crunch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jeffrey Quinn (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CRUNCH-580) FileTargetImpl#handleOutputs Inefficiency on S3NativeFileSystem
Date Tue, 08 Dec 2015 18:04:10 GMT

    [ https://issues.apache.org/jira/browse/CRUNCH-580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15047164#comment-15047164
] 

Jeffrey Quinn commented on CRUNCH-580:
--------------------------------------

Sorry about that [~jwills] did not notice those stack traces, strangely the integration tests
were successful in spite of them..

Attached a new patch where I moved the creation of the _SUCCESS file out of the call back
and into the body of #handleOutputs, integration tests are passing and these traces are no
longer present.

> FileTargetImpl#handleOutputs Inefficiency on S3NativeFileSystem
> ---------------------------------------------------------------
>
>                 Key: CRUNCH-580
>                 URL: https://issues.apache.org/jira/browse/CRUNCH-580
>             Project: Crunch
>          Issue Type: Bug
>          Components: Core, IO
>    Affects Versions: 0.13.0
>         Environment: Amazon Elastic Map Reduce
>            Reporter: Jeffrey Quinn
>            Assignee: Josh Wills
>         Attachments: CRUNCH-580.patch
>
>
> We have run in to a pretty frustrating inefficiency inside of org.apache.crunch.io.impl.FileTargetImpl#handleOutputs.
> This method loops over all of the partial output files and moves them to their ultimate
destination directories, calling org.apache.hadoop.fs.FileSystem#rename(org.apache.hadoop.fs.Path,
org.apache.hadoop.fs.Path) on each partial output in a loop.
> This is no problem when the org.apache.hadoop.fs.FileSystem in question is HDFS where
#rename is a cheap operation, but when an implementation such as S3NativeFileSystem is used
it is extremely inefficient, as each iteration through the loop makes a single blocking S3
API call, and this loop can be extremely long when there are many thousands of partial output
files.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message