cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Marcus Eriksson (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (CASSANDRA-10422) Avoid anticompaction when doing subrange repair
Date Tue, 10 Nov 2015 17:09:10 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-10422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14998893#comment-14998893
] 

Marcus Eriksson edited comment on CASSANDRA-10422 at 11/10/15 5:08 PM:
-----------------------------------------------------------------------

[~aweisberg] (sorry for not replying quicker) - isIncremental is for telling the other nodes
which sstables to use when doing the repair (for merkle tree calculation and streaming, if
isIncremental is true, we will only use the sstables that are not marked as repaired) - isGlobal
is used to tell the other nodes if we can actually do the anticompaction - if all nodes that
have the range are included in the repair.

In 2.2+, even if you add -full we are going to do anticompaction


was (Author: krummas):
[~aweisberg] (sorry for not replying quicker) - isIncremental is for telling the other nodes
which sstables to use when doing the repair (for merkle tree calculation and streaming, if
isIncremental is true, we will only use the sstables that are not marked as repaired) - isGlobal
is used to tell the other nodes if we can actually do the anticompaction - if all nodes that
have the range are included in the repair.


> Avoid anticompaction when doing subrange repair
> -----------------------------------------------
>
>                 Key: CASSANDRA-10422
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-10422
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Marcus Eriksson
>            Assignee: Ariel Weisberg
>             Fix For: 3.1, 2.1.x, 2.2.x
>
>
> If we do split the owned range in say 1000 parts, and then do one repair each, we could
potentially anticompact every sstable 1000 times (ie, we anticompact the repaired range out
1000 times). We should avoid anticompacting at all in these cases.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message