Write-Read Decoupling in Modern Large-Scale Search Engines: Architectures, Techniques, and Emerging Approaches
Abstract
Large-scale search engines face a fundamental tension: the index must be updated frequently to maintain freshness, yet updates create resource contention that inflates query latency. In the dominant Lucene-based architecture, segment merges triggered by writes compete with concurrent queries for CPU cycles, disk I/O bandwidth, and operating-system page cache -- a problem we term \emph{write-read contention}. This survey systematically examines the architectural solutions that industry and academia have developed to decouple write pressure from read latency. We identify five principal patterns: (i)~node-level read-write separation; (ii)~compute-storage separation; (iii)~full in-memory indexing; (iv)~log-structured write paths; and (v)~in-place partial updates. We survey representative systems including Elasticsearch, LinkedIn Galene, Uber Sia, Quickwit, Alibaba Havenask, Algolia, Milvus, and Vespa, and discuss an emerging synthesis -- the ScaleSearch architecture -- that combines compute-storage separation with full in-memory indexing and dedicated write nodes. A key contribution of ScaleSearch is \emph{per-field update routing}: each field is assigned its own Kafka topic and update path, allowing scalar fields (price, stock, tags) to be updated in-place in RAM with immediate visibility while full-text fields follow the segment-based compute-storage path. We conclude with open challenges in hybrid vector-and-full-text retrieval, serverless deployments, and AI-integrated search.
Cite
@article{arxiv.2605.01260,
title = {Write-Read Decoupling in Modern Large-Scale Search Engines: Architectures, Techniques, and Emerging Approaches},
author = {Xin Liang and Qing Yang and Wenru Qiu and Wenjie Mao and Tianyu Ma and Minghui Zhu and Nan Wang},
journal= {arXiv preprint arXiv:2605.01260},
year = {2026}
}
Comments
8 pages, 5 figures