Lightweight Robust Size Aware Cache Management

TL;DR

This paper introduces a simple, efficient extension of the TinyLFU cache policy to manage variable-sized objects, achieving comparable or better cache performance with significantly lower computational overhead.

Contribution

It presents a lightweight, easy-to-implement size-aware cache management algorithm that outperforms complex existing methods in both efficiency and effectiveness.

Findings

Achieves higher hit and byte hit ratios than state-of-the-art algorithms.

Imposes up to three times lower CPU overhead.

Requires only minor modifications to existing caching libraries.

Abstract

Modern key-value stores, object stores, Internet proxy caches, as well as Content Delivery Networks (CDN) often manage objects of diverse sizes, e.g., blobs, video files of different lengths, images with varying resolution, and small documents. In such workloads, size-aware cache policies outperform size-oblivious algorithms. Unfortunately, existing size-aware algorithms tend to be overly complicated and computationally~expensive. Our work follows a more approachable pattern; we extend the prevalent (size-oblivious) TinyLFU cache admission policy to handle variable sized items. Implementing our approach inside two popular caching libraries only requires minor changes. We show that our algorithms yield competitive or better hit-ratios and byte hit-ratios compared to the state of the art size-aware algorithms such as AdaptSize, LHD, LRB, and GDSF. Further, a runtime comparison indicates that our implementation is faster by up to x3 compared to the best alternative, i.e., it imposes much lower CPU overhead.