v0.2.3

Replaces the forcemerge step with an await_index_ready phase that waits for all three engines to signal that indexing and optimisation are complete before benchmarking begins.

• Elasticsearch and OpenSearch: polls _stats/merge until merge activity settles
• Qdrant: polls the collection status until the optimizer reaches a green/ok state

The wait is unbounded — jingra logs progress with elapsed time and indexed/total counts rather than timing out on large collections.

docker pull sachinfrayneelastic/jingra:0.2.3

v0.2.2

Added forcemerge functionality to Elasticsearch and OpenSearch

docker pull sachinfrayneelastic/jingra:0.2.2

v0.2.0

What's new in 0.2.0

• Engine version labels in plots — series labels now include the engine version (e.g. elasticsearch-9.4.0, qdrant-1.17.0) when configured via engine_versions in jingra.yml
• Fuzzy recall matching — the speedup summary and full results now pair engine results within ±0.01 or ±0.02 rounded recall when no exact match exists, rather than silently omitting them
• Single-pass CSV export — full results and summary CSVs share one speedup calculation via exportResultsCsvs, ensuring the values are always consistent
• Elasticsearch client bump — updated to 9.4.0

docker pull sachinfrayneelastic/jingra:0.2.0