hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Doug Cutting (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-54) SequenceFile should compress blocks, not individual entries
Date Thu, 20 Jul 2006 10:41:16 GMT
    [ http://issues.apache.org/jira/browse/HADOOP-54?page=comments#action_12422393 ] 
            
Doug Cutting commented on HADOOP-54:
------------------------------------

Just to be clear, delta compression means something different to me than zero-compression.
 The former represents a lists of integers with their differences, the latter elides leading
zeros in integers.  They're not exclusive.  A sorted list of integers is smaller when delta
compressed and zero-compressed.  A random list of integers is probably not helped by delta
compression, but is helped by zero compression.  If values are in a narrow range, then delta
compression may help.  Thus it may be useful for lists of key lengths and value lengths.

You provide some benchmarks showing the advantage of zero compression for random sequences.
 Eric said delta compression, but I think he meant zero compression.  I agree that we should
use zero compression everywhere.  The only question is if we should also use delta compression
anywhere.


> SequenceFile should compress blocks, not individual entries
> -----------------------------------------------------------
>
>                 Key: HADOOP-54
>                 URL: http://issues.apache.org/jira/browse/HADOOP-54
>             Project: Hadoop
>          Issue Type: Improvement
>          Components: io
>    Affects Versions: 0.2.0
>            Reporter: Doug Cutting
>         Assigned To: Arun C Murthy
>             Fix For: 0.5.0
>
>         Attachments: VIntCompressionResults.txt
>
>
> SequenceFile will optionally compress individual values.  But both compression and performance
would be much better if sequences of keys and values are compressed together.  Sync marks
should only be placed between blocks.  This will require some changes to MapFile too, so that
all file positions stored there are the positions of blocks, not entries within blocks.  Probably
this can be accomplished by adding a getBlockStartPosition() method to SequenceFile.Writer.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message