[ 
https://issues.apache.org/jira/browse/HADOOP-3479?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12607671#action_12607671
 ] 

Steve Loughran commented on HADOOP-3479:
----------------------------------------

+1 to hierarchical; it can be generally useful. Going with long.property.names. 
is simple, but doesnt scale and leads to files like log4j.properties. 

=0 to strong typing. That may seem odd as SmartFrog does have strong types 
-integers, floats, strings- but note that Hadoop Configuration currently does 
java property expansion in its work, so it expects to get raw strings and 
post-process them. FWIW, we work in our types and then toString() everything in 
the hadoop binding, so you can do integer maths (1024 * 1024 * 1024) and 
boolean logic, and hadoop doesn't need to care. 

If the XML format adopts XML Schema types then you couldnt have a reference 
like ${http.Proxy.Port} in your (integer) port property, as the string would 
still be unexpanded. during XSD validation. 

Rob, what you need is "preflight validation", which takes the conf files + 
other settings and validates them, either locally or in the context of the 
target system

I am offering the engineering effort needed to get some of this working with a 
couple of back ends -SmartFrog and JNDI- as both are hierarchical and so test 
things better. That's about 30% of my time, depending on other crises.

> Implement configuration items useful for Hadoop resource manager (v1)
> ---------------------------------------------------------------------
>
>                 Key: HADOOP-3479
>                 URL: https://issues.apache.org/jira/browse/HADOOP-3479
>             Project: Hadoop Core
>          Issue Type: New Feature
>            Reporter: Hemanth Yamijala
>            Assignee: Hemanth Yamijala
>         Attachments: 3479.1.patch, 3479.patch
>
>
> HADOOP-3421 lists requirements for a new resource manager for Hadoop. 
> Implementation for these will require support for new configuration items in 
> Hadoop. This JIRA is to define such configuration, and track it's 
> implementation.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to