[FLUME-2951] Exec Source generate massive logfile in File Channel data dirs

• Type: Bug
• Status: Patch Available
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.6.0
• Component/s: Sinks+Sources
• Assignee: dengkai
• Reporter: dengkai
• Created: 11/Jul/16 05:01
• Updated: 27/Jul/16 07:48

Description

When the file channel is full, exec source always get a filechannel exception.Every time the main thread start to roll, and then the timedFlushService flush the events again.So an exception follows :
07 Jul 2016 17:09:38,789 ERROR [timedFlushExecService63-0] (org.apache.flume.source.ExecSource$ExecRunnable$1.run:328) - Exception occured when processing event batch
org.apache.flume.ChannelException: Commit failed due to IO error [channel=channel-4]
at org.apache.flume.channel.file.FileChannel$FileBackedTransaction.doRollback(FileChannel.java:621)
at org.apache.flume.channel.BasicTransactionSemantics.rollback(BasicTransactionSemantics.java:168)
at org.apache.flume.channel.ChannelProcessor.processEventBatch(ChannelProcessor.java:194)
at org.apache.flume.source.ExecSource$ExecRunnable.flushEventBatch(ExecSource.java:382)
at org.apache.flume.source.ExecSource$ExecRunnable.access$100(ExecSource.java:255)
at org.apache.flume.source.ExecSource$ExecRunnable$1.run(ExecSource.java:324)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:304)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:178)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)
Caused by: java.nio.channels.ClosedByInterruptException
at java.nio.channels.spi.AbstractInterruptibleChannel.end(AbstractInterruptibleChannel.java:202)
at sun.nio.ch.FileChannelImpl.force(FileChannelImpl.java:380)
at org.apache.flume.channel.file.LogFileV3.writeDelimitedTo(LogFileV3.java:148)
at org.apache.flume.channel.file.LogFileV3$Writer.<init>(LogFileV3.java:209)
at org.apache.flume.channel.file.LogFileFactory.getWriter(LogFileFactory.java:77)
at org.apache.flume.channel.file.Log.roll(Log.java:964)
at org.apache.flume.channel.file.Log.roll(Log.java:933)
at org.apache.flume.channel.file.Log.rollback(Log.java:740)
at org.apache.flume.channel.file.FileChannel$FileBackedTransaction.doRollback(FileChannel.java:619)
... 12 more
When this exceptions happens, file channel start another roll, and the ID of the log file keep increasing.
If the exec source is configured as restart = true and file channel is still full, this problem always be there and will generate massive log-ID size about 1.0MB and log-ID.meta size about 47bytes in the channel's data dir(in my try, I even got thousands of files and got too many files error when exec source started tail -F command).

Issue Links

• links to: FLUME-2951-0