[
https://issues.apache.org/jira/browse/HBASE-19133?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16285185#comment-16285185
]
Gali Sheffi commented on HBASE-19133:
-------------------------------------
Hi Everyone!
I have answered all the comments in the Review Board, please take a look and
continue reviewing. So far we have two general questions raised:
1. Whether we should check the appearance of a cell on an MSLAB's chunk via
ByteBufferChunkCell or ByteBufferKeyValue? Currently ByteBufferKeyValue is used
due to possibility of cell being instance of NoTagByteBufferChunkCell.
[~ram_krish] what do you think?
2. The data size of region. So far the data size neither of MemStore nor of
Region wasn't updated upon In-Memory-Compaction or (now) upon Flattening to
CCM. The question is whether we want it to be done? And how did we manage till
now? [~eshcar], [~anoop.hbase], [~ram_krish], what do you think?
Thanks!
> Transfer big cells or upserted/appended cells into MSLAB upon flattening to
> CellChunkMap
> ----------------------------------------------------------------------------------------
>
> Key: HBASE-19133
> URL: https://issues.apache.org/jira/browse/HBASE-19133
> Project: HBase
> Issue Type: Sub-task
> Reporter: Anastasia Braginsky
> Attachments: HBASE-19133.01.patch
>
>
> CellChunkMap Segment index requires all cell data to be written in the MSLAB
> Chunks. Eventhough MSLAB is enabled, cells bigger than chunk size or
> upserted/incremented/appended cells are still allocated on the JVM stack. If
> such cells are found in the process of flattening into CellChunkMap
> (in-memory-flush) they need to be copied into MSLAB.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
