[
https://issues.apache.org/jira/browse/HADOOP-3475?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12601304#action_12601304
]
chris.douglas edited comment on HADOOP-3475 at 5/30/08 4:48 PM:
----------------------------------------------------------------
Corrects accounting and puts allocations in terms of #bytes required per
record. No unit tests are included, as allocations internal to MapOutputBuffer
aren't easily queried or observed.
was (Author: chris.douglas):
Corrects accounting and puts allocations in terms of #bytes required per
record.
> MapOutputBuffer allocates 4x as much space to record capacity as intended
> -------------------------------------------------------------------------
>
> Key: HADOOP-3475
> URL: https://issues.apache.org/jira/browse/HADOOP-3475
> Project: Hadoop Core
> Issue Type: Bug
> Affects Versions: 0.17.0
> Reporter: Chris Douglas
> Assignee: Chris Douglas
> Fix For: 0.18.0
>
> Attachments: 3475-0.patch
>
>
> When setting up its accounting tables, MapTask.MapOutputBuffer calculates the
> size of each of its buffers as:
> {noformat}
> Where #total_bytes is $io.sort.mb * 2^20 and #accounting_bytes is
> $io.sort.record.percent * #total_bytes
> kvbuffer: #total_bytes - #accounting_bytes
> kvoffsets: 1/4 of #accounting_bytes
> kvindices: 3/4 of #accounting_bytes
> {noformat}
> kvoffsets and kvindices are ints, not bytes; the size of both arrays is
> incorrectly calculated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
