hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ramkrishna.s.vasudevan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-7319) Extend Cell usage through read path
Date Tue, 08 Apr 2014 05:20:17 GMT

    [ https://issues.apache.org/jira/browse/HBASE-7319?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13962594#comment-13962594
] 

ramkrishna.s.vasudevan commented on HBASE-7319:
-----------------------------------------------

ScanQueryMatcher is changable.  For now it will still work with KeyValues.  Will take it up
in a follow up issue. It is a core area and so wanted to do it separately after making sure
that these changes does not affect the other read areas.
For the DBE part see HBASE-10801 and the patch attached there.  This change combined with
HBASE-10801 would speed up reads using DBE because we could avoid the deep copy while doing
getKeyValue.  
See 
https://issues.apache.org/jira/browse/HBASE-10801?focusedCommentId=13941459&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13941459.

> Extend Cell usage through read path
> -----------------------------------
>
>                 Key: HBASE-7319
>                 URL: https://issues.apache.org/jira/browse/HBASE-7319
>             Project: HBase
>          Issue Type: Umbrella
>          Components: Compaction, Performance, regionserver, Scanners
>            Reporter: Matt Corgan
>         Attachments: HBASE-7319.patch
>
>
> Umbrella issue for eliminating Cell copying.
> The Cell interface allows us to work with a reference to underlying bytes in the block
cache without copying each Cell into consecutive bytes in an array (KeyValue).



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message