[
https://issues.apache.org/jira/browse/HBASE-13448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14549386#comment-14549386
]
stack commented on HBASE-13448:
-------------------------------
[~lhofhansl]
We are going this route to save CPU -- the length parses show in macro
profiling as costly -- and to simplify code: all over our codebase we are doing
carry-overs, passing a parsed length found in one method as input on other
methods as we walk through the elements of a Cell/KeyValue. These latter won't
always make sense as Cell implementations are different and rather than do them
arbitrarily around the code base, rather let the Cell do the length caching.
See
https://issues.apache.org/jira/browse/HBASE-13448?focusedCommentId=14500517&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14500517
> 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)
