Opportunistic Scheduling for Optimal Spot Instance Savings in the Cloud

TL;DR

This paper develops an analytical framework for scheduling delay-sensitive cloud jobs across spot and on-demand instances, optimizing costs while respecting delay constraints, and introduces an adaptive algorithm with near-optimal performance.

Contribution

It provides the first analytical treatment of scheduling delay-sensitive jobs in cloud environments using queuing theory, deriving optimal policies and designing an adaptive scheduling algorithm.

Findings

Queue length one is optimal for low delay targets.

A knapsack structure is identified for high delay targets.

The proposed adaptive algorithm achieves near-optimal cost savings.

Abstract

We study the problem of scheduling delay-sensitive jobs over spot and on-demand cloud instances to minimize average cost while meeting an average delay constraint. Jobs arrive as a general stochastic process, and incur different costs based on the instance type. This work provides the first analytical treatment of this problem using tools from queuing theory, stochastic processes, and optimization. We derive cost expressions for general policies, prove queue length one is optimal for low target delays, and characterize the optimal wait-time distribution. For high target delays, we identify a knapsack structure and design a scheduling policy that exploits it. An adaptive algorithm is proposed to fully utilize the allowed delay, and empirical results confirm its near-optimality.