Project Background
According to statistics from domestic public cloud vendors, hard disk failures account for as much as 60% of all hardware failures, a figure that far exceeds other types of failures. In client feedback from data centers, issues related to latency account for more than one-third. When the storage latency of business servers spikes, it can directly lead to user request timeouts and service interruptions. Such problems not only affect user experience but can also trigger a chain reaction, such as service degradation, resource wastage, and increased operational costs. Performance issues with storage components, especially latency issues, have become a core bottleneck affecting the efficiency of data centers. Therefore, conducting research on latency in Linux storage systems and establishing a systematic method for latency measurement and analysis is of significant necessity.
Main Content of the Report
The ODCC Storage Special Group has established a project for analyzing latency in Linux storage systems, led by the Xi’an Samsung Electronics Research Institute, which has compiled the white paper titled “Measurement and Analysis of Latency in Linux Storage Systems.” This white paper aims to provide a systematic methodology and tool guide for locating and analyzing storage latency, with the main content including:
Layered Fault Model Definition: Clarifying the latency characteristics and typical fault patterns at the file system layer, block layer, and device layer.
Practical Analysis Framework: Providing a complete methodology from latency data collection/call chain tracing to root cause identification, along with real case analyses of typical manifestations of faults at each layer.
Optimization Suggestions and Trends: Discussing the impact of future storage architectures on latency fault localization and the development direction of intelligent diagnostic tools.
Through this white paper, readers will be able to quickly determine the hierarchical attribution of latency faults and construct an efficient cross-layer diagnostic toolchain. We also call on more partners to join us in building an efficient and systematic method for storage latency analysis.
This achievement will be officially released at the Open Data Center Conference in September this year! Stay tuned!
Project Manager Introduction:
Dou Kun
Xi’an Samsung Electronics Research Institute
Senior Project Manager
Contact for ODCC Storage Special Group
Teacher Gao: 17614721889 (same number on WeChat)
Email: [email protected]
Contact for ODCC Secretariat
Teacher Liu: 13488889649 (same number on WeChat)
Email: [email protected]
Previous Reviews
1. [25’ODCC Achievements] White Paper on Cloud-Edge Collaborative AI Network Technology
2. [25’ODCC Achievements] White Paper on SSD Transparent Compression Function Testing
3. [25’ODCC Achievements] Intelligent Computing Optical Interconnection Enters a New Era of Dynamic Scheduling
4. [25’ODCC Achievements] Low Voltage Power Modules for Data Centers
5. [25’ODCC Achievements] Super Partition Technology and Management Solutions for Large-Scale Distributed Block Storage