This paper presents an indirect methodology to assess IRQ overhead by constructing preliminary approaches to reduce the impact of IRQs. While these approaches are not suitable for general deployment, their corresponding performance observations indirectly confirm the conjecture. Based on these findings, a small modification of a vanilla Linux system is devised that improves the efficiency and performance of traditional kernel-based networking significantly, resulting in up to 45% increased throughput without compromising tail latency