How does one use RedHawk and distribute components to different nodes? If the two processors are the same (eg x86_64) the implementation section of the spd would have only one implementation section. How do I force a component to one of two or more nodes? Or do I add additional implementation sections? Is there an option in the IDE that I do not see to set all this up or do I have to add allocation properties manually to force this?
There are three ways to have a component spill over onto another executable device (GPP). First, if there are two GPPs available in your domain, components will spill over to the second GPP when resources are fully allocated on the first GPP. Second, you can use the device assignment sequence when creating your application, per section 9.3.3.2 of the REDHAWK User Guide. Lastly, you can add or change one of the allocation properties on one of the devices and have your components look specifically for that allocation property.