DocumentCode :
3376227
Title :
Modeling and Verifying Google File System
Author :
Bo Li ; Mengdi Wang ; Yongxin Zhao ; Geguang Pu ; Huibiao Zhu ; Fu Song
Author_Institution :
Shanghai Key Lab. of Trustworthy Comput., East China Normal Univ., Shanghai, China
fYear :
2015
fDate :
8-10 Jan. 2015
Firstpage :
207
Lastpage :
214
Abstract :
Google File System (GFS) is a distributed file system developed by Google for massive data-intensive applications. Its high aggregate performance of delivering massive data to many clients but the inexpensiveness of commodity hardware facilitate GFS to successfully meet the massive storage needs and be widely used in industries. In this paper, we first present a formal model of Google File System in terms of Communicating Sequential Processes (CSP#), which precisely describes the un-derlying read/write behaviors of GFS. On that basis, both relaxed consistency and eventually consistency guaranteed by GFS may be revealed in our framework. Furthermore, the suggested CSP# model is encoded in Process Analysis Toolkit (PAT), thus several properties such as starvation-free and deadlock-free could be automatically checked and verified in the framework of formal methods.
Keywords :
cloud computing; communicating sequential processes; distributed databases; formal verification; storage management; CSP; GFS; Google file system; PAT; communicating sequential processes; distributed file system; eventually consistency; formal model; massive data-intensive applications; process analysis toolkit; relaxed consistency; Analytical models; Computational modeling; Data models; File systems; Google; System recovery; CSP; Consistency; GFS; PAT;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
High Assurance Systems Engineering (HASE), 2015 IEEE 16th International Symposium on
Conference_Location :
Daytona Beach Shores, FL
Print_ISBN :
978-1-4799-8110-6
Type :
conf
DOI :
10.1109/HASE.2015.38
Filename :
7027433
Link To Document :
بازگشت