Straggler-Robust Distributed Optimization in Parameter-Server Networks

TL;DR

This paper introduces a robust distributed optimization algorithm designed for parameter-server networks that effectively handles network failures, delays, and stragglers through coding techniques, improving convergence and reliability.

Contribution

The paper proposes a novel distributed convex optimization algorithm that is resilient to network failures and delays by integrating coding strategies within a parameter-server framework.

Findings

Algorithm demonstrates robustness to node failures and delays.

Coding techniques improve convergence in dynamic network topologies.

Experimental results show enhanced performance over existing methods.

Abstract

Optimization in distributed networks plays a central role in almost all distributed machine learning problems. In principle, the use of distributed task allocation has reduced the computational time, allowing better response rates and higher data reliability. However, for these computational algorithms to run effectively in complex distributed systems, the algorithms ought to compensate for communication asynchrony, network node failures and delays known as stragglers. These issues can change the effective connection topology of the network, which may vary over time, thus hindering the optimization process. In this paper, we propose a new distributed unconstrained optimization algorithm for minimizing a convex function which is adaptable to a parameter server network. In particular, the network worker nodes solve their local optimization problems, allowing the computation of their local coded gradients, which will be sent to different server nodes. Then within this parameter server platform each server node aggregates its communicated local gradients, allowing convergence to the desired optimizer. This algorithm is robust to network s worker node failures, disconnection, or delaying nodes known as stragglers. One way to overcome the straggler problem is to allow coding over the network. We further extend this coding framework to enhance the convergence of the proposed algorithm under such varying network topologies. By using coding and utilizing evaluations of gradients of uniformly bounded delay we further enhance the proposed algorithm performance. Finally, we implement the proposed scheme in MATLAB and provide comparative results demonstrating the effectiveness of the proposed framework