[
https://issues.apache.org/jira/browse/SAMZA-181?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13933501#comment-13933501
]
Martin Kleppmann commented on SAMZA-181:
----------------------------------------
Hi Yan, this looks good, just a few nitpicks:
* Title: "Run Samza with HDFS" sounds to me as though HDFS is used as
input/output of the job, which is not what the tutorial is about. How about
"Deploying Samza jobs from HDFS"?
* Typo: "Buidl"
* As of SAMZA-154, we default {{YARN_HOME}} to be {{$HOME/.samza}}. So you
could simplify the instructions by putting the config in
{{~/.samza/conf/yarn-site.xml}}, then you don't need to set the environment
variable.
> Add a tutorial for running Samza by putting the job package in HDFS
> -------------------------------------------------------------------
>
> Key: SAMZA-181
> URL: https://issues.apache.org/jira/browse/SAMZA-181
> Project: Samza
> Issue Type: Improvement
> Components: docs
> Affects Versions: 0.7.0
> Reporter: Yan Fang
> Priority: Minor
> Fix For: 0.7.0
>
> Attachments: SAMZA-181.patch
>
>
> It is good to have a tutorial to decribe how to make Samza run with the job
> package published to HDFS. Because this option is useful for the users who do
> not have a web server or who just want to use HDFS.
> Added a link in index.md. Created a new doc called run-samza-with-hdfs.md.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
