Job getting failed for 1 node

srini.dw · Post by **srini.dw** » Mon Dec 30, 2013 6:01 am

Hi,

Please need your help to resolve the below issue.
Job Design
ODBC Connector(SQL server) -> Copy Stage -> DataSet.

The job is running fine in the 2 node environment.
But when I run the job in 1 node, its getting failed with the below error.
Parallel job reports failure (code 139)
Below are the 2 node files as configured

Code: Select all

Twonode.apt
{
        node "node1"
        {
                fastname "RAP"
                pools ""
                resource disk "/apps/IBM/dataset1" {pools ""}
                resource scratchdisk "/apps/IBM/scratch" {pools ""}
        }
        node "node2"
        {
                fastname "RAP"
                pools ""
                resource disk "/apps/IBM/dataset2" {pools ""}
                resource scratchdisk "/apps/IBM/scratch" {pools ""}
        }
}


Onenode.apt
{
        node "Conductor"
        {
                fastname "PAR"
                pools "conductor"
                resource disk "/apps/IBM/dataset1" {pools ""}
                resource scratchdisk "/apps/IBM/scratch" {pools ""}
        }
}

Please let me know any issue with the configuration file for 1 node.

Thanks,

chulett · Post by **chulett** » Mon Dec 30, 2013 8:38 am

Did you try doing an exact search for "code 139"? There are 46 other posts on this subject, did none of them help?

ray.wurlod · Post by **ray.wurlod** » Tue Dec 31, 2013 6:14 am

Is "PAR" really a hostname, or have you mis-spelled "RAP" in the one node configuration?

chulett · Post by **chulett** » Tue Dec 31, 2013 8:05 am

Or that.

srini.dw · Post by **srini.dw** » Wed Jan 01, 2014 11:14 pm

Thanks for the reply.

My mistake, 1node.apt is RAP

Code: Select all

Twonode.apt 
{ 
        node "node1" 
        { 
                fastname "RAP" 
                pools "" 
                resource disk "/apps/IBM/dataset1" {pools ""} 
                resource scratchdisk "/apps/IBM/scratch" {pools ""} 
        } 
        node "node2" 
        { 
                fastname "RAP" 
                pools "" 
                resource disk "/apps/IBM/dataset2" {pools ""} 
                resource scratchdisk "/apps/IBM/scratch" {pools ""} 
        } 
} 


Onenode.apt 
{ 
        node "Conductor" 
        { 
                fastname "RAP" 
                pools "conductor" 
                resource disk "/apps/IBM/dataset1" {pools ""} 
                resource scratchdisk "/apps/IBM/scratch" {pools ""} 
        } 
}

We are using a grid configuration.

I have checked the other posts, its not much of help.

Any ideas will be appreciated.

Thanks,

lstsaur · Post by **lstsaur** » Thu Jan 02, 2014 12:45 am

In a grid environment, the APT configuration file is generated dynamically by the system. All you have to do is to specify "1" in the APT_GRID_COMPUTENODES parameter, then the job will run with just 1 node.

ray.wurlod · Post by **ray.wurlod** » Thu Jan 02, 2014 3:37 am

Your single node configuration lacks a default node pool name ("").

srini.dw · Post by **srini.dw** » Fri Jan 03, 2014 12:09 pm

Thanks for the replies, will try out and let you know.