You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-issues@hadoop.apache.org by "Chris Nauroth (JIRA)" <ji...@apache.org> on 2015/01/06 22:00:34 UTC

[jira] [Commented] (HADOOP-11464) Reinstate support for launching Hadoop processes on Windows using Cygwin.

    [ https://issues.apache.org/jira/browse/HADOOP-11464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14266757#comment-14266757 ] 

Chris Nauroth commented on HADOOP-11464:
----------------------------------------

Prior to the Windows compatibility work, there were 2 main use cases for Cygwin in Hadoop: 1) as a Linux emulation layer and 2) as an alternative shell for launching commands.  We learned that #1 cannot work, at least not completely.  We really needed native code calls into the Windows API to achieve the semantics required of the Hadoop code.  Based on that, we dropped Cygwin support.  However, perhaps we were too hasty in eliminating support for #2.  For example, some users really would like to run a Cygwin bash shell to launch MapReduce jobs.  We can make that work easily with a small patch on the bash scripts.

> Reinstate support for launching Hadoop processes on Windows using Cygwin.
> -------------------------------------------------------------------------
>
>                 Key: HADOOP-11464
>                 URL: https://issues.apache.org/jira/browse/HADOOP-11464
>             Project: Hadoop Common
>          Issue Type: Bug
>          Components: bin
>            Reporter: Chris Nauroth
>            Assignee: Chris Nauroth
>
> The Windows compatibility work removed the dependency on Cygwin as a compatibility layer.  However, some users may still wish to run Cygwin on a personal machine for running Hadoop commands that interact with a cluster.  This issue proposes to reinstate support for launching Hadoop processes on Windows using Cygwin to call the bash scripts.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)