[jira] Resolved: (MAPREDUCE-1795) add error option if file-based record-readers fail to consume all input (e.g., concatenated gzip, bzip2)

["Greg Roelofs (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/MAPREDUCE-1795?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Greg Roelofs resolved MAPREDUCE-1795.
-------------------------------------

    Resolution: Won't Fix

Per previous comment, we're going to fix the underlying issue instead (i.e., make decompressors support concatenated streams).  See MAPREDUCE-469.

> add error option if file-based record-readers fail to consume all input (e.g., concatenated gzip, bzip2)
> --------------------------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-1795
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1795
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>            Reporter: Greg Roelofs
>            Assignee: Greg Roelofs
>
> When running MapReduce with concatenated gzip files as input, only the first part ("member" in gzip spec parlance, http://www.ietf.org/rfc/rfc1952.txt) is read; the remainder is silently ignored.  As a first step toward fixing that, this issue will add a configurable option to throw an error in such cases.
> MAPREDUCE-469 is the tracker for the more complete fix/feature, whenever that occurs.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.