%E8%B9%88%E5%AD%A6%E9%99%A2%E4%B8%AD%E5%9B%BD%E8%88%9E%E8%80%83%E7%BA%A7%E6%95%99%E6%9D%90&ref=&u_a=&bro=%E5%BE%AE%E4%BF%A1&os=Android&o_v=8.1.0&eng=Webkit&man=Xiaomi&mod=HM-6&sns=weixin-singlemessage&n_t=wifi&s_i=v3x20200219eb3eadcb&c_i=da1c0dd6ad3b19b9a86f42bdfd31c69a&u_i=&c_p=Android&b_v=2.0&c_e=0.0.1&product=traffic_view&b_t=traffic&x_t=0&wx_o_i=&wx_n_n=&wx_sex=&wx_pro=&wx_cit=&wx_cou=&wx_hea=&wx_u_i=&wx_r_f=singlemessage&scene_id=121973512&scene_c_u=ff80808155520087015556e924cc0e8e&scene_code=DPawp3qi&scene_bizType=0&scene_property_eqAdType=1&scene_ext_yqc_ad=121973512&scene_member_type=&scene_user_type=1&foto_id=&foto_code=&rdt=1&domain=b.scene.eprezi.cn&media_id=1&pid=10000&works_id=121973512&earnings_user_id=ff80808155520087015556e924cc0e8e&publisher_id=24996&ad_unit_id=12&plan_id=201904091007&strategy_id=515&task_id=5156899&ad_platform=gdt&ad_type=tpl&ad_style_id=43&ad_size=&works_open_type=1&ad_page_num=0&man_2=%E5%B0%8F%E7%B1%B3&cou=&pro=&cit=&sex=0&new_user=&is_auth=0&device_size=36&d_t=1&e_t=element_view&count=1&scene_page_curr=&target_url=&conversion_type=ad_position_request&c_t=1582102328147"

复制代码

压测结果：

Concurrency Level:      500
Time taken for tests:   83.936 seconds
Complete requests:      2000000
Failed requests:        0
Write errors:           0
Total transferred:      272000000 bytes
HTML transferred:       0 bytes
Requests per second:    23827.75 [#/sec] (mean)
Time per request:       20.984 [ms] (mean)

复制代码

作者：易企秀大数据总监-Yarn