[
https://issues.apache.org/jira/browse/MAPREDUCE-64?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12740837#action_12740837
]
Todd Lipcon commented on MAPREDUCE-64:
--------------------------------------
Hi Arun,
Have you guys worked on this at all already? I'm interested in playing around
with rewriting part of the mapside sort to get rid of this tunable. Like you
said, for a lot of applications the default values are *way* off. 350K records
in 95MB = 271 bytes average record size, which is larger than probably the
majority of jobs we see in practice. If you already have worked on this I don't
want to duplicate your effort, but if not, I think it would be a good step
towards better average performance
without expert tuning.
> Map-side sort is hampered by io.sort.record.percent
> ---------------------------------------------------
>
> Key: MAPREDUCE-64
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-64
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Reporter: Arun C Murthy
>
> Currently io.sort.record.percent is a fairly obscure, per-job configurable,
> expert-level parameter which controls how much accounting space is available
> for records in the map-side sort buffer (io.sort.mb). Typically values for
> io.sort.mb (100) and io.sort.record.percent (0.05) imply that we can store
> ~350,000 records in the buffer before necessitating a sort/combine/spill.
> However for many applications which deal with small records e.g. the
> world-famous wordcount and it's family this implies we can only use 5-10% of
> io.sort.mb i.e. (5-10M) before we spill inspite of having _much_ more memory
> available in the sort-buffer. The word-count for e.g. results in ~12 spills
> (given hdfs block size of 64M). The presence of a combiner exacerbates the
> problem by piling serialization/deserialization of records too...
> Sure, jobs can configure io.sort.record.percent, but it's tedious and
> obscure; we really can do better by getting the framework to automagically
> pick it by using all available memory (upto io.sort.mb) for either the data
> or accounting.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
