flume file channel 异常解决

1. 错误提示

　　

-- ::, (SinkRunner-PollingRunner-DefaultSinkProcessor) [ERROR - org.apache.flume.SinkRunner$PollingRunner.run(SinkRunner.java:)] Unable to deliver event. Exception follows.
java.lang.IllegalStateException: Channel closed [channel=channel1]. Due to java.io.EOFException: null
        at org.apache.flume.channel.file.FileChannel.createTransaction(FileChannel.java:)
        at org.apache.flume.channel.BasicChannelSemantics.getTransaction(BasicChannelSemantics.java:)
        at org.apache.flume.sink.hdfs.HDFSEventSink.process(HDFSEventSink.java:)
        at org.apache.flume.sink.DefaultSinkProcessor.process(DefaultSinkProcessor.java:)
        at org.apache.flume.SinkRunner$PollingRunner.run(SinkRunner.java:)
        at java.lang.Thread.run(Thread.java:)
Caused by: java.io.EOFException
        at java.io.RandomAccessFile.readInt(RandomAccessFile.java:)
        at java.io.RandomAccessFile.readLong(RandomAccessFile.java:)
        at org.apache.flume.channel.file.EventQueueBackingStoreFactory.get(EventQueueBackingStoreFactory.java:)
        at org.apache.flume.channel.file.Log.replay(Log.java:)
        at org.apache.flume.channel.file.FileChannel.start(FileChannel.java:)
        at org.apache.flume.lifecycle.LifecycleSupervisor$MonitorRunnable.run(LifecycleSupervisor.java:)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:)
        at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$(ScheduledThreadPoolExecutor.java:)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:)
        ...  more
-- ::, (pool--thread-) [ERROR - org.apache.flume.source.SpoolDirectorySource$SpoolDirectoryRunnable.run(SpoolDirectorySource.java:)] FATAL: Spool Directory source source1: { spoolDir: /home/hadoop_admin/movielog }: Uncaught exception in SpoolDirectorySource thread. Restart or reconfigure Flume to continue processing.
java.lang.IllegalStateException: Channel closed [channel=channel1]. Due to java.io.EOFException: null
        at org.apache.flume.channel.file.FileChannel.createTransaction(FileChannel.java:)
        at org.apache.flume.channel.BasicChannelSemantics.getTransaction(BasicChannelSemantics.java:)
        at org.apache.flume.channel.ChannelProcessor.processEventBatch(ChannelProcessor.java:)
        at org.apache.flume.source.SpoolDirectorySource$SpoolDirectoryRunnable.run(SpoolDirectorySource.java:)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:)
        at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$(ScheduledThreadPoolExecutor.java:)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:)
        at java.lang.Thread.run(Thread.java:)
Caused by: java.io.EOFException
        at java.io.RandomAccessFile.readInt(RandomAccessFile.java:)
        at java.io.RandomAccessFile.readLong(RandomAccessFile.java:)
        at org.apache.flume.channel.file.EventQueueBackingStoreFactory.get(EventQueueBackingStoreFactory.java:)
        at org.apache.flume.channel.file.Log.replay(Log.java:)
        at org.apache.flume.channel.file.FileChannel.start(FileChannel.java:)
        at org.apache.flume.lifecycle.LifecycleSupervisor$MonitorRunnable.run(LifecycleSupervisor.java:)
        ...  more

　配置文件：

　　

agent1.sources = source1
agent1.channels = channel1
agent1.sinks = sink1

# Each channel's type is defined.
agent1.channels.channel1.type = file
agent1.channels.channel1.checkpointDir = /home/hadoop_admin/flumeTemp/fchannel/spool/checkpoint
agent1.channels.channel1.dataDirs = /home/hadoop_admin/flumeTemp/fchannel/spool/data
agent1.channels.channel1.capacity =
agent1.channels.channel1.keep-alive =
agent1.channels.channel1.write-timeout =
agent1.channels.channel1.checkpoint-timeout = 

# For each one of the sources, the type is defined
agent1.sources.source1.type = spooldir
agent1.sources.source1.inputCharset = GBK
agent1.sources.source1.spoolDir =/home/hadoop_admin/movielog
agent1.sources.source1.fileHeader = true
agent1.sources.source1.deletePolicy = immediate
agent1.sources.source1.batchSize =
agent1.sources.source1.channels = channel1

# Each sink's type must be defined
agent1.sinks.sink1.type = hdfs
agent1.sinks.sink1.channel = channel1
agent1.sinks.sink1.hdfs.path = hdfs://master:9000/flumeTest
agent1.sinks.sink1.hdfs.filePrefix = master-
agent1.sinks.sink1.hdfs.writeFormat = Text
agent1.sinks.sink1.hdfs.fileType = DataStream
agent1.sinks.sink1.hdfs.rollInterval =
agent1.sinks.sink1.hdfs.idleTimeout =
agent1.sinks.sink1.hdfs.rollCount =
agent1.sinks.sink1.hdfs.rollSize =
agent1.sinks.sink1.hdfs.batchSize =
agent1.sinks.sink1.hdfs.callTimeout = 

2. 解决方法

　　链接： http://serverfault.com/questions/690588/flume-error-log-while-using-filechannel

　　只要把配置文件中的关于channel配置checkpointDir和dataDir目录清空就可以