MetaNet: Automated Dynamic Selection of Scheduling Policies in Cloud Environments

TL;DR

This paper introduces MetaNet, a surrogate model that dynamically selects the most appropriate scheduling policy in cloud environments, balancing cost and performance by choosing from multiple DNN-based schedulers in real-time.

Contribution

MetaNet enables online adaptive selection of scheduling policies, improving cost-efficiency and performance over fixed or static approaches in cloud task scheduling.

Findings

Improves execution costs by up to 11%

Reduces energy consumption by up to 43%

Decreases SLA violations by up to 13%

Abstract

Task scheduling is a well-studied problem in the context of optimizing the Quality of Service (QoS) of cloud computing environments. In order to sustain the rapid growth of computational demands, one of the most important QoS metrics for cloud schedulers is the execution cost. In this regard, several data-driven deep neural networks (DNNs) based schedulers have been proposed in recent years to allow scalable and efficient resource management in dynamic workload settings. However, optimal scheduling frequently relies on sophisticated DNNs with high computational needs implying higher execution costs. Further, even in non-stationary environments, sophisticated schedulers might not always be required and we could briefly rely on low-cost schedulers in the interest of cost-efficiency. Therefore, this work aims to solve the non-trivial meta problem of online dynamic selection of a scheduling policy using a surrogate model called MetaNet. Unlike traditional solutions with a fixed scheduling policy, MetaNet on-the-fly chooses a scheduler from a large set of DNN based methods to optimize task scheduling and execution costs in tandem. Compared to state-of-the-art DNN schedulers, this allows for improvement in execution costs, energy consumption, response time and service level agreement violations by up to 11, 43, 8 and 13 percent, respectively.