You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by Eugene Chekanskiy <ec...@hortonworks.com> on 2014/01/22 17:06:53 UTC
Review Request 17186: Datanode data directory is not created correctly
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17186/
-----------------------------------------------------------
Review request for Ambari and Andrew Onischuk.
Bugs: AMBARI-4383
https://issues.apache.org/jira/browse/AMBARI-4383
Repository: ambari
Description
-------
dfs.datanode.data.dir was handled as directory path, but it can be comma-separated list of directories, in which datanode store its data. Also this patch include fix for potential issue due to low retry count during creation of hdfs directories.
Diffs
-----
ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/hdfs_datanode.py aa7b5e6
ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/utils.py 058fa66
ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/hdfs_datanode.py e0b6c39
ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/utils.py 0c035b4
Diff: https://reviews.apache.org/r/17186/diff/
Testing
-------
Tested on stack 1.3.3 and 2.1.1 with dfs.datanode.data.dir=/hadoop/hdfs/data1,/hadoop/hdfs/data2, all directories created and datanode started normally.
Thanks,
Eugene Chekanskiy
Re: Review Request 17186: Datanode data directory is not created correctly
Posted by Mahadev Konar <ma...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17186/#review32498
-----------------------------------------------------------
Please add unit tests to this.
- Mahadev Konar
On Jan. 22, 2014, 4:06 p.m., Eugene Chekanskiy wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17186/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 4:06 p.m.)
>
>
> Review request for Ambari and Andrew Onischuk.
>
>
> Bugs: AMBARI-4383
> https://issues.apache.org/jira/browse/AMBARI-4383
>
>
> Repository: ambari
>
>
> Description
> -------
>
> dfs.datanode.data.dir was handled as directory path, but it can be comma-separated list of directories, in which datanode store its data. Also this patch include fix for potential issue due to low retry count during creation of hdfs directories.
>
>
> Diffs
> -----
>
> ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/hdfs_datanode.py aa7b5e6
> ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/utils.py 058fa66
> ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/hdfs_datanode.py e0b6c39
> ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/utils.py 0c035b4
>
> Diff: https://reviews.apache.org/r/17186/diff/
>
>
> Testing
> -------
>
> Tested on stack 1.3.3 and 2.1.1 with dfs.datanode.data.dir=/hadoop/hdfs/data1,/hadoop/hdfs/data2, all directories created and datanode started normally.
>
>
> Thanks,
>
> Eugene Chekanskiy
>
>
Re: Review Request 17186: Datanode data directory is not created correctly
Posted by Andrew Onischuk <ao...@hortonworks.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17186/#review32497
-----------------------------------------------------------
Ship it!
Ship It!
- Andrew Onischuk
On Jan. 22, 2014, 4:06 p.m., Eugene Chekanskiy wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17186/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 4:06 p.m.)
>
>
> Review request for Ambari and Andrew Onischuk.
>
>
> Bugs: AMBARI-4383
> https://issues.apache.org/jira/browse/AMBARI-4383
>
>
> Repository: ambari
>
>
> Description
> -------
>
> dfs.datanode.data.dir was handled as directory path, but it can be comma-separated list of directories, in which datanode store its data. Also this patch include fix for potential issue due to low retry count during creation of hdfs directories.
>
>
> Diffs
> -----
>
> ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/hdfs_datanode.py aa7b5e6
> ambari-server/src/main/resources/stacks/HDP/1.3.3/services/HDFS/package/scripts/utils.py 058fa66
> ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/hdfs_datanode.py e0b6c39
> ambari-server/src/main/resources/stacks/HDP/2.1.1/services/HDFS/package/scripts/utils.py 0c035b4
>
> Diff: https://reviews.apache.org/r/17186/diff/
>
>
> Testing
> -------
>
> Tested on stack 1.3.3 and 2.1.1 with dfs.datanode.data.dir=/hadoop/hdfs/data1,/hadoop/hdfs/data2, all directories created and datanode started normally.
>
>
> Thanks,
>
> Eugene Chekanskiy
>
>