中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
LASS: Reducing Cold Startup Latency in Serverless Through Loaded Library Sharing

文献类型:期刊论文

作者Yu, Zishu1,2; Zhang, Yao1,2; Li, Runfu1,2; Wang, Yifan1,2; Peng, Xiaohui1,2; Xu, Zhiwei1,2
刊名IEEE TRANSACTIONS ON COMPUTERS
出版日期2026-03-01
卷号75期号:3页码:901-915
关键词Libraries Loading Containers Computers Message systems Virtual machine monitors Serverless computing Runtime Artificial intelligence Security Cloud computing serverless cold startup library-level sharing local address space service
ISSN号0018-9340
DOI10.1109/TC.2025.3643460
英文摘要In serverless scenario, function invocation runs in an individual container. Lightweight container technology has significantly reduced the startup latency of container. The library loading process now becomes a critical performance bottleneck of serverless function cold startup. The state-of-the-art approaches leverage the process fork operation to reduce the cold startup latency in serverless computing by reusing the loaded libraries. However, the fork operation can only share libraries between parent process and forked process. For security, the libraries loaded by the parent process should be a subset of those required by the forked process, which limits opportunities to eliminate library loading overhead. To address this problem, we propose the LASS system, which enables multiple processes to share initialized libraries in a composable and efficient manner. LASS allows a process to securely reuse libraries loaded by multiple processes, thereby reducing library loading latency to the millisecond level. Compared to the state-of-the-art approaches, LASS can improve average library loading speed by more than 10.3 x, and reduce 99(th) percentile end-to-end latency by 34%-57%.
资助项目National Natural Science Foundation of China[62072434] ; National Natural Science Foundation of China[U23B2004]
WOS研究方向Computer Science ; Engineering
语种英语
WOS记录号WOS:001690579300030
出版者IEEE COMPUTER SOC
源URL[http://119.78.100.204/handle/2XEOYT63/42789]  
专题中国科学院计算技术研究所
通讯作者Peng, Xiaohui
作者单位1.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
2.Univ Chinese Acad Sci, Beijing 100190, Peoples R China
推荐引用方式
GB/T 7714
Yu, Zishu,Zhang, Yao,Li, Runfu,et al. LASS: Reducing Cold Startup Latency in Serverless Through Loaded Library Sharing[J]. IEEE TRANSACTIONS ON COMPUTERS,2026,75(3):901-915.
APA Yu, Zishu,Zhang, Yao,Li, Runfu,Wang, Yifan,Peng, Xiaohui,&Xu, Zhiwei.(2026).LASS: Reducing Cold Startup Latency in Serverless Through Loaded Library Sharing.IEEE TRANSACTIONS ON COMPUTERS,75(3),901-915.
MLA Yu, Zishu,et al."LASS: Reducing Cold Startup Latency in Serverless Through Loaded Library Sharing".IEEE TRANSACTIONS ON COMPUTERS 75.3(2026):901-915.

入库方式: OAI收割

来源:计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。