Building DataFlow Applications : Building DataFlow Applications Using RushScript : RushScript Variables : Cluster Specifier
 
Share this page                  
Cluster Specifier
The dr.cluster() method is used to create a cluster specifier that identifies the cluster execution environment for the current graph. The method returns a Cluster object that can be used to further specify execution properties. The Cluster object supports the following methods:
Function Name
Input Parameters
Returns
Description
host
 
String:cluster manager host name
Gets the current setting of the host name of Cluster Manager.
host
String:host name
 
Sets the host name of Cluster Manager.
option
String:option name, String:option value
 
Sets the value of a run-time setting of the given name.
port
 
int: cluster manager port number
Gets the current setting of the port number of Cluster Manager.
port
int: port number
 
Sets the port number of Cluster Manager.
The Cluster.option() method can be used to set run time options used during graph execution. The following example demonstrates setting the cluster specifier and then setting additional run time options. See Job Settings for more information about the per-job settings that may be overridden using the Cluster.option() function.
Setting Additional Cluster Options
var cluster = dr.cluster("server", 9999);
cluster.option("job.executor.jvm.memory", "2G");
cluster.option("job.executor.jvm.arguments", "-Dmy.system.property=value");
cluster.option("job.executor.logging.level", "DEBUG");