Modeling and Verifying HDFS Using CSP

Abstract

Hadoop Distributed File System (HDFS) is a high fault-tolerant distributed file system, which provides a high throughput access to application data and is suitable for applications that have large data sets. Since HDFS is widely used, analysis on it in a formal framework is of great significance. In this paper, we use Communicating Sequential Processes (CSP) to model and analyze HDFS. We mainly focus on the dominant parts which include reading files and writing files in HDFS and formalize them in detail. Moreover, we use the model checker Process Analysis Toolkit (PAT) to simulate the model constructed and verify whether it caters for the specification and some important properties, including Deadlock-freeness, Minimal Distance Scheme, Mutual Exclusion, Write-Once Scheme and Robustness.