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] [Comment Edited] (HBASE-13448) New Cell implementation with cached component offsets/lengths
Date Tue, 19 May 2015 06:24:00 GMT

    [ https://issues.apache.org/jira/browse/HBASE-13448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14549869#comment-14549869
] 

ramkrishna.s.vasudevan edited comment on HBASE-13448 at 5/19/15 6:23 AM:
-------------------------------------------------------------------------

bq.Not all Cells will have a notion of a key length, I thought we're going to get rid of it
(rather than now cementing its use more by caching the key length).
I would say 'yes', we are getting rid of it even now.  But KeyValue, a type of Cell works
based on keylength and where ever KeyValue is used in read side its better to cache it (the
perf reading part wrt GC need to check and decide though).  Overall the read path is actually
not bothered about how the key or the key length is getting used.  Previously the read path
was directly using the key or keylength and trying to infer the details about a Cell.
As Stack said a Cell would know what to be cached. If you take any cell coming from the encoders
we don't do this.




was (Author: ram_krish):
bq.Not all Cells will have a notion of a key length, I thought we're going to get rid of it
(rather than now cementing its use more by caching the key length).
I would say 'yes', we are getting rid of it even now.  But KeyValue, a type of Cell works
based on keylength and where ever KeyValue is used in read side its better to cache it (the
perf reading part wrt GC need to say though).  Overall the read path is actually not bothered
about how the key or the key length is getting used.  Previously the read path was directly
using the key or keylength and trying to infer the details about a Cell.
As Stack said a Cell would know what to be cached. If you take any cell coming from the encoders
we don't do this.



> New Cell implementation with cached component offsets/lengths
> -------------------------------------------------------------
>
>                 Key: HBASE-13448
>                 URL: https://issues.apache.org/jira/browse/HBASE-13448
>             Project: HBase
>          Issue Type: Sub-task
>          Components: Scanners
>            Reporter: Anoop Sam John
>            Assignee: Anoop Sam John
>             Fix For: 2.0.0
>
>         Attachments: HBASE-13448.patch, HBASE-13448_V2.patch, HBASE-13448_V3.patch, gc.png,
hits.png
>
>
> This can be extension to KeyValue and can be instantiated and used in read path.



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

Mime
View raw message