[ 
https://issues.apache.org/jira/browse/FLUME-1683?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13492199#comment-13492199
 ] 

Hudson commented on FLUME-1683:
-------------------------------

Integrated in flume-trunk #322 (See 
[https://builds.apache.org/job/flume-trunk/322/])
    FLUME-1683: Fix Time Granularity Bug in SpoolingFileLineReader (Revision 
c145b2b8c1ed7bc04e1b018085c519555cf61a21)

     Result = ABORTED
brock : 
http://git-wip-us.apache.org/repos/asf/flume/repo?p=flume.git&a=commit&h=c145b2b8c1ed7bc04e1b018085c519555cf61a21
Files : 
* 
flume-ng-core/src/main/java/org/apache/flume/client/avro/SpoolingFileLineReader.java
* 
flume-ng-core/src/test/java/org/apache/flume/client/avro/TestSpoolingFileLineReader.java

                
> Fix Time Granularity Bug in SpoolingFileLineReader
> --------------------------------------------------
>
>                 Key: FLUME-1683
>                 URL: https://issues.apache.org/jira/browse/FLUME-1683
>             Project: Flume
>          Issue Type: Bug
>            Reporter: Patrick Wendell
>            Assignee: Patrick Wendell
>             Fix For: v1.3.0
>
>         Attachments: FLUME-1683.v1.txt, FLUME-1683.v2.txt
>
>
> This patch fixes a test in the SpoolingFileLineReader, and a bug associated 
> with that test:
> The issue is related to the way that the source sorts files which are 
> candidates for ingestion. It sorts the files by timestamp, and the unit test 
> assumes that if it creates two files:
> x = new File()
> Files.write("something", x)
> y = new File() 
> Files.write("something", y)
> // Read from files
> that x will be processed before y. In my filesystem, the time granularity is 
> only 1 second, so these files were actually processed in a non-deterministic 
> order. On your filesystem, they were processed in the correct order, and the 
> test (correctly) failed.
> I didn't notice that the test was mixed up because it was passing for me due 
> to this issue.
> I updated the sorting order to take into account filename when there is a tie 
> in the timestamp, so at least we have a consistent ordering we can test.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to