[
https://issues.apache.org/jira/browse/HBASE-18232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16067925#comment-16067925
]
Anastasia Braginsky commented on HBASE-18232:
---------------------------------------------
bq. So in coming user request will copy to an ondemand ByteBuffer and when the
actual flattening happens (say to CellChunkMap) then there we create a new
variable size Chunk and copy the content from the ondemand ByteBuffer created
during writes to this new variable chunk(again an ondemand Bytebuffer but with
associated with a chunkId)?
I think you are right. The user request will be served as before, meaning the
cell space will be allocated from JVM. Pay attention that this is a not a MSLAB
chunk (apropos the cell is bigger than a chunk), and there is no relevant chunk
ID. When flattening happens the cell is indeed copied into specially allocated
chunk with chunk ID.
bq. So the similar thing should be done with increment/append cells also?
Exactly.
> Add variable size chunks to the MSLAB
> -------------------------------------
>
> Key: HBASE-18232
> URL: https://issues.apache.org/jira/browse/HBASE-18232
> Project: HBase
> Issue Type: Sub-task
> Reporter: Anastasia Braginsky
>
> Add possibility to create a variable size chunks of memory, so any cell (of
> any size) can reside on a chunk.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
