Evaluate storage capabilities in AI scenario Comprehensively
CCF Storage Evaluation
CCF Storage Evaluation defines and standardizes a set of evaluation methods and tool for evaluating the comprehensive capabilities of storage systems, including fine-grained file system tracing, I/O load replay in real service scenarios, automatic file system test, and low overhead. Generates and executes any configurable combination of workloads based on service scenario characteristics to accurately describe and evaluate file systems, storage behaviors, and features. Here we publish the first toolset aiming for AI applications and scenario.
Storage Evaluation System
Storage-evaluation-system (SES) provides comprehensive capabilities for evaluating the ecology, performance, data security, reliability, capacity, energy efficiency and other aspects that a storage solution can provide.The benchmark is protocol-independent and will run on any storage product that supports block, file, or object storage protocols.
SES needs to run on Linux, and can be used to test any storage services based on these systems. The core evaluation scenarios that SES can provide: baseline function verification, baseline performance evaluation, and interactive evaluation.
Downdload files list
File Version (effective date) Download
Storage Evaluation System(SES) intall package V1.0 Link
Storage Evaluation System(SES) source code V1.0 Link
Storage Evaluation System(SES) user manual V1.0 Link
CCF Storage Evaluation Specifications V1.0 Link
Evaluation Guide in AI Scenario V1.0 Link
Storage Evaluation System(SES) dependencies V1.0 Link
CCF Storage Testing Working Committee
Chairman
XIAO Nong Professor Sun Yat-sen University
Executive chairman
WU Yong Wei Professor Tsinghua University
ZHOU Ke Professor Huazhong University of Science and Technology
Members
CHEN Zhi Guang Professor Sun Yat-sen University
ZHANG Wu Sheng Senior Engineer Tsinghua University
CHEN Jian Xi Associate Professor Huazhong University of Science and Technology
DONG Yong Researcher National University of Defense Technology
ZHANG Xiao Professor Northwestern Polytechnical University
ZENG Ling Fang Researcher Zhijiang Laboratory
REN Feng Research Associate Qiyuan Laboratory
Contact:ZHANG Wu Sheng Email: zws@tsinghua.edu.cn