-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/42021/#review113971
-----------------------------------------------------------



ambari-common/src/main/python/resource_management/libraries/providers/hdfs_resource.py
 (line 25)
<https://reviews.apache.org/r/42021/#comment174768>

    You only implemented this behaviour for webhdfs resource, but will if 
webhdfs is turned of the hdfsResource will fail.



ambari-common/src/main/python/resource_management/libraries/providers/hdfs_resource.py
 (line 297)
<https://reviews.apache.org/r/42021/#comment174766>

    This time now efficient image having a lot of files there.
    
    curl can put multiple files at a connection. Can we put all at one 
connection by specifying multiple -T
    
    cstm-hdfs@test-ss-u14-dev-kerb-mit-3-2:~$ curl -sS -L -w '%{http_code}' -X 
PUT -T /usr/hdp/2.4.0.0-57/hadoop/mapreduce.tar.gz -T /tmp/1 --negotiate -u : 
'http://test-ss-u14-dev-kerb-mit-3-5.novalocal:50070/webhdfs/v1/hdp/apps/2.4.0.0-57/mapreduce/mapreduce?op=CREATE&user.name=cstm-hdfs&overwrite=True&permission=444'



ambari-common/src/main/python/resource_management/libraries/providers/hdfs_resource.py
 (line 298)
<https://reviews.apache.org/r/42021/#comment174767>

    This is conter intuitive to ignore directories, I don't the person which 
uses this resource later would be happy with this behaviour.


- Andrew Onischuk


On Jan. 12, 2016, 8:26 a.m., yao lei wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/42021/
> -----------------------------------------------------------
> 
> (Updated Jan. 12, 2016, 8:26 a.m.)
> 
> 
> Review request for Ambari and Andrew Onischuk.
> 
> 
> Bugs: AMBARI-14536
>     https://issues.apache.org/jira/browse/AMBARI-14536
> 
> 
> Repository: ambari
> 
> 
> Description
> -------
> 
> In some cases,we need to copy local matching wildcard files to hdfs.
> eg. we'd like to copy /usr/lib/spark*.jar to hdfs by following codes in 
> Python:
> params.HdfsResource(hdfs_path,
> type="file",
> action="create_on_execute",
> source="/usr/lib/spark*.jar",
> group=params.user_group,
> owner=params.pdm_user,
> mode=0755
> )
> params.HdfsResource(None, action="execute")
> 
> 
> Diffs
> -----
> 
>   
> ambari-common/src/main/python/resource_management/libraries/providers/hdfs_resource.py
>  731bce7 
> 
> Diff: https://reviews.apache.org/r/42021/diff/
> 
> 
> Testing
> -------
> 
> Manually Tested.
> 
> 
> Thanks,
> 
> yao lei
> 
>

Reply via email to