Modernizing Enterprise Search with OpenSearch, AWS Graviton 4, and Bonsai

Executive Summary

A large enterprise running a mission-critical search platform faced escalating costs and performance limits on a legacy Elasticsearch 7.10.2 deployment. In partnership with Bonsai, the organization executed a direct migration to OpenSearch 2.15 while modernizing its infrastructure on AWS Graviton 4 processors. The result: 25% lower infrastructure costs, faster query performance, 10% more disk capacity and improved disk IO, and a platform ready for AI-powered workloads.

The Challenge

The company’s existing search infrastructure had hit a wall:

• High cost: AWS infrastructure spend and increased velocity of resource expansion
• Limited scalability: 2,632 vCPUs and 116TB storage spread across an aging architecture.
• Technology constraints: Although the company was already using Graviton2, the legacy Elasticsearch 7.10.2 environment limited its ability to fully benefit from ARM performance. This prevented them from realizing efficiency gains such as reducing CPU count, lowering costs, and improving disk capacity/throughput, as well as leveraging advanced compression.
• Modernization risk: Leadership needed a path forward without prolonged disruption.
  

The Solution

Working with Bonsai, a managed search platform provider and active OpenSearch community contributor, the enterprise modernized in one decisive move:

• Direct migration: from Elasticsearch 7.10.2 to OpenSearch 2.15, avoiding the costly path to Elasticsearch 8+. The upgrade would have required reindexing at 7.17 and an upgrade
• Infrastructure rebuild: on AWS Graviton4 processors to unlock modern versions of ARM architecture performance and cost effectiveness gains.
• Adoption of advanced compression: with OpenSearch’s zstd and zstd_no_dict codecs to reduce storage requirements and accelerate queries. (These features are not available in Elasticsearch 8+.)
• Expert management: from Bonsai from day one, minimizing downtime and fine-tuning for optimal performance.

Results

• 26% cost reduction
• Better search performance: Reduced query times and improved responsiveness.
• Higher reliability: ElasticSearch gates features like searchable snapshots and ILM / ISM.
• Future readiness: Scalable platform aligned for AI workloads and modern data pipelines.

Why It Matters

Enterprises are reimagining search as part of the foundation for AI-powered experiences. For this company, migrating to OpenSearch offered them a modern search engine with key capabilities like vector search, an opportunity to reallocate infrastructure budget to human talent, and an efficient path to success by avoiding a confusing, costly, zero-sum migration to Elasticsearch 8+. With a trusted open source foundation, modern infrastructure, and the right partners, organizations can modernize quickly, scale intelligently, and keep full control over their data and future.

Technical Highlights (for deeper readers)

• Deployment scale: 1,440 vCPUs, 167TB Storage (Instance Store NVMe SSDs)
• AWS Graviton 4 processors for ARM performance and efficiency.
• OpenSearch 2.15 with advanced zstd/zstd_no_dict compression codecs.
  

Managed cluster operations by Bonsai.