hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Liang Xie (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-7845) optimize hfile index key
Date Thu, 07 Mar 2013 02:18:13 GMT

    [ https://issues.apache.org/jira/browse/HBASE-7845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13595445#comment-13595445
] 

Liang Xie commented on HBASE-7845:
----------------------------------

Please refer to TestBlocksRead case in v3, it's easy to see the performance gain.
It means after HBASE-4434,HBASE-4433,HBASE-4496,HBASE-4465,HBASE-4585,HBASE-4532, we can optimize
seeking number again, avoid extra potentially disk IO:)

For another expected benefit:  reduce index size,  it depends on the real key characteristic,
if you have a random long key distribution(e.g. md5 hash key) and short value, then the optimization
result will be very outstanding;  and if most of keys are increased sequentially(e.g. key1000,key1001,key1002),then
no obvious index size optimization will be observed.
                
> optimize hfile index key
> ------------------------
>
>                 Key: HBASE-7845
>                 URL: https://issues.apache.org/jira/browse/HBASE-7845
>             Project: HBase
>          Issue Type: Improvement
>          Components: HFile
>    Affects Versions: 0.96.0
>            Reporter: Liang Xie
>            Assignee: Liang Xie
>             Fix For: 0.98.0
>
>         Attachments: HBASE-7845.txt, HBASE-7845-v2.txt, HBASE-7845-v3.txt
>
>
> Leveldb uses ByteWiseComparatorImpl::FindShortestSeparator() & FindShortSuccessor()
to reduce index key size, it would be helpful under special conditions.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message