我在linux ubuntu上运行了一个flume，它运行得比较好，但是在运行一天后它总是会停止，下面是flume的配置：

nginx.channels=ch-spooling ch-tail
    nginx.sources=spooling-source tail-source
    nginx.sinks=hdfs-spooling kafka-tail

    nginx.channels.ch-spooling.type=file
    nginx.channels.ch-spooling.checkpointDir=/home/hadoop/flume/file-channel/ch-spooling/checkpoint
    nginx.channels.ch-spooling.dataDirs=/home/hadoop/flume/file-channel/ch-spooling/data
    nginx.channels.ch-spooling.capacity=1000
    nginx.channels.ch-spooling.transactionCapacity=100
    nginx.channels.ch-spooling.capacity=100000

    nginx.channels.ch-tail.type=file
    nginx.channels.ch-tail.checkpointDir=/home/hadoop/flume/file-channel/ch-tail/checkpoint
    nginx.channels.ch-tail.dataDirs=/home/hadoop/flume/file-channel/ch-tail/data
    nginx.channels.ch-tail.capacity=1000
    nginx.channels.ch-tail.transactionCapacity=100
    nginx.channels.ch-tail.capacity=100000

    nginx.sources.spooling-source.type=spooldir
    nginx.sources.spooling-source.channels=ch-spooling
    nginx.sources.spooling-source.spoolDir=/usr/local/nginx/logs/flume_logs
    nginx.sources.spooling-source.fileHeader=true
    nginx.sources.spooling-source.fileHeaderKey=file
    nginx.sources.spooling-source.basenameHeader=true
    nginx.sources.spooling-source.basenameHeaderKey=basename
    nginx.sources.spooling-source.deletePolicy=never

    nginx.sources.spooling-source.consumeOrder=oldest
    nginx.sources.spooling-source.recursiveDirectorySearch=false
    nginx.sources.spooling-source.batchSize=100
    nginx.sources.spooling-source.inputCharset=UTF-8

    nginx.sources.spooling-source.decodeErrorPolicy=IGNORE
    nginx.sources.spooling-source.selector.type=replicating
    nginx.sources.spooling-source.interceptors=i1 i2
    nginx.sources.spooling-source.interceptors.i1.type=timestamp
    nginx.sources.spooling-source.interceptors.i2.type=host
    nginx.sources.spooling-source.interceptors.i2.useIP=true
    nginx.sources.spooling-source.interceptors.i2.hostHeader=host

    nginx.sources.tail-source.type=TAILDIR
    nginx.sources.tail-source.channels=ch-tail
    nginx.sources.tail-source.filegroups=hnrapi hnrapierror
    nginx.sources.tail-source.filegroups.hnrapi=/usr/local/nginx/logs/hnr-api.access.log
    nginx.sources.tail-source.filegroups.hnrapierror=/usr/local/nginx/logs/error.log
    nginx.sources.tail-source.positionFile=/home/hadoop/flume/file-channel/ch-tail/taildir_position.json
    nginx.sources.tail-source.headers.hnrapi.topic=hnrapi
    nginx.sources.tail-source.headers.hnrapierror.topic=hnrapierror
    nginx.sources.tail-source.skipToEnd=true
    nginx.sources.tail-source.interceptors=i1 i2
    nginx.sources.tail-source.interceptors.i1.type=timestamp
    nginx.sources.tail-source.interceptors.i2.type=host
    nginx.sources.tail-source.interceptors.i2.useIP=true
    nginx.sources.tail-source.interceptors.i2.hostHeader=host

    nginx.sinks.hdfs-spooling.channel=ch-spooling
    nginx.sinks.hdfs-spooling.type=hdfs
    nginx.sinks.hdfs-spooling.hdfs.fileType=DataStream
    nginx.sinks.hdfs-spooling.hdfs.writeFormat=Text

    nginx.sinks.hdfs-spooling.hdfs.path=hdfs://namenode1:9000/flume/nginx/%Y-%m-%d

    nginx.sinks.hdfs-spooling.hdfs.filePrefix=%{basename}.[%{host}]
    nginx.sinks.hdfs-spooling.hdfs.fileSuffix=
    nginx.sinks.hdfs-spooling.hdfs.inUseSuffix=.tmp
    nginx.sinks.hdfs-spooling.hdfs.rollInterval=0
    nginx.sinks.hdfs-spooling.hdfs.rollSize=1073741824
    nginx.sinks.hdfs-spooling.hdfs.rollCount=0
    nginx.sinks.hdfs-spooling.hdfs.idleTimeout=60

    nginx.sinks.kafka-tail.channel=ch-tail
    nginx.sinks.kafka-tail.type=org.apache.flume.sink.kafka.KafkaSink
    nginx.sinks.kafka-tail.kafka.bootstrap.servers=192.47.180.63:9192,192.27.125.123:9192,192.27.124.96:9192
    nginx.sinks.kafka-tail.flumeBatchSize=32
    nginx.sinks.kafka-tail.kafka.producer.acks=1
    nginx.sinks.kafka-tail.useFlumeEventFormat=false

我用taildir和spooldir收集nginx日志，taildir有target到kafka，spooldir有target到hdfs。
以下是Flume监视器日志的一部分：
2017-03-06 18:17:54302（kafka producer网络线程| producer-1）[调试-org.apache.flume.sink.kafka.sinkcallback.oncompletion（kafkasink。java:456)]发送时间：13 2017-03-06 18:18:14，230（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:18:14，408（log backgroundworker ch tail）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.begincheckpoint（eventqueuebackingstorefile）。java:227)]启动/home/hadoop/flume/file channel/ch tail/checkpoint/checkpoint的检查点，要同步的元素=8 2017-03-06 18:18:14，415（log backgroundworker ch tail）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.checkpoint（eventqueuebackingstorefile）。java:252)]正在更新检查点元数据：logwriteorderid:1488782535982，queuesize:0，queuehead:1642 2017-03-06 18:18:14，420（log backgroundworker ch tail）[info-org.apache.flume.channel.file.log.writecheckpoint（日志。java:1052)]更新了文件的检查点：/home/hadoop/flume/file channel/ch tail/data/log-16位置：391719 logwriteorderid:1488782535982 2017-03-06 18:18:14，420（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.log.removeoldlogs（日志。java:1115)]当前使用的文件：[16]2017-03-06 18:18:14，484（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:18:44，230（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]不需要检查点2017-03-06 18:18:44420（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:18:44，484（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:19:14，230（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]不需要检查点2017-03-06 18:19:14421（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:19:14，485（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:19:44，231（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]不需要检查点2017-03-06 18:19:44421（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:19:44，485（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:19:54，577（pollablesourcerunner taildirsource tail source）[info-org.apache.flume.source.taildir.taildirsource.closetailfiles（taildirsource）。java:288)]关闭文件：/usr/local/nginx/logs/hnr-api.access.log，inode:2237741，pos:3183 2017-03-06 18:19:59，307（sinkrunner pollingrunner defaultsinkprocessor）[debug-org.apache.flume.sink.kafka.kafkasink.process（kafkasink）。java:187)]事件#0 2017-03-06 18:19:59310（kafka producer网络线程| producer-1）[调试-org.apache.flume.sink.kafka.sinkcallback.oncompletion（kafkasink。java:455)]确认的消息partition:1 ofset:122 2017-03-06 18:19:59,310（kafka producer网络线程| producer-1）[debug-org.apache.flume.sink.kafka.sinkcallback.oncompletion（kafkasink）。java:456)]发送时间：3 2017-03-06 18:20:14，231（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:20:14，421（log backgroundworker ch tail）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.begincheckpoint（eventqueuebackingstorefile）。java:227)]为/home/hadoop/flume/file channel/ch tail/checkpoint/checkpoint启动检查点，要同步的元素=1 2017-03-06 18:20:14，427（log backgroundworker ch tail）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.checkpoint（eventqueuebackingstorefile）。java:252)]正在更新检查点元数据：logwriteorderid:1488782535987，queuesize:0，queuehead:1642 2017-03-06 18:20:14，432（log backgroundworker ch tail）[info-org.apache.flume.channel.file.log.writecheckpoint（日志。java:1052)]更新了文件的检查点：/home/hadoop/flume/file channel/ch tail/data/log-16位置：392302 logwriteorderid:1488782535987 2017-03-06 18:20:14，432（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.log.removeoldlogs（日志。java:1115)]当前使用的文件：[16]2017-03-06 18:20:14，485（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:20:44，231（log backgroundworker ch spooling）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]不需要检查点2017-03-06 18:20:44433（log backgroundworker ch tail）[debug-org.apache.flume.channel.file.flumeeventqueue.checkpoint（flumeeventqueue）。java:138)]无需检查站2017-03-06 18:20:44，485（conf-file-poller-0）[debug-org.apache.flume.node.pollingpropertiesfileconfigurationprovider$filewatcherrunnable.run（pollingpropertiesfileconfigurationprovider）。java:127)]正在检查文件：/home/hadoop/flume/conf/flume-conf.properties的更改2017-03-06 18:20:46，224（代理关闭钩子）[info-org.apache.flume.lifecycle.lifecyclesupervisor.stop（lifecyclesupervisor。java:78)]停止生命周期主管13 2017-03-06 18:20:46，227（代理关闭钩子）[info-org.apache.flume.channel.file.filechannel.stop（filechannel）。java:324)]正在停止filechannel ch tail{datadirs:[/home/hadoop/flume/file channel/ch tail/data]}。。。2017-03-06 18:20:46227（代理关闭挂钩）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.begincheckpoint（eventqueuebackingstorefile）。java:227)]启动/home/hadoop/flume/file channel/ch tail/checkpoint/checkpoint的检查点，要同步的元素=0 2017-03-06 18:20:46，233（代理关闭挂钩）[info-org.apache.flume.channel.file.eventqueuebackingstorefile.checkpoint（eventqueuebackingstorefile）。java:252)]正在更新检查点元数据：logwriteorderid:1488782535988，queuesize:0，queuehead:1642 2017-03-06 18:20:46，238（代理关闭钩子）[info-org.apache.flume.channel.file.log.writecheckpoint（日志。java:1052)]更新了文件的检查点：/home/hadoop/flume/file channel/ch tail/data/log-16位置：392302 logwriteorderid:1488782535988正在尝试关闭后台工作进程。2017-03-06 18:20:46238（代理关闭挂钩）[info-org.apache.flume.channel.file.log.shutdownworker（日志。java:868)]正在尝试关闭后台工作进程。2017-03-06 18:20:46238（代理关闭挂钩）[info-org.apache.flume.channel.file.logfile$writer.close（日志文件）。java:384)]关闭/home/hadoop/flume/file channel/ch tail/data/log-16 2017-03-06 18:20:46，238（代理关闭钩子）[info-org.apache.flume.channel.file.logfile$randomreader.close（日志文件）。java:520)]关闭randomreader/home/hadoop/flume/file channel/ch tail/data/log-16 2017-03-06 18:20:46，243（代理关闭钩子）[info-org.apache.flume.channel.file.logfile$randomreader.close（日志文件）。java:520)]关闭randomreader/home/hadoop/flume/file channel/ch tail/data/log-15 2017-03-06 18:20:46，249（代理关闭钩子）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:149)]组件类型：通道，名称：ch tail stopped 2017-03-06 18:20:46，249（代理关闭钩子）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:155)]类型：channel，名称：ch tail的关闭度量。channel.start.time==1488782534266 2017-03-06 18:20:46249（代理关闭钩子）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:161)]类型：channel，名称：ch tail的关闭度量。channel.stop.time==1488795646249 2017-03-06 18:20:46249（代理关闭挂钩）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:177)]类型：channel，名称：ch tail的关闭度量。channel.capacity==100000 2017-03-06 18:20:46249（代理关闭挂钩）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:177)]类型：channel，名称：ch tail的关闭度量。channel.current.size==0 2017-03-06 18:20:46249（代理关闭挂钩）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（monitoredcountergroup）。java:177)]类型：channel，名称：ch tail的关闭度量。channel.event.put.attempt==834 2017-03-06 18:20:46249（代理关闭挂钩）[info-org.apache.flume.instrumentation.monitoredcountergroup.stop（监视计数器组）。java:177)]类型：channel，名称：ch tail的关闭度量。channel.event.put.success==834 2017-03-06 18:20:46249（代理关闭挂钩）[info-org.apache.flume.instrumentation.monit