[jira] Commented: (PIG-1869) Job conf parameters not passed to Streaming

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

    [ https://issues.apache.org/jira/browse/PIG-1869?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12999209#comment-12999209 ] 

Ankur commented on PIG-1869:
----------------------------

In my streaming binary, I would like to get hold of the parameter "map.input.file" that give information about the split file being processed. Also, starting with release 0.8, pig automatically combines multiple files into a single split, it is then responsibility of Pig to set this parameter correctly while reading individual split in combined split.

> Job conf parameters not passed to Streaming
> -------------------------------------------
>
>                 Key: PIG-1869
>                 URL: https://issues.apache.org/jira/browse/PIG-1869
>             Project: Pig
>          Issue Type: Bug
>    Affects Versions: 0.8.0
>            Reporter: Ankur
>
> Pig Streaming does not set the job conf parameters as environment variables for the streaming binary. This is unlike Hadoop streaming where all the job conf parameters are available as environment variables.
> Here is a small test script along with the streaming binary
> === stream.pig ===
> A = LOAD 'dat' as (f1:chararray);
> DEFINE MY_BIN `./stream.sh` SHIP('./stream.sh');
> B = STREAM A THROUGH MY_BIN;
> dump B; 
> === stream.sh ===
> #!/bin/bash
> export

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira