Elasticsearch优化 & filebeat配置文件优化 & logstash格式配置 & grok实践

By | 2020年9月10日
目录
[隐藏]

编码转换问题(主要就是中文乱码)

(1)input-中的codec–plain-转码

codec => plain { charset => “GB2312” }

将GB2312 的文本编码,转为UTF-8 的编码

(2)也可以在filebeat中实现编码的转换(推荐)

filebeat.prospectors: – input_type: log paths: – c:\Users\Administrator\Desktop\performanceTrace.txt encoding: GB2312

删除多余日志中的多余行

(1)logstash filter 中drop 删除

if ([message] =~ “^20.*-\ task\ request,.*,start\ time.*”) { #用正则需删除的多余行 drop {} }

(2)日志示例

2020-03-20 10:44:01,523 [33]DEBUG Debug – task request,task Id:1cbb72f1-a5ea-4e73-957c-6d20e9e12a7a,start time:2018-03-20 10:43:59 #需删除的行 — Request String : {“UserName”:“15046699923”,“Pwd”:“ZYjyh727”,“DeviceType”:2,“DeviceId”:“PC-20170525SADY”,“EquipmentNo”:null,“SSID”:“pc”,“RegisterPhones”:null,“AppKey”:“ab09d78e3b2c40b789ddfc81674bc24deac”,“Version”:“2.0.5.3”} — EndResponse String : {“ErrorCode”:0,“Success”:true,“ErrorMsg”:null,“Result”:null,“WaitInterval”:30} — End

grok 处理多种日志不同的行(重点)

(1)日志示例:

2020-03-20 10:44:01,523 [33]DEBUG Debug – task request,task Id:1cbb72f1-a5ea-4e73-957c-6d20e9e12a7a,start time:2018-03-20 10:43:59Request String : {“UserName”:“15046699923”,“Pwd”:“ZYjyh727”,“DeviceType”:2,“DeviceId”:“PC-20170525SADY”,“EquipmentNo”:null,“SSID”:“pc”,“RegisterPhones”:null,“AppKey”:“ab09d78e3b2c40b789ddfc81674bc24deac”,“Version”:“2.0.5.3”} — EndResponse String : {“ErrorCode”:0,“Success”:true,“ErrorMsg”:null,“Result”:null,“WaitInterval”:30} — End

在logstash filter中grok 分别处理3行

match => { “message” => “^20.*-\ task\ request,.*,start\ time\:%{TIMESTAMP_ISO8601:RequestTime}” } match => { “message” => “^–\ Request\ String\ :\ \{\”UserName\“:\”%{NUMBER:UserName:int}\“,\”Pwd\“:\”(?<Pwd>.*)\“,\”DeviceType\“:%{NUMBER:DeviceType:int},\”DeviceId\“:\”(?<DeviceId>.*)\“,\”EquipmentNo\“:(?<EquipmentNo>.*),\”SSID\“:(?<SSID>.*),\”RegisterPhones\“:(?<RegisterPhones>.*),\”AppKey\“:\”(?<AppKey>.*)\“,\”Version\“:\”(?<Version>.*)\“\}\ –\ \End.*” } match => { “message” => “^–\ Response\ String\ :\ \{\”ErrorCode\“:%{NUMBER:ErrorCode:int},\”Success\“:(?<Success>[a-z]*),\”ErrorMsg\“:(?<ErrorMsg>.*),\”Result\“:(?<Result>.*),\”WaitInterval\“:%{NUMBER:WaitInterval:int}\}\ –\ \End.*” } … 等多行

(2)日志示例:

# 这是一条INFO 日志 2018-09-06 21:21:40.536 [490343b4207b39e5,490343b4207b39e5] [reactor-http-epoll-4] INFO c.w.w.p.i.config.SecurityFilter – [filter,75] – skipFlag:false uri:/report-server/daily/queryDailyReportChannel authorization:GbUzq6IElKkvRswreIHd8Xv/YMDd885jyINObc543vx2H+0lhdu0p5bOu0Vd9PT+jgxJpXHYyZiPgQmyio5Sfg== # 这个一条ERROR日志 2018-09-06 21:21:15.863 [548809be071dd887,548809be071dd887] [reactor-http-epoll-4] ERROR c.w.w.c.e.WebExceptionHandler – [handle,34] – 系统异常:/report-server/game/queryPartnerGameReport\ncom.wbgg.wbcommon.core.base.exception.BusinessException: 您的账号未登录,请登录后再操作!\n\tat com.wbgg.wbcommon.core.base.wrapper.Wrapper.check(Wrapper.java:155)\n\tat com.wbgg.wbgateway.pc.infrastructure.config.SecurityFilter.filter(SecurityFilter.java:86)\n\tat org.springframework.cloud.gateway.handler.FilteringWebHandler$GatewayFilterAdapter.filter(FilteringWebHandler.java:135)\n\tat org.springframework.cloud.gateway.filter.OrderedGatewayFilter.filter(OrderedGatewayFilter.java:44)\n\tat org.springframework.cloud.gateway.handler.FilteringWebHandler$DefaultGatewayFilterChain.lambda$filter$0(FilteringWebHandler.java:117)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:52)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.Mono.subscribe(Mono.java:3695)\n\tat reactor.core.publisher.MonoIgnoreThen$ThenIgnoreMain.drain(MonoIgnoreThen.java:172)\n\tat reactor.core.publisher.MonoIgnoreThen.subscribe(MonoIgnoreThen.java:56)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFlatMap$FlatMapMain.onNext(MonoFlatMap.java:150)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxSwitchIfEmpty$SwitchIfEmptySubscriber.onNext(FluxSwitchIfEmpty.java:67)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.MonoNext$NextSubscriber.onNext(MonoNext.java:76)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.innerNext(FluxConcatMap.java:275)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapInner.onNext(FluxConcatMap.java:849)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxMap$MapSubscriber.onNext(FluxMap.java:114)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxSwitchIfEmpty$SwitchIfEmptySubscriber.onNext(FluxSwitchIfEmpty.java:67)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.Operators$MonoSubscriber.complete(Operators.java:1505)\n\tat reactor.core.publisher.MonoFlatMap$FlatMapInner.onNext(MonoFlatMap.java:241)\n\tat reactor.core.publisher.Operators$ScalarSubscription.request(Operators.java:2070)\n\tat reactor.core.publisher.MonoFlatMap$FlatMapInner.onSubscribe(MonoFlatMap.java:230)\n\tat reactor.core.publisher.MonoJust.subscribe(MonoJust.java:54)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFlatMap$FlatMapMain.onNext(MonoFlatMap.java:150)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxMap$MapSubscriber.onNext(FluxMap.java:114)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.MonoNext$NextSubscriber.onNext(MonoNext.java:76)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.innerNext(FluxConcatMap.java:275)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapInner.onNext(FluxConcatMap.java:849)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxOnErrorResume$ResumeSubscriber.onNext(FluxOnErrorResume.java:73)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxPeek$PeekSubscriber.onNext(FluxPeek.java:192)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.Operators$MonoSubscriber.complete(Operators.java:1505)\n\tat reactor.core.publisher.MonoFilterWhen$MonoFilterWhenMain.innerResult(MonoFilterWhen.java:193)\n\tat reactor.core.publisher.MonoFilterWhen$FilterWhenInner.onNext(MonoFilterWhen.java:260)\n\tat reactor.core.publisher.MonoFilterWhen$FilterWhenInner.onNext(MonoFilterWhen.java:228)\n\tat reactor.core.publisher.Operators$ScalarSubscription.request(Operators.java:2070)\n\tat reactor.core.publisher.MonoFilterWhen$FilterWhenInner.onSubscribe(MonoFilterWhen.java:249)\n\tat reactor.core.publisher.MonoJust.subscribe(MonoJust.java:54)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.Mono.subscribe(Mono.java:3695)\n\tat reactor.core.publisher.MonoFilterWhen$MonoFilterWhenMain.onNext(MonoFilterWhen.java:150)\n\tat reactor.core.publisher.Operators$ScalarSubscription.request(Operators.java:2070)\n\tat reactor.core.publisher.MonoFilterWhen$MonoFilterWhenMain.onSubscribe(MonoFilterWhen.java:103)\n\tat reactor.core.publisher.MonoJust.subscribe(MonoJust.java:54)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFilterWhen.subscribe(MonoFilterWhen.java:56)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoPeek.subscribe(MonoPeek.java:71)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoOnErrorResume.subscribe(MonoOnErrorResume.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.Mono.subscribe(Mono.java:3695)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.drain(FluxConcatMap.java:442)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.onNext(FluxConcatMap.java:244)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxDematerialize$DematerializeSubscriber.onNext(FluxDematerialize.java:114)\n\tat reactor.core.publisher.FluxDematerialize$DematerializeSubscriber.onNext(FluxDematerialize.java:42)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.drainAsync(FluxFlattenIterable.java:395)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.drain(FluxFlattenIterable.java:638)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.onNext(FluxFlattenIterable.java:242)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxPeekFuseable$PeekFuseableSubscriber.onNext(FluxPeekFuseable.java:204)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.Operators$MonoSubscriber.complete(Operators.java:1505)\n\tat reactor.core.publisher.MonoCollectList$MonoBufferAllSubscriber.onComplete(MonoCollectList.java:118)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.DrainUtils.postCompleteDrain(DrainUtils.java:131)\n\tat reactor.core.publisher.DrainUtils.postComplete(DrainUtils.java:186)\n\tat reactor.core.publisher.FluxMaterialize$MaterializeSubscriber.onComplete(FluxMaterialize.java:134)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.drainAsync(FluxFlattenIterable.java:325)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.drain(FluxFlattenIterable.java:638)\n\tat reactor.core.publisher.FluxFlattenIterable$FlattenIterableSubscriber.onComplete(FluxFlattenIterable.java:259)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.FluxMapFuseable$MapFuseableSubscriber.onComplete(FluxMapFuseable.java:144)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.Operators$MonoSubscriber.complete(Operators.java:1508)\n\tat reactor.core.publisher.MonoCollectList$MonoBufferAllSubscriber.onComplete(MonoCollectList.java:118)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.FluxFlatMap$FlatMapMain.checkTerminated(FluxFlatMap.java:794)\n\tat reactor.core.publisher.FluxFlatMap$FlatMapMain.drainLoop(FluxFlatMap.java:560)\n\tat reactor.core.publisher.FluxFlatMap$FlatMapMain.drain(FluxFlatMap.java:540)\n\tat reactor.core.publisher.FluxFlatMap$FlatMapMain.onComplete(FluxFlatMap.java:426)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onComplete(ScopePassingSpanSubscriber.java:112)\n\tat reactor.core.publisher.FluxIterable$IterableSubscription.slowPath(FluxIterable.java:265)\n\tat reactor.core.publisher.FluxIterable$IterableSubscription.request(FluxIterable.java:201)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.request(ScopePassingSpanSubscriber.java:79)\n\tat reactor.core.publisher.FluxFlatMap$FlatMapMain.onSubscribe(FluxFlatMap.java:335)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onSubscribe(ScopePassingSpanSubscriber.java:71)\n\tat reactor.core.publisher.FluxIterable.subscribe(FluxIterable.java:139)\n\tat reactor.core.publisher.FluxIterable.subscribe(FluxIterable.java:63)\n\tat reactor.core.publisher.FluxLiftFuseable.subscribe(FluxLiftFuseable.java:70)\n\tat reactor.core.publisher.FluxFlatMap.subscribe(FluxFlatMap.java:97)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.MonoCollectList.subscribe(MonoCollectList.java:59)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoMapFuseable.subscribe(MonoMapFuseable.java:59)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFlattenIterable.subscribe(MonoFlattenIterable.java:101)\n\tat reactor.core.publisher.FluxLiftFuseable.subscribe(FluxLiftFuseable.java:70)\n\tat reactor.core.publisher.FluxMaterialize.subscribe(FluxMaterialize.java:40)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.MonoCollectList.subscribe(MonoCollectList.java:59)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoPeekFuseable.subscribe(MonoPeekFuseable.java:74)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFlattenIterable.subscribe(MonoFlattenIterable.java:101)\n\tat reactor.core.publisher.FluxLiftFuseable.subscribe(FluxLiftFuseable.java:70)\n\tat reactor.core.publisher.FluxDematerialize.subscribe(FluxDematerialize.java:39)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.FluxDefer.subscribe(FluxDefer.java:54)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.FluxConcatMap.subscribe(FluxConcatMap.java:121)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.MonoNext.subscribe(MonoNext.java:40)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoMap.subscribe(MonoMap.java:55)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoFlatMap.subscribe(MonoFlatMap.java:60)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoSwitchIfEmpty.subscribe(MonoSwitchIfEmpty.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoMap.subscribe(MonoMap.java:55)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.Mono.subscribe(Mono.java:3695)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.drain(FluxConcatMap.java:442)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.onNext(FluxConcatMap.java:244)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onNext(ScopePassingSpanSubscriber.java:96)\n\tat reactor.core.publisher.FluxIterable$IterableSubscription.slowPath(FluxIterable.java:243)\n\tat reactor.core.publisher.FluxIterable$IterableSubscription.request(FluxIterable.java:201)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.request(ScopePassingSpanSubscriber.java:79)\n\tat reactor.core.publisher.FluxConcatMap$ConcatMapImmediate.onSubscribe(FluxConcatMap.java:229)\n\tat org.springframework.cloud.sleuth.instrument.reactor.ScopePassingSpanSubscriber.onSubscribe(ScopePassingSpanSubscriber.java:71)\n\tat reactor.core.publisher.FluxIterable.subscribe(FluxIterable.java:139)\n\tat reactor.core.publisher.FluxIterable.subscribe(FluxIterable.java:63)\n\tat reactor.core.publisher.FluxLiftFuseable.subscribe(FluxLiftFuseable.java:70)\n\tat reactor.core.publisher.FluxConcatMap.subscribe(FluxConcatMap.java:121)\n\tat reactor.core.publisher.FluxLift.subscribe(FluxLift.java:46)\n\tat reactor.core.publisher.MonoNext.subscribe(MonoNext.java:40)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoSwitchIfEmpty.subscribe(MonoSwitchIfEmpty.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoFlatMap.subscribe(MonoFlatMap.java:60)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoFlatMap.subscribe(MonoFlatMap.java:60)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:52)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:52)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:52)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat org.springframework.cloud.sleuth.instrument.web.TraceWebFilter$MonoWebFilterTrace.subscribe(TraceWebFilter.java:180)\n\tat reactor.core.publisher.MonoDefer.subscribe(MonoDefer.java:52)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoOnErrorResume.subscribe(MonoOnErrorResume.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoOnErrorResume.subscribe(MonoOnErrorResume.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.MonoPeekTerminal.subscribe(MonoPeekTerminal.java:61)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoOnErrorResume.subscribe(MonoOnErrorResume.java:44)\n\tat reactor.core.publisher.MonoLift.subscribe(MonoLift.java:45)\n\tat reactor.core.publisher.Mono.subscribe(Mono.java:3695)\n\tat reactor.core.publisher.MonoIgnoreThen$ThenIgnoreMain.drain(MonoIgnoreThen.java:172)\n\tat reactor.core.publisher.MonoIgnoreThen.subscribe(MonoIgnoreThen.java:56)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoPeekFuseable.subscribe(MonoPeekFuseable.java:70)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.core.publisher.MonoPeekTerminal.subscribe(MonoPeekTerminal.java:61)\n\tat reactor.core.publisher.MonoLiftFuseable.subscribe(MonoLiftFuseable.java:55)\n\tat reactor.netty.http.server.HttpServerHandle.onStateChange(HttpServerHandle.java:64)\n\tat reactor.netty.tcp.TcpServerBind$ChildObserver.onStateChange(TcpServerBind.java:226)\n\tat reactor.netty.http.server.HttpServerOperations.onInboundNext(HttpServerOperations.java:434)\n\tat reactor.netty.channel.ChannelOperationsHandler.channelRead(ChannelOperationsHandler.java:141)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:374)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:360)\n\tat io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:352)\n\tat reactor.netty.http.server.HttpTrafficHandler.channelRead(HttpTrafficHandler.java:160)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:374)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:360)\n\tat io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:352)\n\tat io.netty.channel.CombinedChannelDuplexHandler$DelegatingChannelHandlerContext.fireChannelRead(CombinedChannelDuplexHandler.java:438)\n\tat io.netty.handler.codec.ByteToMessageDecoder.fireChannelRead(ByteToMessageDecoder.java:328)\n\tat io.netty.handler.codec.ByteToMessageDecoder.channelRead(ByteToMessageDecoder.java:302)\n\tat io.netty.channel.CombinedChannelDuplexHandler.channelRead(CombinedChannelDuplexHandler.java:253)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:374)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:360)\n\tat io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:352)\n\tat io.netty.channel.DefaultChannelPipeline$HeadContext.channelRead(DefaultChannelPipeline.java:1422)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:374)\n\tat io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:360)\n\tat io.netty.channel.DefaultChannelPipeline.fireChannelRead(DefaultChannelPipeline.java:931)\n\tat io.netty.channel.epoll.AbstractEpollStreamChannel$EpollStreamUnsafe.epollInReady(AbstractEpollStreamChannel.java:799)\n\tat io.netty.channel.epoll.EpollEventLoop.processReady(EpollEventLoop.java:433)\n\tat io.netty.channel.epoll.EpollEventLoop.run(EpollEventLoop.java:330)\n\tat io.netty.util.concurrent.SingleThreadEventExecutor$6.run(SingleThreadEventExecutor.java:1044)\n\tat io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)\n\tat java.lang.Thread.run(Thread.java:748)

在logstash filter中grok 规则进行匹配处理

input { kafka { id => “test-kafka-input” bootstrap_servers => [“192.168.0.250:9092”] # kafka地址 group_id => “logstash” # kafka group topics => [“test”, “filebeat”] # kafka topics codec => json # 设定输入类型为json } } filter { # mutate { # gsub => [ “message”, “\r”, “” ] # 替换掉换行符 # } grok { match => [“message”,“%{TIMESTAMP_ISO8601:timestamp}\s+%{SYSLOG5424SD:uid}\s+%{SYSLOG5424SD:threadid}\s+%{LOGLEVEL:loglevel}\s+%{JAVACLASS:javaclass}\s+.?\s+%{SYSLOG5424SD}\s+.?\s+%{GREEDYDATA:message}”] # 配置正则表达式和标签匹配日志 overwrite => [“message”] # 将上面%{GREEDYDATA:message} 标签覆盖到message上 } date { match => [ “timestamp”, “yyyy-MM-dd HH:mm:ss,SSS” ] # 配置timestamp 时间格式 target => “@timestamp” # 将上面grok正则匹配的标签timestamp 覆盖到默认date “@timestamp” 上面,以便kibana中看到打印的最新时间 } # 下面这段是为了解决Elasticsearch 默认时间是0时区,不是东八区,所以默认显示时间比东八区少8个小时,这时我们通过ruby 进行时间格式的修改,增加8个小时,示例如下: ruby { code => “event.set(‘timestamp’, event.get(‘@timestamp’).time.localtime + 8*60*60)” } ruby { code => “event.set(‘@timestamp’,event.get(‘timestamp’))” } # 配置要删除的多余的一些字符串,通过mutate模块进行删除 mutate { remove_field => [“timestamp”,“hostname”,“tags”,“stream”,“agent”,“ecs”,“input”,“[kubernetes][container][name]”,“[kubernetes][labels][pod-template-hash]”,“[kubernetes][pod][uid]”,“[kubernetes][replicaset]”,“@version”,“[log][offset]”] } json { source => “@fields” # 删除filebeat 自带的不需要的元数据 remove_field => [ “beat”,“@fields”,“fields”,“index_name”,“offset”,“source”,“message”,“time”,“tags”] } # json { # source => “message” # remove_field => [ “message” ] # } # multiline { # pattern => “^\d{4}-\d{1,2}-\d{1,2}\s\d{1,2}:\d{1,2}:\d{1,2}” # negate => true # what => “previous” # } } output { elasticsearch { hosts => [“http://192.168.0.250:9200”] user => logstash_admin password => “YHkdypsPKqw5gaWKE” index => “game-filebeat-%{+YYYY.MM.dd}” } #file { # path => “/test/bak/test.txt” #} }

日志多行合并处理—multiline插件(重点)

(1)示例:

① 日志

2018-03-20 10:44:01,523 [33]DEBUG Debug – task request,task Id:1cbb72f1-a5ea-4e73-957c-6d20e9e12a7a,start time:2018-03-20 10:43:59Request String : {“UserName”:“15046699923”,“Pwd”:“ZYjyh727”,“DeviceType”:2,“DeviceId”:“PC-20170525SADY”,“EquipmentNo”:null,“SSID”:“pc”,“RegisterPhones”:null,“AppKey”:“ab09d78e3b2c40b789ddfc81674bc24deac”,“Version”:“2.0.5.3”} — EndResponse String : {“ErrorCode”:0,“Success”:true,“ErrorMsg”:null,“Result”:null,“WaitInterval”:30} — End

② logstash grok 对合并后多行的处理(合并多行后续都一样,如下)

filter {   grok {     match => {       “message” => “^%{TIMESTAMP_ISO8601:InsertTime}\ .*-\ task\ request,.*,start\ time:%{TIMESTAMP_ISO8601:RequestTime}\n–\ Request\ String\ :\ \{\”UserName\“:\”%{NUMBER:UserName:int}\“,\”Pwd\“:\”(?<Pwd>.*)\“,\”DeviceType\“:%{NUMBER:DeviceType:int},\”DeviceId\“:\”(?<DeviceId>.*)\“,\”EquipmentNo\“:(?<EquipmentNo>.*),\”SSID\“:(?<SSID>.*),\”RegisterPhones\“:(?<RegisterPhones>.*),\”AppKey\“:\”(?<AppKey>.*)\“,\”Version\“:\”(?<Version>.*)\“\}\ –\ \End\n–\ Response\ String\ :\ \{\”ErrorCode\“:%{NUMBER:ErrorCode:int},\”Success\“:(?<Success>[a-z]*),\”ErrorMsg\“:(?<ErrorMsg>.*),\”Result\“:(?<Result>.*),\”WaitInterval\“:%{NUMBER:WaitInterval:int}\}\ –\ \End”     }   } }

(2)在filebeat中使用multiline 插件(推荐)

① 介绍multiline

pattern:正则匹配从哪行合并

negate:true/false,匹配到pattern 部分开始合并,还是不配到的合并

match:after/before(需自己理解)

after:匹配到pattern 部分后合并,注意:这种情况最后一行日志不会被匹配处理

before:匹配到pattern 部分前合并(推荐)

② 5.5版本之后(before为例)

filebeat.prospectors:input_type: log paths:/root/performanceTrace* fields: type: zidonghualog multiline.pattern: ‘.*\”WaitInterval\”:.*–\ End’ multiline.negate: true multiline.match: before

③ 5.5版本之前(after为例)

filebeat.prospectors: – input_type: log paths: /root/performanceTrace* input_type: log multiline: pattern: ‘^20.*’ negate: true match: after

(3)在logstash input中使用multiline 插件(没有filebeat 时推荐)

① 介绍multiline

pattern:正则匹配从哪行合并

negate:true/false,匹配到pattern 部分开始合并,还是不配到的合并

what:previous/next(需自己理解)

previous:相当于filebeat 的after

next:相当于filebeat 的before

② 用法

input { file { path => [“/root/logs/log2”] start_position => “beginning” codec => multiline { pattern => “^20.*” negate => true what => “previous” } } }

(4)在logstash filter中使用multiline 插件(不推荐)

(a)不推荐的原因:

① filter设置multiline后,pipline worker会自动将为1

② 5.5 版本官方把multiline 去除了,要使用的话需下载,下载命令如下:

/usr/share/logstash/bin/logstash-plugin install logstash-filter-multiline

(b)示例:

filter { multiline { pattern => “^20.*” negate => true what => “previous” } }

logstash filter 中的date使用

(1) 日志示例

2018-03-20 10:44:01 [33]DEBUG Debug task request,task Id:1cbb72f1-a5ea-4e73-957c-6d20e9e12a7a,start time:2018-03-20 10:43:59

(2) date 使用

date { match => [“InsertTime”,“YYYY-MM-dd HH:mm:ss “] remove_field => “InsertTime” }

注:

match => [“timestamp” ,“dd/MMM/YYYY H:m:s Z”]

匹配这个字段,字段的格式为:日日/月月月/年年年年 时/分/秒 时区

也可以写为:match => [“timestamp”,”ISO8601″](推荐)

(3)date 介绍

就是将匹配日志中时间的key 替换为@timestamp 的时间,因为@timestamp 的时间是日志送到logstash 的时间,并不是日志中真正的时间。

对多类日志分类处理(重点)

① 在filebeat 的配置中添加type 分类

filebeat: prospectors: – paths: – /mnt/data_total/WebApiDebugLog.txt* fields: type: WebApiDebugLog_total – paths: – /mnt/data_request/WebApiDebugLog.txt* fields: type: WebApiDebugLog_request – paths: – /mnt/data_report/WebApiDebugLog.txt* fields: type: WebApiDebugLog_report

② 在logstash filter中使用if,可进行对不同类进行不同处理

filter { if [fields][type] == “WebApiDebugLog_request” { #对request 类日志 if ([message] =~ “^20.*-\ task\ report,.*,start\ time.*”) { #删除report 行 drop {} } grok { match => {“… …”} } }

③ 在logstash output中使用if

if [fields][type] == “WebApiDebugLog_total” { elasticsearch { hosts => [“6.6.6.6:9200”] index => “logstashl-WebApiDebugLog_total-%{+YYYY.MM.dd}” document_type => “WebApiDebugLog_total_logs” }

对elk 整体性能的优化

2|1性能分析

(1)服务器硬件Linux:1cpu 4GRAM

假设每条日志250 Byte

(2)分析

① logstash硬件Linux:1cpu 4GRAM

每秒500条日志 去掉ruby每秒660条日志 去掉grok后每秒1000条数据

② filebeat硬件Linux:1cpu 4GRAM

每秒2500-3500条数据 每天每台机器可处理:24h*60min*60sec*3000*250Byte=64,800,000,000Bytes,约64G

③ 瓶颈在logstash 从redis中取数据存入ES,开启一个logstash,每秒约处理6000条数据;开启两个logstash,每秒约处理10000条数据(cpu已基本跑满);

④ logstash的启动过程占用大量系统资源,因为脚本中要检查java、ruby以及其他环境变量,启动后资源占用会恢复到正常状态。

2|2关于收集日志的选择:logstash/filter

(1)没有原则要求使用filebeat或logstash,两者作为shipper的功能是一样的,区别在于:

① logstash由于集成了众多插件,如grok,ruby,所以相比beat是重量级的;

② logstash启动后占用资源更多,如果硬件资源足够则无需考虑二者差异;

③ logstash基于JVM,支持跨平台;而beat使用golang编写,AIX不支持;

④ AIX 64bit平台上需要安装jdk(jre) 1.7 32bit,64bit的不支持;

⑤ filebeat可以直接输入到ES,但是系统中存在logstash直接输入到ES的情况,这将造成不同的索引类型造成检索复杂,最好统一输入到els 的源。

(2)总结

logstash/filter 总之各有千秋,但是,我推荐选择:在每个需要收集的日志服务器上配置filebeat,因为轻量级,用于收集日志;再统一输出给logstash,做对日志的处理;最后统一由logstash 输出给es。中间也开增加kafka消息队列进行缓存。

2|3logstash的优化相关配置

(1)可以优化的参数,可根据自己的硬件进行优化配置

① pipeline 线程数,官方建议是等于CPU内核数

默认配置 —> pipeline.workers: 2

可优化为 —> pipeline.workers: CPU内核数(或几倍cpu内核数)

② 实际output 时的线程数

默认配置 —> pipeline.output.workers: 1

可优化为 —> pipeline.output.workers: 不超过pipeline 线程数

③ 每次发送的事件数

默认配置 —> pipeline.batch.size: 125

可优化为 —> pipeline.batch.size: 1000

④ 发送延时

默认配置 —> pipeline.batch.delay: 5

可优化为 —> pipeline.batch.size: 10

(2)总结

通过设置-w参数指定pipeline worker数量,也可直接修改配置文件logstash.yml。这会提高filter和output的线程数,如果需要的话,将其设置为cpu核心数的几倍是安全的,线程在I/O上是空闲的。

默认每个输出在一个pipeline worker线程上活动,可以在输出output 中设置workers设置,不要将该值设置大于pipeline worker数。

还可以设置输出的batch_size数,例如ES输出与batch size一致。

filter设置multiline后,pipline worker会自动将为1,如果使用filebeat,建议在beat中就使用multiline,如果使用logstash作为shipper,建议在input 中设置multiline,不要在filter中设置multiline。

(3)Logstash中的JVM配置文件

Logstash是一个基于Java开发的程序,需要运行在JVM中,可以通过配置jvm.options来针对JVM进行设定。比如内存的最大最小、垃圾清理机制等等。JVM的内存分配不能太大不能太小,太大会拖慢操作系统。太小导致无法启动。默认如下:

-Xms256m # 最小使用内存 -Xmx1g # 最大使用内存

2|4引入Redis 的相关问题

(1)filebeat可以直接输入到logstash(indexer),但logstash没有存储功能,如果需要重启需要先停所有连入的beat,再停logstash,造成运维麻烦;另外如果logstash发生异常则会丢失数据;引入Redis作为数据缓冲池,当logstash异常停止后可以从Redis的客户端看到数据缓存在Redis中;

(2)Redis可以使用list(最长支持4,294,967,295条)或发布订阅存储模式;

(3)redis 做elk 缓冲队列的优化:

​ ① bind 0.0.0.0 #不要监听本地端口

​ ② requirepass ilinux.io #加密码,为了安全运行

​ ③ 只做队列,没必要持久存储,把所有持久化功能关掉:快照(RDB文件)和追加式文件(AOF文件),性能更好

  save “” 禁用快照   appendonly no 关闭RDB

​ ④ 把内存的淘汰策略关掉,把内存空间最大

  maxmemory 0 #maxmemory为0的时候表示我们对Redis的内存使用没有限制

2|5elasticsearch 节点优化配置

(1)服务器硬件配置,OS 参数

(a) /etc/sysctl.conf 配置

vim /etc/sysctl.conf
vm.swappiness = 1 # ES 推荐将此参数设置为 1,大幅降低 swap 分区的大小,强制最大程度的使用内存,注意,这里不要设置为 0, 这会很可能会造成 OOM net.core.somaxconn = 65535 # 定义了每个端口最大的监听队列的长度 vm.max_map_count= 262144 # 限制一个进程可以拥有的VMA(虚拟内存区域)的数量。虚拟内存区域是一个连续的虚拟地址空间区域。当VMA 的数量超过这个值,OOM fs.file-max = 518144 # 设置 Linux 内核分配的文件句柄的最大数量
[root@elasticsearch]# sysctl -p 生效一下

(b)limits.conf 配置

vim /etc/security/limits.conf elasticsearch soft nofile 65535 elasticsearch hard nofile 65535 elasticsearch soft memlock unlimited elasticsearch hard memlock unlimited

(c)为了使以上参数永久生效,还要设置两个地方

vim /etc/pam.d/common-session-noninteractive vim /etc/pam.d/common-session 添加如下属性: session required pam_limits.so 可能需重启后生效

(2)elasticsearch 中的JVM配置文件

-Xms2g -Xmx2g

① 将最小堆大小(Xms)和最大堆大小(Xmx)设置为彼此相等。

② Elasticsearch可用的堆越多,可用于缓存的内存就越多。但请注意,太多的堆可能会使您长时间垃圾收集暂停。

③ 设置Xmx为不超过物理RAM的50%,以确保有足够的物理内存留给内核文件系统缓存。

④ 不要设置Xmx为JVM用于压缩对象指针的临界值以上;确切的截止值有所不同,但接近32 GB。不要超过32G,如果空间大,多跑几个实例,不要让一个实例太大内存

(3)elasticsearch 配置文件优化参数

① vim elasticsearch.yml

bootstrap.memory_lock: true #锁住内存,不使用swap #缓存、线程等优化如下 bootstrap.mlockall: true transport.tcp.compress: true indices.fielddata.cache.size: 40% indices.cache.filter.size: 30% indices.cache.filter.terms.size: 1024mb threadpool: search: type: cached size: 100 queue_size: 2000

② 设置环境变量

vim /etc/profile.d/elasticsearch.sh export ES_HEAP_SIZE=2g #Heap Size不超过物理内存的一半,且小于32G

(4)集群的优化(我未使用集群)

① ES是分布式存储,当设置同样的cluster.name后会自动发现并加入集群;

② 集群会自动选举一个master,当master宕机后重新选举;

③ 为防止”脑裂”,集群中个数最好为奇数个

④ 为有效管理节点,可关闭广播 discovery.zen.ping.multicast.enabled: false,并设置单播节点组discovery.zen.ping.unicast.hosts: [“ip1”, “ip2”, “ip3”]

2|6性能的检查

(1)检查输入和输出的性能

Logstash和其连接的服务运行速度一致,它可以和输入、输出的速度一样快。

(2)检查系统参数

① CPU

注意CPU是否过载。在Linux/Unix系统中可以使用top -H查看进程参数以及总计。

如果CPU使用过高,直接跳到检查JVM堆的章节并检查Logstash worker设置。

② Memory

注意Logstash是运行在Java虚拟机中的,所以它只会用到你分配给它的最大内存。

检查其他应用使用大量内存的情况,这将造成Logstash使用硬盘swap,这种情况会在应用占用内存超出物理内存范围时。

③ I/O 监控磁盘I/O检查磁盘饱和度

使用Logstash plugin(例如使用文件输出)磁盘会发生饱和。

当发生大量错误,Logstash生成大量错误日志时磁盘也会发生饱和。

在Linux中,可使用iostat,dstat或者其他命令监控磁盘I/O

④ 监控网络I/O

当使用大量网络操作的input、output时,会导致网络饱和。

在Linux中可使用dstat或iftop监控网络情况。

(3)检查JVM heap

heap设置太小会导致CPU使用率过高,这是因为JVM的垃圾回收机制导致的。

一个快速检查该设置的方法是将heap设置为两倍大小然后检测性能改进。不要将heap设置超过物理内存大小,保留至少1G内存给操作系统和其他进程。

你可以使用类似jmap命令行或VisualVM更加精确的计算JVM heap

发表评论

您的电子邮箱地址不会被公开。 必填项已用*标注