[
https://issues.apache.org/jira/browse/HBASE-13448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14557190#comment-14557190
]
stack commented on HBASE-13448:
-------------------------------
[~anoop.hbase] My rig currently occupied. Will try when freed up. Will try on
0.98 so can compare to Lars' runs.
[~lhofhansl] You've pasted your rig somewhere? It is a straight scan end to
end? How to replicate?
bq. I think trunk would give a better picture because we are now not explicitly
passing the rowlength, rowkeys etcs to the comparators.
True but even not doing this, I would think that the caching of a few vital
lengths that are reconstructed repeatedly would make our perf be the same if
not better (Lars is finding otherwise).
> 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: 13448-0.98.txt, 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)
