[
https://issues.apache.org/jira/browse/FLUME-2282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13855444#comment-13855444
]
Hari Shreedharan commented on FLUME-2282:
-----------------------------------------
That is possible - in which case the errors are really valid. Spool dir runs
threads which write to the channel, so it is possible the thread had not died
before the channel closed. In SpoolDir source, it can be fixed by using a
CyclicBarrier.await() in the stop method, and another call at the end of the
SpoolDirSourceRunnable.
> File Channel throws IllegalStateException causing some sources to not be to
> recover from a reconfig
> ---------------------------------------------------------------------------------------------------
>
> Key: FLUME-2282
> URL: https://issues.apache.org/jira/browse/FLUME-2282
> Project: Flume
> Issue Type: Bug
> Components: Configuration, File Channel
> Reporter: Roshan Naik
> Labels: channel
>
> when the config file is updated and the file channel is shut, the Source
> receives a java.lang.IllegalState exception from FileChannel. Unfortunately
> unless sources are designed to catch all exception types, they will not know
> this is a channel exception and likely to go into fuzzy states.
> For instance the FLUME-2255 patch catches only ChannelException.
> The below stack trace can be seen in logs when this happens with FLUME-2255
> applied:
> {code}
> 21 Dec 2013 21:15:03,030 ERROR [pool-17-thread-1]
> (org.apache.flume.source.SpoolDirectorySource$SpoolDirectoryRunnable.run:194)
> - Uncaught exception in SpoolDirSource thread
> java.lang.IllegalStateException: Channel closed [channel=channel_nontx]
> at
> org.apache.flume.channel.file.FileChannel.createTransaction(FileChannel.java:351)
> at
> org.apache.flume.channel.BasicChannelSemantics.getTransaction(BasicChannelSemantics.java:122)
> at
> org.apache.flume.channel.ChannelProcessor.processEventBatch(ChannelProcessor.java:181)
> at
> org.apache.flume.source.SpoolDirectorySource$SpoolDirectoryRunnable.run(SpoolDirectorySource.java:179)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
> at
> java.util.concurrent.FutureTask$Sync.innerRunAndReset(FutureTask.java:351)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:178)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:165)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:267)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1146)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
> at java.lang.Thread.run(Thread.java:679)
> {code}
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
