Intelligent Resource Scheduling for Co-located Latency-critical Services: A Multi-Model Collaborative Learning Approach

TL;DR

This paper introduces OSML, a multi-model machine learning-based scheduler that improves resource allocation for co-located latency-critical services by avoiding resource cliffs and enhancing QoS stability.

Contribution

It presents a novel collaborative ML approach that predicts QoS variations and intelligently guides resource scheduling to improve efficiency and stability in cloud environments.

Findings

Supports higher load levels with QoS guarantees

Reduces scheduling overhead and convergence time

Effectively avoids resource cliffs during scheduling

Abstract

Latency-critical services have been widely deployed in cloud environments. For cost-efficiency, multiple services are usually co-located on a server. Thus, run-time resource scheduling becomes the pivot for QoS control in these complicated co-location cases. However, the scheduling exploration space enlarges rapidly with the increasing server resources, making the schedulers hardly provide ideal solutions quickly. More importantly, we observe that there are "resource cliffs" in the scheduling exploration space. They affect the exploration efficiency and always lead to severe QoS fluctuations. Resource cliffs cannot be easily avoided in previous schedulers. To address these problems, we propose a novel ML-based intelligent scheduler - OSML. It learns the correlation between architectural hints (e.g., IPC, cache misses, memory footprint, etc.), scheduling solutions and the QoS demands based on a data set we collected from 11 widely deployed services running on off-the-shelf servers. OSML employs multiple ML models to work collaboratively to predict QoS variations, shepherd the scheduling, and recover from QoS violations in complicated co-location cases. OSML can intelligently avoid resource cliffs during scheduling and reach an optimal solution much faster than previous approaches for co-located LC services. Experimental results show that OSML supports higher loads and meets QoS targets with lower scheduling overheads and shorter convergence time than previous studies.