I’ve moved this post into the CPCluster forum since it seems more relevant there.
Several others have asked the same question, specifically for Linux. There is no CPCluster 2.0 in the same way as there was one for 1.0. Basically, you run CP 2.0 in “headless” mode (i.e, without the GUI) on the cluster machines, and feed in the relevant data via command line arguments. You can see a basic workflow under Help > General Help > Batch Processing. However, since the implementation tends to be specific to the group/institution using it, we hesitate on posting a one-size-fits-all solution.
Still, I’m sure people would benefit from a few helpful hints. If/when we pull one together, we’ll post it here.