[
https://issues.apache.org/jira/browse/PHOENIX-1471?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14235668#comment-14235668
]
Gabriel Reid commented on PHOENIX-1471:
---------------------------------------
[~xufeng] it's possible to supply any custom Hadoop configuration parameters to
the batch load tool by supplying them as properties when starting the tool. For
example, you could currently accomplish the same thing as follows:
{code}hadoop phoenix-xxx-client.jar
org.apache.phoenix.mapreduce.CsvBulkLoadTool -Dmapred.queue.name=my.queue.name
--table EXAMPLE --input /data/example.csv{code}
I'd personally prefer to keep things like this instead of adding a specific
parameter for things that can be supplied as a config parameter like this. Is
that ok for you?
In any case, this definitely shows that we need to add some documentation on
this way of supplying custom hadoop config parameters to the bulk loader at
http://phoenix.apache.org/bulk_dataload.html
> Should we provide queuename option when do mr job like bulkload operation
> -------------------------------------------------------------------------
>
> Key: PHOENIX-1471
> URL: https://issues.apache.org/jira/browse/PHOENIX-1471
> Project: Phoenix
> Issue Type: Improvement
> Affects Versions: 4.0.0
> Reporter: xufeng
> Priority: Minor
> Attachments: patch_v1.txt
>
>
> In acl scenario, people must specify the queue to execute their job.
> When Do bulkload with phoenix,there are no option to specify the queue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
