cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jack Money (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (CASSANDRA-11511) disable Anticompaction after repair
Date Wed, 06 Apr 2016 11:39:25 GMT

     [ https://issues.apache.org/jira/browse/CASSANDRA-11511?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Jack Money updated CASSANDRA-11511:
-----------------------------------
    Description: 
 I use write consistency level ALL.
Sometimes when my data grows to fast i cant normal add new node, [CASSANDRA-10797 issue occurs|https://issues.apache.org/jira/browse/CASSANDRA-10797].
After this i set auto_bootstrap to false, and run nodetool repair - full on all tables one
by one.
70% percent time of repair is taken by "Anticompaction after repair".
But i don't need it, because i do not need to run repair (consistency level ALL.). 
To avoid anticompaction i run repair with token range [Avoid anticompaction when doing subrange
repair|https://issues.apache.org/jira/browse/CASSANDRA-10422] but i think it should be options
to avoid anticompaction, i.e. nodetool repair -full *-disableanticompaction*  

  was:
 I use write consistency level ALL.
Sometimes when my data grows to fast i cant normal add new node, [CASSANDRA-10797 issue occurs|https://issues.apache.org/jira/browse/CASSANDRA-10797].
After this bug i set auto_bootstrap to false, and run nodetool repair - full on all tables
one by one.
70% percent time of repair is taken by "Anticompaction after repair".
But i don't need it, because i do not need to run repair (consistency level ALL.). 
To avoid anticompaction i run repair with token range [Avoid anticompaction when doing subrange
repair|https://issues.apache.org/jira/browse/CASSANDRA-10422] but i think it should be options
to avoid anticompaction, i.e. nodetool repair -full *-disableanticompaction*  


> disable Anticompaction after repair
> -----------------------------------
>
>                 Key: CASSANDRA-11511
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-11511
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Jack Money
>            Priority: Minor
>
>  I use write consistency level ALL.
> Sometimes when my data grows to fast i cant normal add new node, [CASSANDRA-10797 issue
occurs|https://issues.apache.org/jira/browse/CASSANDRA-10797].
> After this i set auto_bootstrap to false, and run nodetool repair - full on all tables
one by one.
> 70% percent time of repair is taken by "Anticompaction after repair".
> But i don't need it, because i do not need to run repair (consistency level ALL.). 
> To avoid anticompaction i run repair with token range [Avoid anticompaction when doing
subrange repair|https://issues.apache.org/jira/browse/CASSANDRA-10422] but i think it should
be options to avoid anticompaction, i.e. nodetool repair -full *-disableanticompaction* 




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

Mime
View raw message