Request latency directly affects the performance of modern cloud applications. Due to various causes in hosts and networks, requests can suffer from request latency anomalies (RLAs), which may violate the Service-Level Agreement. However, existing performance monitoring tools have incomplete coverage and inconsistent semantics for monitoring requests and cannot accurately diagnose RLAs. This paper presents <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope , a high-coverage and low-overhead request event monitoring system, which monitors <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">buffers to capture most RLA-related abnormal events with consistent request-level semantics in the end-to-end datapath of request. First, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope models the datapath of request as a buffer chain and defines events based on three properties of buffers, so as to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">end-to-end monitor the root causes of RLA. Then, to achieve <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">consistent semantics for captured events, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope designs a request-level semantics injection mechanism to make events captured in networks have the victim requests’ ID. Finally, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope offloads the semantics operations and event collection in software to SmartNICs for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">low CPU overhead . We have implemented <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope on commodity SmartNICs and programmable switches. Evaluation results show that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BufScope can diagnose 98% RLAs with < 0.08% network bandwidth overhead and 0.6% application throughput decline.

Kaihui Gao

Chen Sun

Shuai Wang

Dan Li

Yu Zhou

Hongqiang Harry Liu

Lingjun Zhu

Ming Zhang

Xiang Deng

Cheng Zhou

Lu Lu

IEEE/ACM Transactions on Networking

IEEE/ACM TRANSACTIONS ON NETWORKING 的高层次目标是发表来自通信/计算机网络领域的理论或实验探索的高质量原创研究成果，涵盖各种物理层技术上的各种信息传输网络，有线（各种引导媒体：例如，铜、光学）和无线（例如，射频、声学（例如，水下）、红外线），或这些的混合。该杂志欢迎报告新经验和实际系统实验的应用贡献。

The IEEE/ACM Transactions on Networking aims to publish high-quality original research results from the field of communication and computer networking. These studies can be theoretical or experimental, covering various physical layer technologies for information transmission networks, including wired networks (such as copper and optical media) and wireless networks (such as radio frequency, acoustic (e.g., underwater), and infrared), as well as hybrid applications of these technologies. The journal welcomes reports of new experiences and practical system experimental contributions.

IEEE/ACM Transactions on Networking 的目标是发表通信和计算机网络领域的高质量原创研究成果。这些研究可以是理论性的，也可以是实验性的，涵盖各种物理层技术的信息传输网络，包括有线网络（如铜缆和光纤）和无线网络（如射频、声学（例如水下声学）和红外线），以及这些技术的混合应用。该期刊欢迎报告新的经验和实际系统实验的应用贡献。

Buffer-Based High-Coverage and Low-Overhead Request Event Monitoring in the Cloud

BufScope is presented, a high-coverage and low-overhead request event monitoring system, which monitors buffers to capture most RLA-related abnormal events with consistent request-level semantics in the end-to-end datapath of request.

Category	Quartile
计算机科学	3
计算机科学, 计算机理论方法	2
计算机科学, 计算机硬件	3
计算机科学, 工程电子与电气	3
计算机科学, 电信学	3

Buffer-Based High-Coverage and Low-Overhead Request Event Monitoring in the Cloud

ivySCI AI Smartly Parses PDF, Answers Researchers' Questions, and Helps You Understand Papers in Seconds

Journal Info