[
https://issues.apache.org/jira/browse/HDFS-13237?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16432008#comment-16432008
]
maobaolong commented on HDFS-13237:
-----------------------------------
[~elgoiri] Thank you for this documentation update.
Would you like to introduce where we use those
"HASH|LOCAL|RANDOM|HASH_ALL|SPACE" modes, and in each mode, what we should
ensure?
Such as, RANDOM mode, i think we should know a series mkdir operation under a
parent dir could transfer to any of the targets, so if one of the target ns
down, the listing or du operation will get the wrong result. I think this is a
terrible thing.
> [Documentation] RBF: Mount points across multiple subclusters
> -------------------------------------------------------------
>
> Key: HDFS-13237
> URL: https://issues.apache.org/jira/browse/HDFS-13237
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Íñigo Goiri
> Assignee: Íñigo Goiri
> Priority: Minor
> Attachments: HDFS-13237.000.patch, HDFS-13237.001.patch,
> HDFS-13237.002.patch, HDFS-13237.003.patch
>
>
> Document the feature to spread mount points across multiple subclusters.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
