small hadoop cluster setup question

[Ski Gh3 <sk...@gmail.com>]

Hi all,

I'm trying to set up a small cluster with 3 machines.  I'd like to have one
machine serves as the namenode and the jobtracker, while the 3 all serve as
the datanode and tasktrackers.

After following the set up instructions, I got an exception running
 $HADOOP_HOME/bin/start-dfs.sh: java.net.BindException:Address already in
use
and my secondarynamenode cannot be started.  But as I can see, the namenode
started and the three data nodes all started. Does this matter?

However, when I go to the webinterface at localhost:50070, I saw there is 1
datanode.
Is there any reason why the data nodes are not started on the other two
machines?
The site.xml file and the java path, installation path etc are all set up.

Does anyone have this problem before? I 'd really appreciate any help!

Thanks!

Re: small hadoop cluster setup question

[Samuel Guo <gu...@gmail.com>]

could you please attach your configurations and logs?

On Fri, Sep 26, 2008 at 6:12 AM, Ski Gh3 <sk...@gmail.com> wrote:

> Hi all,
>
> I'm trying to set up a small cluster with 3 machines.  I'd like to have one
> machine serves as the namenode and the jobtracker, while the 3 all serve as
> the datanode and tasktrackers.
>
> After following the set up instructions, I got an exception running
>  $HADOOP_HOME/bin/start-dfs.sh: java.net.BindException:Address already in
> use
> and my secondarynamenode cannot be started.  But as I can see, the namenode
> started and the three data nodes all started. Does this matter?
>
> However, when I go to the webinterface at localhost:50070, I saw there is 1
> datanode.
> Is there any reason why the data nodes are not started on the other two
> machines?
> The site.xml file and the java path, installation path etc are all set up.
>
> Does anyone have this problem before? I 'd really appreciate any help!
>
> Thanks!
>