Advancing Data Resilience: An In-Depth Study of Pyramid Codes for Fault Tolerance and Repair Cost Optimization

In the digital era, distributed storage systems are essential for managing vast volumes of data generated by cloud computing, big data, and the Internet of Things. These systems face challenges such as scalability, fault tolerance, and cost control. Pyramid Codes have emerged as an effective solutio...

Full description

Saved in:
Bibliographic Details
Main Author: Chen Yanqiu
Format: Article
Language:English
Published: EDP Sciences 2025-01-01
Series:ITM Web of Conferences
Online Access:https://www.itm-conferences.org/articles/itmconf/pdf/2025/04/itmconf_iwadi2024_03020.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the digital era, distributed storage systems are essential for managing vast volumes of data generated by cloud computing, big data, and the Internet of Things. These systems face challenges such as scalability, fault tolerance, and cost control. Pyramid Codes have emerged as an effective solution, enhancing fault tolerance and reducing repair costs. This paper presents a thorough analysis of Pyramid Codes, examining their structure, performance, and advantages over traditional Reed-Solomon Codes. The study delves into fault tolerance mechanisms, demonstrating Pyramid Codes’ ability to handle single, double, and triple disk failures efficiently. Additionally, strategies are proposed for optimizing reliability, such as increasing parity length and creating additional parity layers, which enhance repair capabilities without substantially increasing costs. The potential applications of Pyramid Codes in scenarios requiring high fault tolerance and low repair costs are discussed, emphasizing their role in advancing distributed storage resilience. Future directions include exploring adaptive frameworks for real-time data repair and optimizing parity distribution to balance security and cost-efficiency, paving the way for robust and efficient storage solutions in distributed systems.
ISSN:2271-2097