You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by Allan Avendaño <aa...@fiec.espol.edu.ec> on 2008/05/27 03:08:32 UTC

Masters and Slaves in Slaves machines

Regards for all!

I'm configuring a HDFS with multi-nodes, I've some problems with slaves
machines, specifically configuring these archives:

   masters
   slaves
   hadoop-site.xml


I don't have idea which ip address I should have to write in "masters"
archive in slaves machines...

localhost or "my master machine".

And so, with others archives.

Thanks for ur help,

-- 
--------

Allan Roberto Avendaño Sudario
Guayaquil-Ecuador


Re: Masters and Slaves in Slaves machines

Posted by Taeho Kang <tk...@gmail.com>.
In masters file, you should write your master's address.
In slaves file, you should write all of your slaves' addresses, seperated by
'\n' (newline char)



On Tue, May 27, 2008 at 10:08 AM, Allan Avendaño <aa...@fiec.espol.edu.ec>
wrote:

> Regards for all!
>
> I'm configuring a HDFS with multi-nodes, I've some problems with slaves
> machines, specifically configuring these archives:
>
>   masters
>   slaves
>   hadoop-site.xml
>
>
> I don't have idea which ip address I should have to write in "masters"
> archive in slaves machines...
>
> localhost or "my master machine".
>
> And so, with others archives.
>
> Thanks for ur help,
>
> --
> --------
>
> Allan Roberto Avendaño Sudario
> Guayaquil-Ecuador
>
>

Re: Masters and Slaves in Slaves machines

Posted by phonechen <ph...@gmail.com>.
1. config the namenode and jobtracker node in  hadoop-site.xml :

<property>
  <name>fs.default.name</name>
  <value>your-name-node:9527</value>
</property>

<property>
  <name>mapred.job.tracker</name>
  <value>your-job-tracker-node :9528</value>
</property>
2. config all the datanode and tasktracker node by the "slaves"   file,
node1
 node2
 node3
....


3."masters" is for secondary namenode,so if you just want to run the hadoop
,my suggestion is remove the content and
 leave it as an empty  file.





On 5/27/08, Allan Avendaño <aa...@fiec.espol.edu.ec> wrote:
>
> Regards for all!
>
> I'm configuring a HDFS with multi-nodes, I've some problems with slaves
> machines, specifically configuring these archives:
>
>   masters
>   slaves
>   hadoop-site.xml
>
>
> I don't have idea which ip address I should have to write in "masters"
> archive in slaves machines...
>
> localhost or "my master machine".
>
> And so, with others archives.
>
> Thanks for ur help,
>
> --
> --------
>
> Allan Roberto Avendaño Sudario
> Guayaquil-Ecuador
>
>


-- 
--~--~---------~--~----~------------~-------~--

Best Regards,

Yours
Phonechen

-~----------~----~----~----~------~----~------

Re: Masters and Slaves in Slaves machines

Posted by Samuel Guo <gu...@gmail.com>.
Allan Avendaño 写道:
> Regards for all!
>
> I'm configuring a HDFS with multi-nodes, I've some problems with slaves
> machines, specifically configuring these archives:
>
>    masters
>    slaves
>    hadoop-site.xml
>
>
> I don't have idea which ip address I should have to write in "masters"
> archive in slaves machines...
>
> localhost or "my master machine".
>   
your master machine in your masters file
slaves file is no neccessary in slaves machines,because slave machine 
doesn't talk to each other, it just talk to master
> And so, with others archives.
>
> Thanks for ur help,
>
>