Exploring Domestic Alternatives to Elasticsearch When ES Is Restricted

Why Look for Elasticsearch Alternatives?

Certain industries such as government and critical infrastructure impose strict software‑国产化 and controllability requirements, preventing reliance on foreign technology. Since version 7.11 Elasticsearch switched from the Apache 2.0 license to a proprietary Elastic license, creating compliance and cost concerns. Large enterprises also desire full source‑code control for customization and long‑term sustainability, even if they lack immediate capability.

Potential Domestic Substitutes

OpenSearch – an open‑source fork of Elasticsearch 7.10 maintained by AWS and the community, Apache 2.0 licensed, retains API compatibility, and can be self‑hosted in China. The international community is active, but Chinese community activity is limited.

XunSearch (讯搜) – a Chinese open‑source full‑text search solution built on C/C++ and Xapian, optimized for Chinese tokenization. Lightweight and suitable for small‑to‑medium projects, but the official website became unavailable in Dec 2024 and the last code update was two years ago.

Lucene / Solr Customization – Lucene is the core search library behind Elasticsearch; Solr is a search server built on Lucene. Both are pure Apache‑licensed, mature, and allow deep customization, but building a distributed, analytics‑rich system comparable to Elasticsearch requires significant development and operational effort.

Domestic Database Extensions – databases such as Dameng, Kingbase, and OceanBase can be extended with Lucene plugins or custom full‑text modules to add search capabilities. These solutions meet国产化 compliance but generally offer weaker search features than Elasticsearch.

Milvus (Zilliz) – an open‑source vector database for AI‑driven similarity search (image, audio, text embeddings). It targets next‑generation AI search scenarios rather than traditional full‑text retrieval.

EasySearch – a Chinese‑developed search‑oriented database that mirrors Elasticsearch’s distributed architecture while providing deep Chinese‑language optimizations, built‑in security, multi‑version cluster replication, full SQL support, and native compatibility with Chinese CPUs and operating systems. It has achieved top rankings in domestic search‑database benchmarks since October 2023.

Key Comparison Highlights

Security : EasySearch offers built‑in kernel‑level encryption, role‑based access control, LDAP/AD integration, and field‑level encryption, whereas basic Elasticsearch lacks these features and the Platinum edition adds only limited security.

Cluster Management : EasySearch supports unlimited version management, audit trails, and cloud‑native (K8s) orchestration, while Elasticsearch basic provides one‑to‑one management and the Platinum edition adds one‑to‑many management.

Scalability : EasySearch enables multi‑version cross‑cluster data replication and query expansion, surpassing Elasticsearch’s cross‑cluster search and Platinum’s same‑version replication.

Data Management : EasySearch fully supports SQL parsing (JDBC), asynchronous search, and a suite of Chinese tokenizers (ik, pinyin, etc.), whereas Elasticsearch basic offers only snapshot/recovery APIs and the Platinum edition adds limited SQL and async features.

Kernel Optimizations : EasySearch provides custom index compression, query acceleration, ZSTD compression, and unrestricted kNN vector search, while Elasticsearch basic only has basic inverted indexing.

Domestic Adaptation : EasySearch runs on Chinese CPUs (申威, 龙芯, 飞腾, 鲲鹏, 海光, 光芯) and operating systems (华为欧拉, 统信UOS, 银河麒麟), meeting strict国产化 requirements; OpenSearch and Lucene/Solr have no such adaptations.

Support & Service : EasySearch offers 7×24 rapid response and localized technical support, whereas Elasticsearch basic has no official support and the Platinum edition promises a 2‑hour response.

Localization : EasySearch includes built‑in Chinese semantic analysis plugins without extra cost, while Elasticsearch’s international design lacks Chinese‑specific optimizations.

Recommendation and Selection Guidance

Replacing Elasticsearch is a complex, multi‑step effort that requires thorough feasibility studies. Small‑to‑medium enterprises without strict国产化 mandates can continue using Elasticsearch safely. When a replacement is necessary—especially for large‑scale clusters—choose the solution that aligns with your business scenario, team expertise, and compliance needs. Early preparation is advisable if the shift becomes inevitable.

Future articles will provide deeper technical walkthroughs of EasySearch and other domestic search engines.