{"ID":2853263,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.17015","arxiv_id":"2510.17015","title":"Justitia: Fair and Efficient Scheduling of Task-parallel LLM Agents with Selective Pampering","abstract":"LLM agents, which often comprise parallel inference tasks, are commonly adopted to solve real-world problems. When serving such task-parallel LLM agents in shared GPU servers, the scheduler is expected to attain fast agent completion with guaranteed worst-case performance. For that objective, our insight is to selectively pampering agents based on their completion order under idealized fair-sharing. We design Justitia, a fair and also efficient scheduler for task-parallel LLM agents. Noticing that memory is prevalently a bottleneck in LLM serving, Justitia quantifies the true agent cost in a memory-centric manner. It also adopts a light-weight yet accurate method to predict agent costs. Finally, Justitia adopts a virtual-time based fair queuing algorithm to reduce the overall performance with guaranteed worst-case delay. We have implemented Justitia atop vLLM, and experimental results involving diverse agents show that it can substantially enhance the scheduling efficiency with fairness preserved.","short_abstract":"LLM agents, which often comprise parallel inference tasks, are commonly adopted to solve real-world problems. When serving such task-parallel LLM agents in shared GPU servers, the scheduler is expected to attain fast agent completion with guaranteed worst-case performance. For that objective, our insight is to selectiv...","url_abs":"https://arxiv.org/abs/2510.17015","url_pdf":"https://arxiv.org/pdf/2510.17015v2","authors":"[\"Mingyan Yang\",\"Guanjie Wang\",\"Manqi Luo\",\"Yifei Liu\",\"Chen Chen\",\"Han Zhao\",\"Yu Feng\",\"Quan Chen\",\"Minyi Guo\"]","published":"2025-10-19T21:34:34Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.DC\"]","methods":"[\"Large Language Model\"]","has_code":false}