Getting the MOST out of your Storage Hierarchy with Mirror-Optimized Storage Tiering

TL;DR

MOST is a new storage tiering approach that combines mirroring and tiering to improve load balancing and bandwidth utilization in modern storage hierarchies, especially under intensive workloads.

Contribution

The paper introduces Mirror-Optimized Storage Tiering (MOST), a novel method that efficiently balances load and space in storage hierarchies by combining mirroring with tiering.

Findings

Cerberus with MOST achieves higher throughput than competitors.

MOST improves bandwidth utilization under I/O-intensive workloads.

The approach balances load dynamically without costly data migrations.

Abstract

We present Mirror-Optimized Storage Tiering (MOST), a novel tiering-based approach optimized for modern storage hierarchies. The key idea of MOST is to combine the load balancing advantages of mirroring with the space-efficiency advantages of tiering. Specifically, MOST dynamically mirrors a small amount of hot data across storage tiers to efficiently balance load, avoiding costly migrations. As a result, MOST is as space-efficient as classic tiering while achieving better bandwidth utilization under I/O-intensive workloads. We implement MOST in Cerberus, a user-level storage management layer based on CacheLib. We show the efficacy of Cerberus through a comprehensive empirical study: across a range of static and dynamic workloads, Cerberus achieves better throughput than competing approaches on modern storage hierarchies especially under I/O-intensive and dynamic workloads.