Using two ports at the HAProxy server I have managed to now run two raging scripts in parallel next to each other using the same set of 6 nodes all
having a single Ollama instance running.
At this time they only used the same model simultaniously dedicated to RAGing.
Eventually, this set up will handle complex routing between dedicated Expert AIs.
