[
https://issues.apache.org/jira/browse/HADOOP-1558?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12512206
]
Arun C Murthy commented on HADOOP-1558:
---------------------------------------
I firmly am on Doug's side of the fence about the need keep the kernel free of
user-code, however w.r.t to this issue I'd like to bring some complications to
everyone's attention:
Eventually we need to move {{Task.saveOutput}} and {{Task.discardOutput}} to
the OutputFormats; however this means that we *have* to call these methods from
the {{JobTracker}} (only there do we have a global picture of the tasks and the
job), there-by ruling out these being done at the child-jvms since I believe
doing this isn't feasible performance-wise (I'd love to hear
thoughts/arguments/ideas); hence I'd agree with Alejandro's take on static
output-file handlers which cannot be user-supplied or user-overridden for now.
> changes to OutputFormat to work on temporary directory to enable re-running
> crashed jobs (Issue: 1121)
> ------------------------------------------------------------------------------------------------------
>
> Key: HADOOP-1558
> URL: https://issues.apache.org/jira/browse/HADOOP-1558
> Project: Hadoop
> Issue Type: Improvement
> Components: mapred
> Environment: all
> Reporter: Alejandro Abdelnur
> Fix For: 0.14.0
>
> Attachments: hadoop-1558-JUN1007-1934.txt,
> hadoop-1558-JUN1107-1533.txt
>
>
> Add OutputFormat methods like:
> /** Called to initialize output for this job. */
> void initialize(JobConf job) throws IOException;
> /** Called to finalize output for this job. */
> void commit(JobConf job) throws IOException;
> In the base implemenation for FileSystem output, initialize() might then
> create a temporary directory for the job, removing any that already exists,
> and commit could rename the temporary output directory to the final name.
> The existing checkOutputSpecs() would continue to throw an exception if the
> final output already exists.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
