crunch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gabriel Reid (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CRUNCH-329) Re-add type info to TupleWritable to make fields sort correctly
Date Mon, 03 Feb 2014 17:30:11 GMT

    [ https://issues.apache.org/jira/browse/CRUNCH-329?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13889660#comment-13889660
] 

Gabriel Reid commented on CRUNCH-329:
-------------------------------------

After a look at the code I see what you mean. I was thinking (incorrectly) that the inner
types would only be needed by the MapFns used by the PType, and seeing as those get initialized
then they would have the information that they need.

I guess this means that Writables.WRITABLE_CODES probably needs to get updated somehow based
on the Configuration (so that its state is published to remote JVMs). I guess this could be
as simple as a static Writables.initialize(Configuration) function that would merge in the
codes contained in the Configuration, which feels really ugly to me but I don't see a lot
of other easily-implementable options. What do you think?

> Re-add type info to TupleWritable to make fields sort correctly
> ---------------------------------------------------------------
>
>                 Key: CRUNCH-329
>                 URL: https://issues.apache.org/jira/browse/CRUNCH-329
>             Project: Crunch
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.10.0, 0.8.3
>            Reporter: Josh Wills
>            Assignee: Josh Wills
>             Fix For: 0.10.0, 0.8.3
>
>         Attachments: CRUNCH-329.patch, fix-ss-writables.patch
>
>
> Secondary sorts aren't currently working correctly for Writable types after we hacked
the TupleWritable impl to make all of the fields BytesWritables (e.g., secondary IntWritable
values will no longer be sorted correctly, even though everything is still grouped correctly.)
> The least-bad way that I came up with to fix this is to use integer codes for each possible
WritableComparable type in a pipeline that we can use to decode what Writable type each tuple
field corresponds to. This allows us to keep the various fields sortable while still doing
a reasonable job of minimizing the serialization required to pass the type information along.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message