Straggler Mitigation with Tiered Gradient Codes

Sasi, S and Lalitha, V and Aggarwal, V and Rajan, BS (2020) Straggler Mitigation with Tiered Gradient Codes. In: IEEE Transactions on Communications, 68 (8). pp. 4632-4647.

PDF iee_tra_com_68-08_4632-4647_2020.pdf - Published Version Restricted to Registered users only Download (1MB) | Request a copy

Abstract

Coding theoretic techniques have been proposed for synchronous Gradient Descent (GD) on multiple servers to mitigate stragglers. These techniques provide the flexibility that the job is complete when any k out of n servers finish their assigned tasks. The task size on each server is found based on the values of k and n. However, it is assumed that all the n jobs are started when the job is requested. In contrast, we assume a tiered system, where we start with n1 � k tasks, and on completion of c tasks, we start n2-n1 more tasks. The aim is that as long as k servers can execute their tasks, the job gets completed. This paper exploits the flexibility that not all servers are started at the request time to obtain the achievable task sizes on each server. The task sizes are in general lower than starting all n2 tasks at the request times thus helping achieve lower task sizes which helps to reduce both the job completion time and the total server utilization. © 2020 IEEE.

Item Type:	Journal Article
Publication:	IEEE Transactions on Communications
Publisher:	Institute of Electrical and Electronics Engineers Inc.
Additional Information:	The copyright of this article belongs to Institute of Electrical and Electronics Engineers Inc.
Keywords:	Communication; Electronics engineering, Gradient descent; Job completion; K-out-of-n; K-server; Multiple servers; Tiered System, Gradient methods
Department/Centre:	Division of Electrical Sciences > Electrical Communication Engineering
Date Deposited:	08 Oct 2020 11:05
Last Modified:	08 Oct 2020 11:05
URI:	http://eprints.iisc.ac.in/id/eprint/66607

Actions (login required)

View Item