FuxiShuffle: An Adaptive and Resilient Shuffle Service for Distributed Data Processing on Alibaba Cloud

TL;DR

FuxiShuffle is a novel adaptive and resilient shuffle service designed for large-scale distributed data processing, dynamically optimizing performance and fault tolerance in Alibaba Cloud's MaxCompute platform.

Contribution

It introduces a dynamic shuffle mode selection, progress-aware scheduling, and active failure resilience mechanisms tailored for ultra-large, highly dynamic environments.

Findings

Reduces end-to-end job completion time

Lowers aggregate resource consumption

Improves adaptability and failure resilience

Abstract

Shuffle exchanges intermediate results between upstream and downstream operators in distributed data processing and is usually the bottleneck due to factors such as small random I/Os and network contention. Several systems have been designed to improve shuffle efficiency, but from our experiences of running ultra-large clusters at Alibaba Cloud MaxCompute platform, we observe that they can not adapt to highly dynamic job characteristics and cluster resource conditions, and their fault tolerance mechanisms are passive and inefficient when failures are inevitable. To tackle their limitations, we design and implement FuxiShuffle as a general data shuffle service for the ultra-large production environment of MaxCompute, featuring good adaptability and efficient failure resilience. Specifically, to achieve good adaptability, FuxiShuffle dynamically selects the shuffle mode based on runtime information, conducts progress-aware scheduling for the downstream workers, and automatically determines the most suitable backup strategy for each shuffle data chunk. To make failure resilience efficient, FuxiShuffle actively ensures data availability with multi-replica failover, prevents memory overflow with careful memory management, and employs an incremental recovery mechanism that does not lose computation progress. Our experiments show that, compared to baseline systems, FuxiShuffle significantly reduces not only end-to-end job completion time but also aggregate resource consumption. Micro experiments suggest that our designs are effective in improving adaptability and failure resilience.