[ https://issues.apache.org/jira/browse/ACCUMULO-3178?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14235811#comment-14235811
]
Christopher Tubbs commented on ACCUMULO-3178:
---------------------------------------------
Another (awesome, I think) use case for this preferred volumes chooser was brought to my attention
today:
Imagine you have two HDFS clusters, A and B, and you want to bulk import files from A into
Accumulo running on B. Right now, bulk import has the restriction that you cannot import from
a nameservice that is not configured as an Accumulo volume. So, you typically have to run
distcp first before you can bulk import. If you list both A and B in {{instance.volumes}},
but have a volume chooser only choose A. Now, you can bulk import files currently on B, but
they'll get stored on A when they are compacted, eliminating the need to distcp.
> Create example preferred volumes chooser
> ----------------------------------------
>
> Key: ACCUMULO-3178
> URL: https://issues.apache.org/jira/browse/ACCUMULO-3178
> Project: Accumulo
> Issue Type: Sub-task
> Reporter: Jenna Huston
> Assignee: Jenna Huston
>
> This example VolumeChooser chooses a volume from a user specified list of preferred volumes
for that specific table.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
|