[
https://issues.apache.org/jira/browse/HBASE-16417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15662731#comment-15662731
]
ramkrishna.s.vasudevan commented on HBASE-16417:
------------------------------------------------
ACtually we had a discussion around this. But as Anoop said we cannot go with
first memstore and then HFiles as per the reasons stated above.
For a given row - if there are 100 qualifiers (single CF), you could add them
in 100 different puts one by one.
So in this case when we say get that row - we are not sure if by this time if
any flush had happened moving some of the qualifiers to the file.
The InternalScan is not exposed but if you want a behaviour where you are sure
that you do frequent updates and fetch only the recent one (the catch is that
if the recent is not in memstore - you won't get any result) then we may have
to expose readOnlyMemstore type of API in scan. But am not sure how beneficial
it i will be where there are more chances for missing results.
> In-Memory MemStore Policy for Flattening and Compactions
> --------------------------------------------------------
>
> Key: HBASE-16417
> URL: https://issues.apache.org/jira/browse/HBASE-16417
> Project: HBase
> Issue Type: Sub-task
> Reporter: Anastasia Braginsky
> Assignee: Eshcar Hillel
> Fix For: 2.0.0
>
> Attachments: HBASE-16417-benchmarkresults-20161101.pdf,
> HBASE-16417-benchmarkresults-20161110.pdf
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
