hadoop-hdfs-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Wang (JIRA)" <j...@apache.org>
Subject [jira] [Resolved] (HDFS-8954) Not balance block in RAM_DISK use lazy_persist policy on datanode local copy
Date Thu, 27 Aug 2015 03:28:46 GMT

     [ https://issues.apache.org/jira/browse/HDFS-8954?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Andrew Wang resolved HDFS-8954.
-------------------------------
    Resolution: Not A Problem

This is working as intended, only the local replica goes to RAM_DISK, the others go to DISK.

> Not balance block in RAM_DISK use lazy_persist policy  on datanode local copy
> -----------------------------------------------------------------------------
>
>                 Key: HDFS-8954
>                 URL: https://issues.apache.org/jira/browse/HDFS-8954
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: datanode
>    Affects Versions: 2.7.1
>            Reporter: Xu Chen
>            Priority: Critical
>
> I was try to use Heterogeneous Storage feature and I config RAM_DISK volume for each
DN size of 20g , and set one directory policy of lazy_persist at HDFS , then command copyFromLocal
to load a 79g file to the directory HDFS ,  it is have 593 block , and only 141 block in RAM_DISK
, I use command "df" to look tmpfs usage it only my execute command DN fill enough the tmpfs
, other DN tmpfs is empty !   
> when I use other machine without datanode installed , it is worked fine all DN tmpfs
is fill enough
> Regards



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

Mime
View raw message