Hi all,

          I have successfully installed pig on the machine which serves as the
namenode/master node. The only problem is when I run pig in hadoop mode it
only makes use of the namenode to perform tasks and not the whole cluster.
I tried including the hadoop/conf directory of all nodes (7) in the
HADOOPDIR variable but tha didn't work.please help. Thanks
      

How are you invoking the pig query ? When you say it makes use of only the namenode, does it mean that it is running in local mode ?Or do you mean the namedone also a datanode and it is using only that datanode?

-Thejas
On 8/4/10 10:32 PM, "ife joeseph" <[EMAIL PROTECTED]> wrote:

 Hi all,

          I have successfully installed pig on the machine which serves as the
namenode/master node. The only problem is when I run pig in hadoop mode it
only makes use of the namenode to perform tasks and not the whole cluster.
I tried including the hadoop/conf directory of all nodes (7) in the
HADOOPDIR variable but tha didn't work.please help. Thanks

Perhaps you've forgotten to set the PIG_CLASSPATH variable?

On Thu, Aug 5, 2010 at 11:02 AM, ife joeseph <[EMAIL PROTECTED]> wrote:
>  Hi all,
>
>          I have successfully installed pig on the machine which serves as the
> namenode/master node. The only problem is when I run pig in hadoop mode it
> only makes use of the namenode to perform tasks and not the whole cluster.
> I tried including the hadoop/conf directory of all nodes (7) in the
> HADOOPDIR variable but tha didn't work.please help. Thanks
>
>
>

--
Harsh J
www.harshj.com