TraceFlow: Efficient Trace Analysis for Large-Scale Parallel Applications via Interaction Pattern-Aware Trace Distribution
Yuyang Jin, Xirui Shui, Mingshu Zhai, Zan Zong, Feng Zhang, Felix Wolf, Jidong Zhai
2025年份
3被引次数
摘要
Trace analysis of large-scale parallel applications is crucial for understanding and optimizing performance. It primarily focuses on the interaction behaviors between different parallel processes, such as synchronization waits and asynchronous overlaps. The trace size explodes as the parallel scale of applications, thus current methods analyze traces in parallel to ensure analysis speed. However, due to the interaction pattern-agnostic trace distribution, they often introduce inter-process communications to fetch non-local event data during interaction analysis, leading to excessively long trace analysis time.
问问这篇 Paper
问问你的智能体。
Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。
相关 Paper
- PerFlow: a domain specific framework for automatic performance analysis of parallel applicationsYuyang Jin, Haojie Wang, Runxin Zhong, Chen Zhang 等PPoPP 2022 · 被引用 10 次
- Traveler: Navigating Task Parallel Traces for Performance AnalysisSayef Azad Sakin, Alex Bigelow, R. Tohid, Connor Scully-Allison 等IEEE VIS 2022 · 被引用 10 次
- ScalAna: automating scaling loss detection with graph analysisYuyang Jin, Haojie Wang, Teng Yu, Xiongchao Tang 等SC 2020 · 被引用 16 次
- C.A.T.S.: Memory and Control Flow Tracing for Whole-Program Performance AnalysisPhilipp Schaad, Tal Ben-Nun, Torsten HoeflerSC 2025 · 被引用 1 次
- EXIST: Enabling Extremely Efficient Intra-Service Tracing Observability in DatacentersXinkai Wang, Xiaofeng Hou, Chao Li, Yuancheng Li 等ASPLOS 2025 · 被引用 4 次
