Making Wide Stripes Practical: Cascaded Parity LRCs for Efficient Repair and High Reliability

2025-12-11

AI Summary

Local Reconstruction Codes (LRCs) in wide-stripe erasure coding suffer from structural limitations: enlarged local groups increase single-node repair overhead; frequent multi-node failures trigger costly global repairs; and reliability degrades sharply. Method: This paper proposes Cascaded Parity LRC (CP-LRC), the first LRC design to establish structured dependencies between local and global parity blocks. It decomposes global parity information and embeds it into local parity groups, forming a cascaded parity structure that enables coordinated local–global repair—while preserving MDS fault tolerance. CP-LRC is instantiated via a finite-field coefficient generation framework and cascade-aware repair algorithms, yielding two variants: CP-Azure and CP-Uniform. Contribution/Results: Real-world deployment on Alibaba Cloud demonstrates 41% and 26% reductions in repair time for single- and double-node failures, respectively, significantly improving both repair efficiency and system reliability.

Technology Category

• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Search and Optimization: Distributed Search

Application Category

• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Erasure coding with wide stripes is increasingly adopted to reduce storage overhead in large-scale storage systems. However, existing Locally Repairable Codes (LRCs) exhibit structural limitations in this setting: inflated local groups increase single-node repair cost, multi-node failures frequently trigger expensive global repair, and reliability degrades sharply. We identify a key root cause: local and global parity blocks are designed independently, preventing them from cooperating during repair. We present Cascaded Parity LRCs (CP-LRCs), a new family of wide stripe LRCs that embed structured dependency between parity blocks by decomposing a global parity block across all local parity blocks. This creates a cascaded parity group that preserves MDS-level fault tolerance while enabling low-bandwidth single-node and multi-node repairs. We provide a general coefficient-generation framework, develop repair algorithms exploiting cascading, and instantiate the design with CP-Azure and CP-Uniform. Evaluations on Alibaba Cloud show reductions in repair time of up to 41% for single-node failures and 26% for two-node failures.

Problem

• Reduces single-node repair cost in wide stripe LRCs
• Minimizes expensive global repair for multi-node failures
• Enhances reliability while preserving MDS fault tolerance

Innovation

• Cascaded parity LRCs embed structured dependency between parity blocks
• Decompose global parity across local parity blocks for cooperation
• Enable low-bandwidth single-node and multi-node repairs with MDS-level fault tolerance