[jira] Updated: (MAPREDUCE-931) rumen should use its own interpolation classes to create runtimes for simulated tasks

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

     [ https://issues.apache.org/jira/browse/MAPREDUCE-931?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dick King updated MAPREDUCE-931:
--------------------------------

    Attachment: patch-931.patch

There is no separate test case for this patch, which is a code cleanup.  ZombieJob was introduced in MAPREDUCE-751's patch, but it had an ad hoc interpolation engine.  We refactored it here to use the new interpolation engine.  This can affect performance during simulations, because ZombieJob created a whole new local interpolation table every time it needed an interpolation, even if it's identical to the one that would have been created for another interpolation in the same job.  We now only create one interpolator for each job.

TestZombieJob and TestPiecewiseLinearInterpolation are test cases for this technology.

> rumen should use its own interpolation classes to create runtimes for simulated tasks
> -------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-931
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-931
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>            Reporter: Dick King
>            Assignee: Dick King
>            Priority: Minor
>         Attachments: patch-931.patch
>
>
> Currently, when a simulator or benchmark is running and simulating hadoop jobs using rumen data, and rumen's runtime system is used to get execution times for the tasks in the simulated jobs, rumen would use some ad hoc code, despite the fact that rumen has a perfectly good interpolation framework to generate random variables that fit discrete CDFs.
> We should use the interpolation framework.

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