Overview

On this page are the results of benchmarking different content types (JSON and SMILE) comparing indexing performance of Elasticsearch master (git revision 3505732).

Benchmarks

  • Geonames: for evaluating the performance of structured data.
  • Geopoint: for evaluating the performance of geo queries.
  • PMC: for evaluating the performance of full text search.
  • NYC taxis: for evaluating the performance for highly structured data.
  • Nested: for evaluating the performance for nested documents.

Benchmarking Methodology

All benchmarks are run by Rally. The benchmark uses two bare-metal server-class machines. On one we run the benchmark driver (Rally), on the other one the benchmark candidate (one or at most two Elasticsearch nodes).

Benchmarking Environment

Current environment

All benchmarks are run on bare-metal machines with the following specifications:

History of changes

  • 2017-04-13: Upgrade kernel from version 4.4.0-38 to 4.4.0-72