I don't know the answers to your questions.  I learned what I know by
grepping through the DataNode.java source for "rack".

To use it, I set the "dfs.network.script" property in my hadoop config.
This contains the path to an executable which prints the name of the rack.

There's also a way to pass it on the command line to the DataNode class, so
it depends on how you start up your datanodes, I suppose.

-Michael

On 9/6/07 3:35 PM, "Torsten Curdt" <[EMAIL PROTECTED]> wrote:

> Is there a jira issue for it already?
> 
> Whom could I bug (bribe?) to document the rack-awareness feature? :)
> 
> cheers
> --
> Torsten
> 
> On 07.09.2007, at 00:26, Michael Bieniosek wrote:
> 
>> Well, there is an (undocumented?) way to get rack-awareness in the
>> Datanode,
>> so you could co-opt this to represent datacenter-awareness.  I
>> don't think
>> there is such a rack-awareness ability for the DFSClient or
>> TaskTracker
>> though.
>> 
>> -Michael
>> 
>> On 9/6/07 3:10 PM, "Torsten Curdt" <[EMAIL PROTECTED]> wrote:
>> 
>>>> Another big question:  Has anybody tried using HADOOP / HDFS across
>>>> multiple geographic sites?
>>> 
>>> That's actually a biggy I would be very much interested in, too
>>> 
>>> cheers
>>> --
>>> Torsten
>> 
> 

Reply via email to