[
https://issues.apache.org/jira/browse/HBASE-13448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14537685#comment-14537685
]
Anoop Sam John commented on HBASE-13448:
----------------------------------------
Doing more tests around Scan (with out ExplicitColumnTracker), it depends on
the number of rows/cols also
On the 1st cell in 1st row we have 6 calls to getRowLength and on subsequent
cols it is 4
On the 2nd row onwards instead of 6, it will be 9 calls.
Again in all these tests, the cells come from memstore alone (one scanner).
When we have 1 memstore scanner + N StoreFiles, it will be more calls.
Am checking the ECT case, there the calls to getRowLength() is many more
because it is being indirectly used by almost all other getXXXLength/Offset
methods
> 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, 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)
