[
https://issues.apache.org/jira/browse/HBASE-14460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14979933#comment-14979933
]
Jingcheng Du commented on HBASE-14460:
--------------------------------------
bq. Or maybe these counter types(append, increase, etc.) of cells and the
latest put that is older than counter cells with the same row key can be kept
always, maybe the TTL can be honored for this latest put?
bq. I mean if all the counter cells are only summed up in major compaction, the
special column family is not needed.
Sorry for the misleading words. I mean these counter cells and latest put that
is older than counter cells with the same row key can be kept in minor
compaction. In this case, at least one put (even it is expired) for each key
will be kept even if there is not counter cells for that key in this minor
compaction.
bq. Does it have to be a special counter column family so it easy to treat
counters special?
And yes, a special column family is a better idea to avoid breaking the
compaction rules in normal column families.
> [Perf Regression] Merge of MVCC and SequenceId (HBASE-HBASE-8763) slowed
> Increments, CheckAndPuts, batch operations
> -------------------------------------------------------------------------------------------------------------------
>
> Key: HBASE-14460
> URL: https://issues.apache.org/jira/browse/HBASE-14460
> Project: HBase
> Issue Type: Bug
> Components: Performance
> Reporter: stack
> Assignee: stack
> Priority: Critical
> Attachments: 14460.txt, region_lock.png
>
>
> As reported by 鈴木俊裕 up on the mailing list -- see "Performance degradation
> between CDH5.3.1(HBase0.98.6) and CDH5.4.5(HBase1.0.0)" -- our unification of
> sequenceid and MVCC slows Increments (and other ops) as the mvcc needs to
> 'catch up' to our current point before we can read the last Increment value
> that we need to update.
> We can say that our Increment is just done wrong, we should just be writing
> Increments and summing on read, but checkAndPut as well as batching
> operations have the same issue. Fix.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
