Efficient Request Queueing – Optimizing LLM Performance