[
https://issues.apache.org/jira/browse/MAHOUT-809?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13104570#comment-13104570
]
Hudson commented on MAHOUT-809:
-------------------------------
Integrated in Mahout-Quality #1037 (See
[https://builds.apache.org/job/Mahout-Quality/1037/])
MAHOUT-809 avoid overwriting prior chunk when updating to next writer
srowen : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1170629
Files :
*
/mahout/trunk/integration/src/main/java/org/apache/mahout/text/ChunkedWriter.java
> Bad bug in ChunkedWriter
> ------------------------
>
> Key: MAHOUT-809
> URL: https://issues.apache.org/jira/browse/MAHOUT-809
> Project: Mahout
> Issue Type: Bug
> Components: Clustering
> Affects Versions: 0.5, 0.6
> Environment: Occurred on single node, not tested on cluster.
> Reporter: Florian Bausch
> Assignee: Sean Owen
> Fix For: 0.6
>
> Original Estimate: 0.25h
> Remaining Estimate: 0.25h
>
> org.apache.mahout.text.ChunkedWriter has a bug, that causes data loss, if the
> maximal chunk size is reached. The first chunk is overwritten, then it
> continues normally.
> This is caused in line 58:
> writer = new SequenceFile.Writer(fs, conf, getPath(currentChunkID++),
> Text.class, Text.class);
> The fix should look like this:
> writer = new SequenceFile.Writer(fs, conf, getPath(++currentChunkID),
> Text.class, Text.class);
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
