{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T05:12:08Z","timestamp":1788844328284,"version":"build-2803163510"},"reference-count":38,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,7]]},"DOI":"10.1109\/icws72778.2026.00129","type":"proceedings-article","created":{"date-parts":[[2026,9,7]],"date-time":"2026-09-07T19:10:21Z","timestamp":1788808221000},"page":"1083-1090","source":"Crossref","is-referenced-by-count":0,"title":["QueueBreak: A Trace-to-Diagnosis Pipeline for Tail Latency in Agentic LLM Services"],"prefix":"10.1109","author":[{"given":"Xingrao","family":"Ma","sequence":"first","affiliation":[{"name":"Lingnan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zongxi","family":"Li","sequence":"additional","affiliation":[{"name":"Lingnan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chengzu","family":"Dong","sequence":"additional","affiliation":[{"name":"Lingnan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Di","family":"Shao","sequence":"additional","affiliation":[{"name":"Pengcheng Lab","place":["Shenzhen, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiao","family":"Liu","sequence":"additional","affiliation":[{"name":"School of IT, Deakin University","place":["Melbourne, Australia"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"ReAct: Synergizing reasoning and acting in language models","volume-title":"The Eleventh International Conference on Learning Representations","author":"Yao","year":"2023"},{"key":"ref2","doi-asserted-by":"crossref","DOI":"10.52202\/075280-2997","article-title":"Toolformer: Language models can teach themselves to use tools","author":"Schick","year":"2023","journal-title":"Advances in Neural Information Processing Systems 36"},{"key":"ref3","doi-asserted-by":"crossref","DOI":"10.52202\/079017-4020","article-title":"Gorilla: Large language model connected with massive APIs","author":"Patil","year":"2024","journal-title":"Advances in Neural Information Processing Systems 37"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1016\/j.caeai.2025.100417"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/2408776.2408794"},{"key":"ref6","article-title":"Flashattention-2: Faster attention with better parallelism and work partitioning","volume-title":"The Twelfth International Conference on Learning Representations","author":"Dao","year":"2024"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref8","doi-asserted-by":"crossref","DOI":"10.52202\/079017-2000","article-title":"SGLang: Efficient execution of structured language model programs","volume-title":"Advances in Neural Information Processing Systems 37","author":"Zheng","year":"2024"},{"key":"ref9","first-page":"193","article-title":"Distserve: Disaggregating prefill and decoding for goodput-optimized large language model serving","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Zhong","year":"2024"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3419111.3421285"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507721"},{"issue":"1","key":"ref12","first-page":"1","article-title":"Dapper, a large-scale distributed systems tracing infrastructure","volume":"44","author":"Sigelman","year":"2010","journal-title":"ACM SIGOPS Operating Systems Review"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3472883.3487005"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3642970.3655835"},{"key":"ref15","article-title":"S-LoRA: Serving thousands of concurrent LoRA adapters","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Sheng","year":"2024"},{"key":"ref16","first-page":"117","article-title":"Taming Throughput-Latency tradeoff in LLM inference with Sarathi-Serve","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Agrawal","year":"2024"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/isca59077.2024.00019"},{"key":"ref18","first-page":"173","article-title":"Llumnix: Dynamic scheduling for large language model serving","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Sun","year":"2024"},{"key":"ref19","first-page":"155","article-title":"Mooncake: Trading more storage for less computation-a KVCache-centric architecture for serving LLM chatbot","volume-title":"23rd USENIX Conference on File and Storage Technologies (FAST 25)","author":"Qin","year":"2025"},{"key":"ref20","first-page":"31094","article-title":"Flexgen: Highthroughput generative inference of large language models with a single GPU","volume-title":"Proceedings of the 40th International Conference on Machine Learning","author":"Sheng","year":"2023"},{"key":"ref21","first-page":"521","article-title":"ORCA: A distributed serving system for transformer-based generative models","volume-title":"16th USENIX Symposium on Operating Systems Design and Implementation (OSDI 22)","author":"Yu","year":"2022"},{"key":"ref22","first-page":"135","article-title":"ServerlessLLM: Low-latency serverless inference for large language models","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Fu","year":"2024"},{"key":"ref23","article-title":"LoRA: Low-rank adaptation of large language models","volume-title":"The Tenth International Conference on Learning Representations","author":"Hu","year":"2022"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1016\/j.ipm.2025.104568"},{"key":"ref25","article-title":"ShadowPEFT: Shadow network for parameter-efficient fine-tuning","author":"Li","year":"2026","journal-title":"arXiv Preprint"},{"key":"ref26","first-page":"38087","article-title":"Smoothquant: Accurate and efficient post-training quantization for large language models","volume-title":"Proceedings of the 40th International Conference on Machine Learning","author":"Xiao","year":"2023"},{"key":"ref27","article-title":"AWQ: Activation-aware weight quantization for on-device LLM compression and acceleration","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Lin","year":"2024"},{"key":"ref28","article-title":"QServe: W4A8KV4 quantization and system co-design for efficient LLM serving","volume-title":"Proceedings of Machine Learning and Systems","volume":"7","author":"Lin","year":"2025"},{"key":"ref29","first-page":"9397","article-title":"GreenTune: Energy-efficient low-rank tuning of LLMs with ThreeE evaluation under 4-\/8-bit quantization","volume-title":"Proceedings of the ACM Web Conference 2026","author":"Ma","year":"2026"},{"key":"ref30","article-title":"Retrieval-augmented generation for knowledge-intensive NLP tasks","volume-title":"Advances in Neural Information Processing Systems 33","author":"Lewis","year":"2020"},{"key":"ref31","article-title":"Self-RAG: Learning to retrieve, generate, and critique through self-reflection","volume-title":"The Twelfth International Conference on Learning Representations","author":"Asai","year":"2024"},{"key":"ref32","first-page":"3006","article-title":"Recurrent attention networks for long-text modeling","author":"Li","year":"2023","journal-title":"Findings of the Association for Computational Linguistics: ACL 2023. Association for Computational Linguistics"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/3691620.3695360"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1145\/3472883.3486985"},{"key":"ref35","first-page":"613","article-title":"Clipper: A low-latency online prediction serving system","volume-title":"14th USENIX Symposium on Networked Systems Design and Implementation (NSDI 17)","author":"Crankshaw","year":"2017"},{"key":"ref36","first-page":"443","article-title":"Serving DNNs like clockwork: Performance predictability from the bottom up","volume-title":"14th USENIX Symposium on Operating Systems Design and Implementation (OSDI 20)","author":"Gujarati","year":"2020"},{"key":"ref37","first-page":"663","article-title":"AlpaServe: Statistical multiplexing with model parallelism for deep learning serving","volume-title":"17th USENIX Symposium on Operating Systems Design and Implementation (OSDI 23)","author":"Li","year":"2023"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1145\/2815400.2815415"}],"event":{"name":"2026 IEEE International Conference on Web Services (ICWS)","location":"Sydney, Australia","start":{"date-parts":[[2026,7,13]]},"end":{"date-parts":[[2026,7,18]]}},"container-title":["2026 IEEE International Conference on Web Services (ICWS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11670018\/11670022\/11670158.pdf?arnumber=11670158","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T04:53:33Z","timestamp":1788843213000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11670158\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7]]},"references-count":38,"URL":"https:\/\/doi.org\/10.1109\/icws72778.2026.00129","relation":{},"subject":[],"published":{"date-parts":[[2026,7]]}}}