OpenSearch benchmarking against Elasticsearch

Hi all,

Just a general topic seeking for our community’s opinion on this benchmarking conducted by ElasticCloud Elasticsearch vs OpenSearch: Performance and Resource Utilization Analysis | Elastic Blog.

I assume there are still quite a few common codes and still based on Lucene, curious what’s the performance gap can come from? Or has anyone trying to reproduce the benchmarking?


The OpenSearch community sees performance as one of the most important attributes of any search engine. The community is continuously engaged with improving performance from the core through the plugins. OpenSearch maintainers have also been working on reliable benchmarks to ensure the most effective use of its many dependencies, such as Lucene, and to ensure every bit of performance is optimized. You can follow along at OpenSearch or contribute to this analysis effort yourself on github at GitHub - opensearch-project/opensearch-benchmark: OpenSearch Benchmark - a community driven, open source project to run performance tests for OpenSearch. We would love for more members of the community to join hands and contribute.


Seems strange that for a code base that forked a bit less than two years ago, there should be so many performance disparities.
I’d like to see the VM sizing for these tests.

Can we use Opensearch benchmark for testing with Search engine as Elastic Search?