[
https://issues.apache.org/jira/browse/HADOOP-1033?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12475559
]
James P. White commented on HADOOP-1033:
----------------------------------------
The value of a public AMI is nearly zero if the step-by-step (which should be
as automatic as it can be and the requisite scripts included) to create it are
not documented.
My point about the page being an example of Java-on-EC2 is to point out the
value of unintended consequences in promoting the project. Also I'm defending
the interests of users here, which are understandably given little
consideration by open source developers. The key accomodation that developers
can do is not rip out and throw away things that users are using. For a great
example of a OSS project that gets that right, see the Spring Framework.
> Rewrite AmazonEC2 wiki page
> ---------------------------
>
> Key: HADOOP-1033
> URL: https://issues.apache.org/jira/browse/HADOOP-1033
> Project: Hadoop
> Issue Type: Improvement
> Components: contrib/streaming, scripts
> Reporter: Tom White
> Assigned To: Tom White
> Priority: Minor
>
> The wiki page that describes running Hadoop on Amazon EC2
> (http://wiki.apache.org/lucene-hadoop/AmazonEC2) includes a large number of
> instructions for manually building images and launching instances using EC2
> tools. We now have a set of automated scripts for doing these tasks (see
> HADOOP-884, HADOOP-952), so I propose removing the manual instructions (also,
> they are likely to become inaccurate over time, and may make the idea of
> using Hadoop on EC2 daunting to new users).
> As a part of this I will keep the parts of the documentation that explain the
> concepts, and document what the scripts do in more detail.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
