hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anoop Sam John (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-17662) Disable in-memory flush when replaying from WAL
Date Fri, 24 Feb 2017 01:09:44 GMT

    [ https://issues.apache.org/jira/browse/HBASE-17662?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15881686#comment-15881686
] 

Anoop Sam John commented on HBASE-17662:
----------------------------------------

I was just asking whether the WAL replay handling (And so this boolean write and read) being
done by single thread only or not.  Did not read the code.   My worry on the 1st patch was
that, it was doing a volatile read on every Cell addition..  Now we have changed it so that
once the in memory flush size limit reached, we will do the boolean read to confirm it is
not a replay time.  So we are ok IMO.    Understood u made it AtomicBoolean in a preventive
way.  Am ok with that way also provided we moved the boolean check as 2nd check now.

> Disable in-memory flush when replaying from WAL
> -----------------------------------------------
>
>                 Key: HBASE-17662
>                 URL: https://issues.apache.org/jira/browse/HBASE-17662
>             Project: HBase
>          Issue Type: Sub-task
>            Reporter: Anastasia Braginsky
>            Assignee: Anastasia Braginsky
>         Attachments: HBASE-17662-V02.patch, HBASE-17662-V03.patch, HBASE-17662-V04.patch,
HBASE-17662-V05.patch
>
>
> When replaying the edits from WAL, the region's updateLock is not taken, because a single
threaded action is assumed. However, the thread-safeness of the in-memory flush of CompactingMemStore
is based on taking the region's updateLock. 
> The in-memory flush can be skipped in the replay time (anyway everything is flushed to
disk just after the replay). Therefore it is acceptable to just skip the in-memory flush action
while the updates come as part of replay from WAL.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message