[
https://issues.apache.org/jira/browse/HBASE-14921?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15392571#comment-15392571
]
Anastasia Braginsky commented on HBASE-14921:
---------------------------------------------
First, I agree that the sizing issue is ugly and need to be improved. Which is
partially done in this patch and I planned to improve it further.
However, I think it is unnecessary and not urgent, to open another JIRA for
this fix. This is not an issue of rebase only, because we take the code in two
different directions. We could live with the code as is (or at least could see
the final outcome of 14921) and later we could agree how to arrange the sizes
(if what we have is not good enough)...
Your two concerns are very clear.
1. The flattening without compaction is causing many small segments in
pipeline, and they are not flushed all together.
2. The issue of compaction prediction cost.
Please correct me if I am wrong.
We understand those concerns. There is no argument that your first concern will
be fixed. For your second concern we are going to benchmark it ourselves with
PE.
> Memory optimizations
> --------------------
>
> Key: HBASE-14921
> URL: https://issues.apache.org/jira/browse/HBASE-14921
> Project: HBase
> Issue Type: Sub-task
> Affects Versions: 2.0.0
> Reporter: Eshcar Hillel
> Assignee: Anastasia Braginsky
> Attachments: CellBlocksSegmentInMemStore.pdf,
> CellBlocksSegmentinthecontextofMemStore(1).pdf, HBASE-14921-V01.patch,
> HBASE-14921-V02.patch, HBASE-14921-V03.patch, HBASE-14921-V04-CA-V02.patch,
> HBASE-14921-V04-CA.patch, HBASE-14921-V05-CAO.patch,
> HBASE-14921-V06-CAO.patch, InitialCellArrayMapEvaluation.pdf,
> IntroductiontoNewFlatandCompactMemStore.pdf
>
>
> Memory optimizations including compressed format representation and offheap
> allocations
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
