这篇“基于FLink如何实现实时安全检测”文章的知识点大部分人都不太理解，所以小编给大家总结了以下内容，内容详细，步骤清晰，具有一定的借鉴价值，希望大家阅读完这篇文章能有所收获，下面我们一起来看看这篇“基于FLink如何实现实时安全检测”文章吧。
场景描述
针对一个内部系统，如邮件系统，公司员工的访问行为日志，存放于kafka，我们希望对于一个用户账号在同一个IP下，任意的3分钟时间内，连续登录邮件系统20次失败，下一次登录成功，这种场景能够及时获取并推送到企业微信某个指定的安全接口人。kafka中的数据，能够通过某个关键字，区分当前网络访问是否一次登录事件，且有访问时间（也就是事件时间）。在解析到符合需求的用户账号之后，第一时间进行企业微信告警推送，并将其这段时间内的访问行为，写入下游ElasticSearch。
组件版本
Flink-1.14.4
Java8
ElasticSearch-7.3.2
Kafka-2.12_2.8.1
日志结构
IP和账号皆为测试使用。
{ "user": "wangxm", "client_ip": "110.68.6.182", "source": "login", "loginname": "wangxm@test.com", "IP": "110.8.148.58", "timestamp": "17:58:12", "@timestamp": "2022-04-20T09:58:13.647Z", "ip": "110.7.231.25", "clienttype": "POP3", "result": "success", "@version": "1" }技术方案
上述场景，可考虑使用FlinkCEP及Flink的滑动窗口进行实现。由于本人在采用FlinkCEP的方案进行代码编写调试后，发现并不能满足，因此改用滑动窗口进行实现。
关键代码
主入口类
主入口类，创建了flink环境、设置了基础参数，创建了kafkaSource,接入消息后，进行了映射、过滤，并设置了水位线，进行了分组，之后设置了滑动窗口，在窗口内进行了事件统计，将复合条件的事件收集返回并写入ElasticSearch。
针对map、filter、keyBy、window等算子，都单独进行了编写，后面会一一列出来。
package com.data.dev.flink.mailTopic.main;import com.data.dev.common.javabean.BaseBean;import com.data.dev.common.javabean.kafkaMailTopic.MailMsgAlarm;import com.data.dev.elasticsearch.ElasticSearchInfo;import com.data.dev.elasticsearch.SinkToEs;import com.data.dev.flink.FlinkEnv;import com.data.dev.flink.mailTopic.OperationForLoginFailCheck.*;import com.data.dev.kafka.KafkaSourceBuilder;import com.data.dev.key.ConfigurationKey;import com.data.dev.utils.TimeUtils;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.common.eventtime.WatermarkStrategy;import org.apache.flink.connector.kafka.source.KafkaSource;import org.apache.flink.streaming.api.datastream.DataStreamSource;import org.apache.flink.streaming.api.datastream.KeyedStream;import org.apache.flink.streaming.api.datastream.SingleOutputStreamOperator;import org.apache.flink.streaming.api.datastream.WindowedStream;import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;import org.apache.flink.streaming.api.windowing.assigners.SlidingEventTimeWindows;import org.apache.flink.streaming.api.windowing.time.Time;import org.apache.flink.streaming.api.windowing.windows.TimeWindow;import java.time.Duration;@Slf4jpublic class MailMsg extends BaseBean { public static final String JobName = "告警采集平台——连续登录失败后登录成功告警"; public static final String KafkaSourceName = "Kafka Source for AlarmPlatform About Mail Topic"; public MailMsg(){ log.info("初始化滑动窗口场景告警程序"); } public static void execute(){ //① 创建Flink执行环境并设置checkpoint等必要的参数 StreamExecutionEnvironment env = FlinkEnv.getFlinkEnv(); KafkaSource<String> kafkaSource = KafkaSourceBuilder.getKafkaSource(ConfigurationKey.KAFKA_MAIL_TOPIC_NAME,ConfigurationKey.KAFKA_MAIL_CONSUMER_GROUP_ID) ; DataStreamSource<String> kafkaMailMsg = env.fromSource(kafkaSource, WatermarkStrategy.forBoundedOutOfOrderness(Duration.ofMillis(10)), KafkaSourceName); //② 筛选登录消息，创建初始登录事件流 SingleOutputStreamOperator<com.data.dev.common.javabean.kafkaMailTopic.MailMsg> loginMapDs = kafkaMailMsg.map(new MsgToBeanMapper()).name("Map算子加工"); SingleOutputStreamOperator<com.data.dev.common.javabean.kafkaMailTopic.MailMsg> loginFilterDs = loginMapDs.filter(new MailMsgForLoginFilter()).name("Filter算子加工"); //③ 设置水位线 WatermarkStrategy<com.data.dev.common.javabean.kafkaMailTopic.MailMsg> watermarkStrategy = WatermarkStrategy.<com.data.dev.common.javabean.kafkaMailTopic.MailMsg>forBoundedOutOfOrderness(Duration.ofMinutes(1)) .withTimestampAssigner((mailMsg, timestamp) -> TimeUtils.switchUTCToBeijingTimestamp(mailMsg.getTimestamp_datetime())); SingleOutputStreamOperator<com.data.dev.common.javabean.kafkaMailTopic.MailMsg> loginWmDs = loginFilterDs.assignTimestampsAndWatermarks(watermarkStrategy.withIdleness(Duration.ofMinutes(3))).name("增加水位线"); //④ 设置主键 KeyedStream<com.data.dev.common.javabean.kafkaMailTopic.MailMsg, String> loginKeyedDs = loginWmDs.keyBy(new LoginKeySelector()); //⑥ 转化为滑动窗口 WindowedStream<com.data.dev.common.javabean.kafkaMailTopic.MailMsg, String, TimeWindow> loginWindowDs = loginKeyedDs.window(SlidingEventTimeWindows.of(Time.seconds(180L),Time.seconds(90L))); //⑦ 在窗口内进行逻辑统计 SingleOutputStreamOperator<MailMsgAlarm> loginWindowsDealDs = loginWindowDs.process(new WindowProcessFuncImpl()).name("窗口处理逻辑"); //⑧ 将结果转化为通用DataStream<String>格式 SingleOutputStreamOperator<String> resultDs = loginWindowsDealDs.map(new AlarmMsgToStringMapper()).name("窗口结果转化为标准格式"); //⑨ 将最终结果写入ES resultDs.addSink(SinkToEs.getEsSinkBuilder(ElasticSearchInfo.ES_LOGIN_FAIL_INDEX_NAME,ElasticSearchInfo.ES_INDEX_TYPE_DEFAULT).build()); //⑩ 提交Flink集群进行执行 FlinkEnv.envExec(env,JobName); }}mapper算子
package com.data.dev.flink.mailTopic.OperationForLoginFailCheck;import com.alibaba.fastjson.JSON;import com.data.dev.common.javabean.BaseBean;import com.data.dev.common.javabean.kafkaMailTopic.MailMsgAlarm;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.common.functions.MapFunction;@Slf4jpublic class AlarmMsgToStringMapper extends BaseBean implements MapFunction<MailMsgAlarm, String> { @Override public String map(MailMsgAlarm mailMsgAlarm) throws Exception { return JSON.toJSONString(mailMsgAlarm); }}filter算子
package com.data.dev.flink.mailTopic.OperationForLoginFailCheck;import com.data.dev.common.javabean.BaseBean;import com.data.dev.common.javabean.kafkaMailTopic.MailMsg;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.common.functions.FilterFunction;@Slf4jpublic class MailMsgForLoginFilter extends BaseBean implements FilterFunction<MailMsg> { @Override public boolean filter(MailMsg mailMsg) { if("login".equals(mailMsg.getSource())) { log.info("筛选原始的login事件：【" + mailMsg + "】"); } return "login".equals(mailMsg.getSource()); }}keyBy算子
package com.data.dev.flink.mailTopic.OperationForLoginFailCheck;import com.data.dev.common.javabean.BaseBean;import com.data.dev.common.javabean.kafkaMailTopic.MailMsg;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.java.functions.KeySelector;@Slf4jpublic class LoginKeySelector extends BaseBean implements KeySelector<MailMsg, String> { @Override public String getKey(MailMsg mailMsg) { return mailMsg.getUser() + "@" + mailMsg.getClient_ip(); }}窗口函数（核心代码）
这里我们主要考虑使用一个事件列表，用来存储每一个窗口期内得到的连续登录，当检测到登陆失败的事件，即存入事件列表中，之后判断下一次登录失败事件，如果检测到登录成功事件，但此时登录失败的次数不足20次，则清空loginEventList，等待下一次检测。一旦符合窗口内连续登录失败超过20次且下一次登录成功这个事件，则清空此时的loginEventList并将当前登录成功的事件进行告警推送。
package com.data.dev.flink.mailTopic.OperationForLoginFailCheck;import com.data.dev.common.javabean.kafkaMailTopic.MailMsg;import com.data.dev.common.javabean.kafkaMailTopic.MailMsgAlarm;import com.data.dev.utils.HttpUtils;import com.data.dev.utils.IPUtils;import lombok.extern.slf4j.Slf4j;import org.apache.flink.streaming.api.functions.windowing.ProcessWindowFunction;import org.apache.flink.streaming.api.windowing.windows.TimeWindow;import org.apache.flink.util.Collector;import java.io.Serializable;import java.util.ArrayList;import java.util.List;@Slf4jpublic class WindowProcessFuncImpl extends ProcessWindowFunction<MailMsg, MailMsgAlarm, String, TimeWindow> implements Serializable { @Override public void process(String key, ProcessWindowFunction<MailMsg, MailMsgAlarm, String, TimeWindow>.Context context, Iterable<MailMsg> iterable, Collector<MailMsgAlarm> collector) { List<MailMsg> loginEventList = new ArrayList<>(); MailMsgAlarm mailMsgAlarm; for (MailMsg mailMsg : iterable) { log.info("收集到的登录事件【" + mailMsg + "】"); if (mailMsg.getResult().equals("fail")) { //开始检测当前窗口内的事件，并将失败的事件收集到loginEventList log.info("开始检测当前窗口内的事件，并将失败的事件收集到loginEventList"); loginEventList.add(mailMsg); } else if (mailMsg.getResult().equals("success") && loginEventList.size() < 20) {//如果检测到登录成功事件，但此时登录失败的次数不足20次，则清空loginEventList，等待下一次检测 log.info("检测到登录成功事件，但此时登录失败的次数为【" + loginEventList.size() + "】不足20次，清空loginEventList，等待下一次检测"); loginEventList.clear(); } else if (mailMsg.getResult().equals("success") && loginEventList.size() >= 20) { mailMsgAlarm = getMailMsgAlarm(loginEventList,mailMsg); log.info("检测到登录成功的事件,此时窗口内连续登录失败的次数为【" + mailMsgAlarm.getFailTimes() + "】"); //一旦符合窗口内连续登录失败超过20次且下一次登录成功这个事件，则清空此时的loginEventList并将当前登录成功的事件进行告警推送； loginEventList.clear(); doAlarmPush(mailMsgAlarm); collector.collect(mailMsgAlarm);//将当前登录成功的事件进行收集上报 } else { log.info(mailMsg.getUser() + "当前已连续：【" + loginEventList.size() + "】 次登录失败"); } } } public static MailMsgAlarm getMailMsgAlarm(List<MailMsg> eventList,MailMsg eventCurrent){ String alarmKey = eventCurrent.getUser() + "@" + eventCurrent.getClient_ip(); String loginFailStartTime = eventList.get(0).getTimestamp_datetime(); String loginSuccessTime = eventCurrent.getTimestamp_datetime(); int loginFailTimes = eventList.size(); MailMsgAlarm mailMsgAlarm = new MailMsgAlarm(); mailMsgAlarm.setMailMsg(eventCurrent); mailMsgAlarm.setAlarmKey(alarmKey); mailMsgAlarm.setStartTime(loginFailStartTime); mailMsgAlarm.setEndTime(loginSuccessTime); mailMsgAlarm.setFailTimes(loginFailTimes); return mailMsgAlarm; } public void doAlarmPush(MailMsgAlarm mailMsgAlarm){ String userKey = mailMsgAlarm.getAlarmKey(); String clientIp = mailMsgAlarm.mailMsg.getClient_ip(); boolean isWhiteListIp = IPUtils.isWhiteListIp(clientIp); if(isWhiteListIp){//如果是白名单IP，不告警 log.info("当前登录用户【" + userKey + "】属于白名单IP"); }else { //IP归属查询结果、企业微信推送告警 String user = HttpUtils.getUserByClientIp(clientIp); HttpUtils.pushAlarmMsgToWechatWork(user,mailMsgAlarm.toString()); } }}最后一次map算子
package com.data.dev.flink.mailTopic.OperationForLoginFailCheck;import com.alibaba.fastjson.JSON;import com.data.dev.common.javabean.BaseBean;import com.data.dev.common.javabean.kafkaMailTopic.MailMsgAlarm;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.common.functions.MapFunction;@Slf4jpublic class AlarmMsgToStringMapper extends BaseBean implements MapFunction<MailMsgAlarm, String> { @Override public String map(MailMsgAlarm mailMsgAlarm) throws Exception { return JSON.toJSONString(mailMsgAlarm); }}ElasticSearch工具类
package com.data.dev.elasticsearch;import com.data.dev.common.javabean.BaseBean;import com.data.dev.key.ConfigurationKey;import com.data.dev.key.ElasticSearchKey;import lombok.extern.slf4j.Slf4j;import org.apache.flink.api.common.functions.RuntimeContext;import org.apache.flink.streaming.connectors.elasticsearch.ElasticsearchSinkFunction;import org.apache.flink.streaming.connectors.elasticsearch.RequestIndexer;import org.apache.flink.streaming.connectors.elasticsearch7.ElasticsearchSink;import org.apache.flink.streaming.connectors.elasticsearch7.RestClientFactory;import org.apache.http.HttpHost;import org.apache.http.auth.AuthScope;import org.apache.http.auth.UsernamePasswordCredentials;import org.apache.http.client.CredentialsProvider;import org.apache.http.impl.client.BasicCredentialsProvider;import org.elasticsearch.action.index.IndexRequest;import org.elasticsearch.client.Requests;import java.util.ArrayList;import java.util.HashMap;import java.util.List;import java.util.Map;@Slf4jpublic class SinkToEs extends BaseBean { public static final long serialVersionUID = 2L; private static final HashMap<String,String> ES_PROPS_MAP = ConfigurationKey.getApplicationProps(); private static final String HOST = ES_PROPS_MAP.get(ConfigurationKey.ES_HOST); private static final String PASSWORD = ES_PROPS_MAP.get(ConfigurationKey.ES_PASSWORD); private static final String USERNAME = ES_PROPS_MAP.get(ConfigurationKey.ES_USERNAME); private static final String PORT = ES_PROPS_MAP.get(ConfigurationKey.ES_PORT); public static HashMap<String,String > getElasticSearchInfo(){ log.info("获取ES连接信息：【 " + "HOST="+HOST + "PORT="+PORT+"USERNAME="+USERNAME+"PASSWORD=********" + " 】"); HashMap<String,String> esInfoMap = new HashMap<>(); esInfoMap.put(ElasticSearchKey.HOST,HOST); esInfoMap.put(ElasticSearchKey.PASSWORD,PASSWORD); esInfoMap.put(ElasticSearchKey.USERNAME,USERNAME); esInfoMap.put(ElasticSearchKey.PORT,PORT); return esInfoMap; } public static ElasticsearchSink.Builder<String> getEsSinkBuilder(String esIndexName,String esType){ HashMap<String, String> esInfoMap = getElasticSearchInfo(); List<HttpHost> httpHosts = new ArrayList<>(); httpHosts.add(new HttpHost(String.valueOf(esInfoMap.get(ElasticSearchKey.HOST)), Integer.parseInt(esInfoMap.get(ElasticSearchKey.PORT)), "http")); ElasticsearchSink.Builder<String> esSinkBuilder = new ElasticsearchSink.Builder<>( httpHosts, new ElasticsearchSinkFunction<String>() { public IndexRequest createIndexRequest() { Map<String, String> json = new HashMap<>(); //log.info("写入ES的data:【"+json+"】"); IndexRequest index = Requests.indexRequest() .index(esIndexName) .type(esType) .source(json); return index; } @Override public void process(String element, RuntimeContext ctx, RequestIndexer indexer) { indexer.add(createIndexRequest()); } } ); //定义es的连接配置 带用户名密码 RestClientFactory restClientFactory = restClientBuilder -> { CredentialsProvider credentialsProvider = new BasicCredentialsProvider(); credentialsProvider.setCredentials( AuthScope.ANY, new UsernamePasswordCredentials( String.valueOf(esInfoMap.get(ElasticSearchKey.USERNAME)), String.valueOf(esInfoMap.get(ElasticSearchKey.PASSWORD)) ) ); restClientBuilder.setHttpClientConfigCallback(httpAsyncClientBuilder -> { httpAsyncClientBuilder.disableAuthCaching(); return httpAsyncClientBuilder.setDefaultCredentialsProvider(credentialsProvider); }); }; esSinkBuilder.setRestClientFactory(restClientFactory); return esSinkBuilder; }}事件实体类
package com.data.dev.common.javabean.kafkaMailTopic;import com.data.dev.common.javabean.BaseBean;import lombok.Data;import java.util.Objects;@Datapublic class MailMsgAlarm extends BaseBean { public MailMsg mailMsg; public String alarmKey; public String startTime; public String endTime; public int failTimes; @Override public String toString() { return "{" + " 'mailMsg_login_success':'" + mailMsg + "'" + ", 'alarmKey':'" + alarmKey + "'" + ", 'start_login_time_in3min':'" +startTime + "'" + ", 'end_login_time_in3min':'" +endTime + "'" + ", 'login_fail_times':'" +failTimes + "'" + "}"; } public MailMsgAlarm() { } @Override public boolean equals(Object o) { if (this == o) return true; if (!(o instanceof MailMsgAlarm)) return false; MailMsgAlarm that = (MailMsgAlarm) o; return getFailTimes() == that.getFailTimes() && getMailMsg().equals(that.getMailMsg()) && getAlarmKey().equals(that.getAlarmKey()) && getStartTime().equals(that.getStartTime()) && getEndTime().equals(that.getEndTime()); } @Override public int hashCode() { return Objects.hash(getMailMsg(), getAlarmKey(), getStartTime(), getEndTime(), getFailTimes()); }}消息实体类
package com.data.dev.common.javabean.kafkaMailTopic;import com.data.dev.common.javabean.BaseBean;import lombok.Data;import java.util.Objects;@Datapublic class MailMsg extends BaseBean { public String user; public String client_ip; public String source; public String loginName; public String mailSenderSourceIp; public String timestamp_time; public String timestamp_datetime; public String ip; public String clientType; public String result; public String version; public MailMsg() { } public MailMsg(String user, String client_ip, String source, String loginName, String mailSenderSourceIp, String timestamp_time, String timestamp_datetime, String ip, String clientType, String result, String version) { this.user = user; this.client_ip = client_ip; this.source = source; this.loginName = loginName; this.mailSenderSourceIp = mailSenderSourceIp; this.timestamp_time = timestamp_time; this.timestamp_datetime = timestamp_datetime; this.ip = ip; this.clientType = clientType; this.result = result; this.version = version; } @Override public boolean equals(Object o) { if (this == o) return true; if (!(o instanceof MailMsg)) return false; MailMsg mailMsg = (MailMsg) o; return getUser().equals(mailMsg.getUser()) && getClient_ip().equals(mailMsg.getClient_ip()) && getSource().equals(mailMsg.getSource()) && getLoginName().equals(mailMsg.getLoginName()) && getMailSenderSourceIp().equals(mailMsg.getMailSenderSourceIp()) && getTimestamp_time().equals(mailMsg.getTimestamp_time()) && getTimestamp_datetime().equals(mailMsg.getTimestamp_datetime()) && getIp().equals(mailMsg.getIp()) && getClientType().equals(mailMsg.getClientType()) && getResult().equals(mailMsg.getResult()) && getVersion().equals(mailMsg.getVersion()); } @Override public int hashCode() { return Objects.hash(getUser(), getClient_ip(), getSource(), getLoginName(), getMailSenderSourceIp(), getTimestamp_time(), getTimestamp_datetime(), getIp(), getClientType(), getResult(), getVersion()); } @Override public String toString() { return "{" + " 'user':'" + user + "'" + ", 'client_ip':'" + client_ip + "'" + ", 'source':'" + source + "'" + ", 'loginName':'" + loginName + "'" + ", 'IP':'" + mailSenderSourceIp + "'" + ", 'timestamp':'" + timestamp_time + "'" + ", '@timestamp':'" + timestamp_datetime + "'" + ", 'ip':'" + "'" + ", 'clientType':'" + clientType + "'" + ", 'result':'" + result + "'" + ", 'version':'" + version + "'" + "}"; }}以上就是关于“基于FLink如何实现实时安全检测”这篇文章的内容，相信大家都有了一定的了解，希望小编分享的内容对大家有帮助，若想了解更多相关的知识内容，请关注编程网行业资讯频道。
