[
https://issues.apache.org/jira/browse/HADOOP-4633?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alok Parlikar updated HADOOP-4633:
----------------------------------
Description:
If a streaming job fails in the map/reduce phase, streaming exits with a
non-zero status. However, if it fails before the map-reduce started (such as
output path already exists), streaming still exits with a zero status.
was:
If a streaming job fails in the map/reduce phase, streaming exits with a
non-zero status. However, if it fails before the map-reduce started (such as
output path already exists), streaming still exits with a zero status.
A probable fix to this might be in StreamJob.java:
-------------------------------------------------------------------
912c912
<
---
> int jobExitCode = 0;
950c950
< return 1;
---
> jobExitCode = 1;
957c957
< return 2;
---
> jobExitCode = 2;
960c960
< return 3;
---
> jobExitCode = 3;
964c964
< return 4;
---
> jobExitCode = 4;
967c967
< return 5;
---
> jobExitCode = 5;
975c975
< return 0;
---
> return jobExitCode;
-------------------------------------------------------------------
> Streaming exits with zero-status in cases even on failure
> ---------------------------------------------------------
>
> Key: HADOOP-4633
> URL: https://issues.apache.org/jira/browse/HADOOP-4633
> Project: Hadoop Core
> Issue Type: Bug
> Components: contrib/streaming
> Affects Versions: 0.18.2
> Environment: Linux
> Reporter: Alok Parlikar
> Original Estimate: 2h
> Remaining Estimate: 2h
>
> If a streaming job fails in the map/reduce phase, streaming exits with a
> non-zero status. However, if it fails before the map-reduce started (such as
> output path already exists), streaming still exits with a zero status.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
