hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gopal V (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-4755) Rewrite MapOutputBuffer to use direct buffers & allow parallel sort+collect
Date Mon, 29 Oct 2012 18:48:12 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-4755?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13486251#comment-13486251

Gopal V commented on MAPREDUCE-4755:

This is rather rough at the moment. Since I'm spilling via mmap(), I can't compress the spill
till I am done sorting it.

As far as memory bounding goes, the real advantage is that we have a split between java GC
space and the sort mb space, something which didn't exist before. Unfortunately, NIO doesn't
provide me a portable unmap() operation, which makes a bit of this stuff fairly painful to
unload from memory (perhaps an fadvise/madvise to unload pages).

Considering my experiment was to only sort the kvmeta here and not the kvbuffer, compression
would cause significant trouble because I assume I can seek & read fast within the MappedByteBuffer.

I did look at sorting L2 sized chunks - which raises an interesting question, the comparator
is responsible for blowing off the cache, the actual data (in say, a terasort) is actually
not staying in cache during the loops. I might get a better bang for buck fixing the number
of comparator calls over messing with the size of the sort block - perhaps hybridizing with
something like a smoothsort after checking a few random positions for being sorted.
> Rewrite MapOutputBuffer to use direct buffers & allow parallel sort+collect
> ---------------------------------------------------------------------------
>                 Key: MAPREDUCE-4755
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4755
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>    Affects Versions: 3.0.0
>         Environment: Ubuntu 12.10 x86_64 (Bulldozer 8-core)
>            Reporter: Gopal V
>            Assignee: Gopal V
>              Labels: optimization, sort
> The MapOutputBuffer has been written with a very severe constraint on the amount of memory
it can consume. This results in code that has to page-in & page-out (i.e spill) data as
it passes through the map buffers.
> With the advent of the java.nio package, there is a fast and portable MMap alternative
to handling your own buffers. This exists outside the GC space of Java and yet provides decently
fast memory access to all the data.
> The suggestion is that using mmap() direct buffers can be faster when a spill is involved
and simpler than the current spill logic when given enough address space & uses the buffer
caches to deliver best effort I/O.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

View raw message