Cache-Aware I/O Cost Modeling for Disk-Based Learned Indexes

2026-06-20

AI Summary

Existing learned indexes lack a cache-aware I/O cost model, hindering effective tuning and query optimization. This work proposes the Cache-Aware Model (CAM), which, for the first time, integrates realistic cache replacement policies into I/O modeling for learned indexes. By estimating page access distributions and coupling them with cache mechanisms, CAM efficiently predicts physical I/O costs without requiring full trace replay. The model is readily adaptable to mainstream learned index structures such as PGM-index and RMI, and further extends to learned index–driven join optimization. Experimental results demonstrate that CAM-guided tuning improves throughput by 1.17× for PGM and 1.66× for RMI, while end-to-end join query performance achieves up to an 8.8× speedup.

Abstract

Learned indexes have shown attractive space-time trade-offs in main-memory settings, yet a principled I/O cost model for their disk-resident deployments is still missing, which is a prerequisite for index tuning and query optimization. The practically employed page buffer makes the problem even harder: under typical cache policies, many of the logical page references issued by the index are served by the buffer rather than reaching disk, so the effective physical I/O depends jointly on the workload, the cache policy, and the index configuration. In this paper, we propose CAM, the \textit{first} cache-aware I/O cost model for learned indexes that takes practical cache eviction policies into consideration. CAM is not tied to a particular learned index design: it estimates page access distributions without full trace replay for mainstream learned index designs, and then combines them with I/O cost models to estimate effective physical I/Os. This formulation enables principled knob tuning by explicitly modeling the trade-off between index footprint and buffer capacity. We instantiate CAM for disk-based PGM-index and RMI, and further apply the same modeling principle to learned-index-based joins through a hybrid strategy that adaptively chooses point or range probes based on local key density. Extensive experiments on real benchmarks show that CAM provides \textit{accurate and efficient} I/O estimation across diverse workloads: CAM-guided tuning improves PGM throughput by \textbf{1.17$\times$} over multicriteria PGM tuning and improves RMI throughput by \textbf{1.66$\times$} over CDFShop with I/O-related considerations. For learned-index-based joins, our hybrid strategy improves end-to-end performance by up to \textbf{8.8$\times$} over disk-based index nested-loop join.

Problem

• learned indexes
• I/O cost modeling
• cache-aware
• disk-based indexing
• buffer management

Innovation

• learned indexes
• I/O cost modeling
• cache-aware
• disk-based indexing
• adaptive join strategy