hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "chenrongwei (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-11283) why should we not introduce distributed database to storage hdfs's metadata?
Date Fri, 30 Dec 2016 04:06:58 GMT

     [ https://issues.apache.org/jira/browse/HDFS-11283?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

chenrongwei updated HDFS-11283:
-------------------------------
    Description: 
why should we not  introduce distributed database to storage hdfs's metadata?
In my opinion,it maybe loss some performance,but it has below improvements:
1、enhance NN's extend ability,such as NN can support much more files and blocks. The problem
of massive little files always make me headache.
2、In most MR cluster aren't care the performance loss,but more care the cluster's scale.
3、NN's HA implements maybe more simpler and reasonable.
so I think maybe we should add a new work mode building on distributed database for NN.

  was:
why should we not  introduce distributed database to storage hdfs's metadata?
In my opinion,it maybe loss some performance,but it has below improvements:
1、enhance NN's extend ability,such as NN can support much more files and blocks. The problem
of massive little files always make me headache.
2、In most MR cluster aren't care the performance loss,but more care the cluster's scale.
3、NN's HA implements maybe more simpler and reasonable.



> why should we not  introduce distributed database to storage hdfs's metadata?
> -----------------------------------------------------------------------------
>
>                 Key: HDFS-11283
>                 URL: https://issues.apache.org/jira/browse/HDFS-11283
>             Project: Hadoop HDFS
>          Issue Type: Wish
>            Reporter: chenrongwei
>
> why should we not  introduce distributed database to storage hdfs's metadata?
> In my opinion,it maybe loss some performance,but it has below improvements:
> 1、enhance NN's extend ability,such as NN can support much more files and blocks. The
problem of massive little files always make me headache.
> 2、In most MR cluster aren't care the performance loss,but more care the cluster's scale.
> 3、NN's HA implements maybe more simpler and reasonable.
> so I think maybe we should add a new work mode building on distributed database for NN.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org


Mime
View raw message