On Distributed Storage Allocations of Large Files for Maximum Service Rate

TL;DR

This paper analyzes how different distributed storage allocations impact the service rate for large files, revealing that optimal strategies depend on system parameters when download times are significant.

Contribution

It provides a detailed service rate analysis for quasi-uniform storage allocations under realistic download time assumptions, highlighting when optimal strategies vary.

Findings

Optimal storage allocation depends on system parameters.

Minimal spreading is not always optimal when download time scales with data size.

Different data access models influence the best storage allocation strategy.

Abstract

Allocation of (redundant) file chunks throughout a distributed storage system affects important performance metrics such as the probability of file recovery, data download time, or the service rate of the system under a given data access model. This paper is concerned with the service rate under the assumption that the stored data is large and its download time is not negligible. We focus on quasi-uniform storage allocations and provide a service rate analysis for two common data access models. We find that the optimal allocation varies in accordance with different system parameters. This was not the case under the assumption that the download time does not scale with the size of data, where the minimal spreading allocation was previously found to be universally optimal.