[
https://issues.apache.org/jira/browse/HBASE-16421?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15782176#comment-15782176
]
ramkrishna.s.vasudevan commented on HBASE-16421:
------------------------------------------------
bq. I assume that in all experiments (except DefaultMemStore) you had MSLAB and
Chunk Pool on,
Yes.
bq.What is the "total time" that you are presenting? Total time of the
experiment? Isn’t it 16 minutes?
Yes it is 16 mins
bq.You can get specifically the scan latency from YCSB.
This does not complete as I said. Because every scan is a full scan from the
randomly selected start row. I may have to reduce the record count and the
operation count may be - to complete this experiment.
Also since we don't know if the scan comes from memstore or files - how to
ensure that the garbage gnerated on a memstore scan reduces/increases
performance ? I don't think that is possible unless we have the memstore first
Scan approach.
bq.Anyway, we can say that CellChunkMap is not decreasing the performance too
much, and it worth it at least as off-heaping worth it. Do you agree?
Yes I think so. But let us wait for some more feedback or cases that I may be
missing. More tests can be done from beginning of next week.
> Introducing the CellChunkMap as a new additional index variant in the MemStore
> ------------------------------------------------------------------------------
>
> Key: HBASE-16421
> URL: https://issues.apache.org/jira/browse/HBASE-16421
> Project: HBase
> Issue Type: Umbrella
> Reporter: Anastasia Braginsky
> Attachments: CellChunkMapRevived.pdf, ChunkCell_creation.png,
> IntroductiontoNewFlatandCompactMemStore.pdf
>
>
> Follow up for HBASE-14921. This is going to be the umbrella JIRA to include
> all the parts of integration of the CellChunkMap to the MemStore.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
