[
https://issues.apache.org/jira/browse/HBASE-16417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15451474#comment-15451474
]
Edward Bortnikov commented on HBASE-16417:
------------------------------------------
[~anoop.hbase], all explanations correct. One of the goals of this jira is to
install some order in terminology (smile). Correct, we now distinguish between
two types of compaction: index compaction, which only merges the cell arrays
but does not actually copy the data, and data compaction, which also eliminates
the redundant versions and copies the remaining data. We are working on
benchmarks (ycsb and PE) to verify that the former is inexpensive, and can be
used as default. Part of these benchmarks is fine-tuning the upper bound for
the # of segments in the pipeline.
> 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: Anastasia Braginsky
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
