Hello, I have a TOP network that is intaking JSON data and using the data to generate some geometry. When I run the TOPnet on a small sample of JSON data, everything works as desired. When I run the TOPnet on a larger JSON, everything also works as desired — except that the TOPs UI seems to not be able to keep up with the actual output of the network. For example, I can cook the network and wait 2 minutes to see all my geometry properly created and output to the right places on my computer; meanwhile the TOP network appears to still be churning through the network, with the nodes that generated the geometry allegedly "waiting" (with the 3 green dots) for upstream items—even though I know those "waiting" nodes must have finished, because they have already cooked out the appropriate geometry.

This isn't a huge blocker for me since I'm still producing the geometry I need, but it is strange because I have to force quit Houdini to get back to the file—attempting to stop the TOPnet freezes Houdini. I've tried distributing "Wait for all" nodes and allocating more CPU in the local scheduler. I'm running a moderately beefy system with 64gb of RAM and a 12core CPU, and I'm not seeing myself run out of RAM.

This feels like a bug to me—but wondering if anyone can shed some light. Thanks!