I'm currently cooking a set of assets on a build farm using the “Submit Graph as Job” festure in the HQueue scheduler.
There are multiple Jobs needed for creating all the assets.
It works just fine, but I wanted to streamline the process a bit.
Instead of clicking the “submit” button multiple times to trigger all the Jobs I need, I would like to automate that - ideally in a second TOPs network.
Currently, I'm trying to do that via a python node in a separate TOPs network.
I copy&pasted some code I found in /pdg/job/top.py:
import hou node = hou.node("/obj/TOPNET") # find TOPs for current asset node.cook(force=True) # create PDG graph node.executeGraph(False, True, False, True) # execute PDG graph
I run the TOPs network containing tis code via “Tasks->Dirty and cook output node” on my local PC.
It does do the job, but unlike the “Submit graph as Job”, it doesn't seem to run he TOPNET on the build farm (as I hoped it would), but rather on my local PC.
I don't see any job coming up on the HQueue dashboard (the web frontend) when I run that script.
Playing around with the booelan arguments didn't help either.
Any ideas what I might be doing wrong?