1.12.0

What's Changed

• bump patch version by @OVI3D0 in #695
• Change min/max to overall_min/overall_max + update comparison results publisher by @OVI3D0 in #692
• Fix workload parameter processing when a key is mapped to a JSON file by @gkamat in #690
• Update README Badges by @IanHoang in #697
• Add calculate-recall parameter to vector search and skip calculating recall if number clients > cpu cores by @finnroblin in #626
• Add support for custom iteration values by @OVI3D0 in #696
• fix fetch for workload params by @OVI3D0 in #704
• update integ-test + unit-test yaml files by @OVI3D0 in #707
• Refactor aggregate by @OVI3D0 in #708
• Fix URLs for ml-model runners by @gkamat in #719
• Make query randomization more flexible by @peteralfonsi in #712
• Replaced occurrences of "Benchmark" with "OSB" by @gkamat in #705
• update maintainers list by @cgchinmay in #723
• Push image to Docker Hub staging repository on a successful build by @gkamat in #717
• Changes to support ramp-up feature by @rishabh6788 in #725
• Change OSB Config directory from ~/.benchmark to ~/.osb by @OVI3D0 in #732
• update to JDK 21 for integ tests by @OVI3D0 in #737
• add improved error messaging in Worker class by @OVI3D0 in #741
• Update Docker-build Workflow to Address Segfault Issue by @IanHoang in #756
• Interim workaround for QEMU bug that causes the Docker ARM build to fail by @gkamat in #757
• Fixes to update the Java version for integration tests by @gkamat in #751
• Revert "Update Docker-build Workflow to Address Segfault Issue " by @IanHoang in #758
• Add load-test-qps parameter for dynamic client ramp up by @OVI3D0 in #739
• Set GitHub ID token to be able to acquire credentials for the Docker build by @gkamat in #761
• Fix repo specification for push to Docker Hub by @gkamat in #762
• Workaround for the QEMU bug for the DockerHub build by @gkamat in #763
• Add shorthand version of opensearch-benchmark invocation, osb by @IanHoang in #745
• Skip running the Docker release workflow in forks of the authoritative repository by @gkamat in #766
• Change Topic name used by PyPI back to Benchmark from OSB (#768) by @gkamat in #770

Full Changelog: 1.11.0...1.12.0

1.11.0

What's Changed

• add automatic testing and aggregation to OSB by @OVI3D0 in #655
• Add support for multi-part data corpora downloads by @gkamat in #677
• Minor updates for multi-part downloads by @gkamat in #678
• Add shorthand alternatives for command line options in OSB by @OVI3D0 in #660
• Bump minor version in main by @IanHoang in #674
• Add new issue template by @IanHoang in #676
• store aggregated results in separate folder by @OVI3D0 in #683
• add relative standard deviation to aggregated test execution metrics by @OVI3D0 in #681
• Add new maintainer Michael Oviedo by @IanHoang in #686
• Remove unnecessary logging by @VijayanB in #689

Full Changelog: 1.10.0...1.11.0

1.10.0

What's Changed

• Support Vector Search Nested benchmark by @finnroblin in #584
• Bump version to 1.10.0 by @gkamat in #623
• Add approximate filters by @finnroblin in #607
• Refactored the OSB build system by @gkamat in #632
• Specify multiple search clients for easier benchmarking by @finnroblin in #614
• Add tag as the release branch in jenkinsfile release by @peterzhuamazon in #640
• More updates to the OSB build by @gkamat in #635
• Address string issue in Groovy by @IanHoang in #646
• Update release guide with more details by @IanHoang in #651
• Remove small-nested-works.hdf5 by @finnroblin in #637
• Add aggregate command by @OVI3D0 in #638
• change op-metrics to op_metrics in aggregator file by @OVI3D0 in #657
• Fix --number-of-docs bug in create-workload by @IanHoang in #659

Full Changelog: 1.9.0...1.10.0

1.9.1

What's Changed

• Support Vector Search Nested benchmark by @finnroblin in #584
• Bump version to 1.10.0 by @gkamat in #623
• Bump 1.9 to 1.9.1 by @IanHoang in #624
• [Backport 1.9] Add tag as the release branch in jenkinsfile release by @opensearch-trigger-bot in #642
• Build updates and fix for docker mount issue in v1.9 by @gkamat in #643

Full Changelog: 1.9.0...1.9.1

1.9.0

What's Changed

• Add PIP_ONLY_BINARY to Dockerfile by @IanHoang in #596
• Added runners for enabling concurrent segment search via settings by @martin-gaievski in #591
• Add a Jinja filter to aid with version checking by @gkamat in #599
• Bump version to 1.9.0 by @IanHoang in #597
• Add docker build as part of checks for x64/arm64 by @peterzhuamazon in #601
• Preset opensearch-benchmark-git dir to support dockerfile change in #601 by @peterzhuamazon in #603
• Corrected setting name for max_slice_count by @martin-gaievski in #602
• Refactor build env logic to switch between testing and production by @peterzhuamazon in #604
• Add benchmark support for vector radial search by @junqiu-lei in #546
• Add docker-compose-tests.yml back in to fix integ tests by @IanHoang in #605
• Enforce docker-compose to use the v1 version in IT by @peterzhuamazon in #606
• Provide instructions on how to run integration tests locally by @gkamat in #608
• [Create-Workload Enhancements] Rearchitect Create-Workload Feature by @IanHoang in #609
• Updates to the expand-data-corpus script by @gkamat in #612
• add default value for master-node when provisioning cluster using OSB by @OVI3D0 in #621

New Contributors

• @martin-gaievski made their first contribution in #591
• @junqiu-lei made their first contribution in #546
• @OVI3D0 made their first contribution in #621

Full Changelog: 1.8.0...1.9.0

1.8.0

What's Changed

• Bump 1.7.0 to 1.8.0 by @IanHoang in #560
• Fix Numpy & h5py bug in 1.7.0 by @IanHoang in #562
• Add back h5py restriction in Dockerfile to unblock Docker Build for ARM images by @IanHoang in #568
• added Triage document for weekly triages by @cgchinmay in #547
• Package pbzip2 binaries, so they get installed automatically by @gkamat in #551
• Fix h5py build issues encountered during the release process by @gkamat in #574
• Backport feature for easier backport when patching by @IanHoang in #575
• Fixed an error in DeleleMlModel runner. by @vpehkone in #537
• Adds --randomization-alpha flag controlling distribution of repeated queries by @peteralfonsi in #564
• Download pre-generated offset files only when needed by @gkamat in #573
• Check that the in-memory metrics store does not run out of memory by @gkamat in #580
• Add TrainKNN Runner/Operation for Benchmarking Approximate KNN Algorithms by @finnroblin in #556
• Add Badges to README by @IanHoang in #588
• Fix bracket in Chat on Forum Badge by @IanHoang in #590
• Publish recall as a kpi metric by @finnroblin in #581

New Contributors

• @finnroblin made their first contribution in #556

Full Changelog: 1.7.0...1.8.0

1.7.1

What's Changed

• Bump patch version from 1.7.0 to 1.7.1 by @IanHoang in #561
• Patch 1.7 by @IanHoang in #566

Full Changelog: 1.7.0...1.7.1

1.7.0

What's Changed

• Bump version to 1.7.0 after 1.6.0 release by @IanHoang in #539
• Download pre-generated offset file from data corpus repository, if present by @gkamat in #542
• Update issue templates for better user experience and incorporate OSB logo into README by @IanHoang in #534
• Update README to show actual image logo by @IanHoang in #544
• Excluding opensearch-benchmarkd process while finding benchmark process by @backslasht in #552
• Removing the duplicate process check based on truncated name and relying on cmdline check by @backslasht in #553
• Add email addresses for Maintainers on MAINTAINERS.md by @IanHoang in #554

New Contributors

• @backslasht made their first contribution in #552

Full Changelog: 1.6.0...1.7.0

1.6.0

What's Changed

• Update h5py in dockerfile by @IanHoang in #521
• Bump version to 1.6.0 by @IanHoang in #523
• Add support to big-ann ground truth file format by @VijayanB in #528
• Update setup.py to reflect current Maintainers and Authors and updated project urls by @IanHoang in #529
• Adding Vijayan as OpenSearch Benchmark Maintainer by @IanHoang in #533

Full Changelog: 1.5.0...1.6.0

1.5.0

What's Changed

• Increment Version to 1.5.0 and add new guidelines to MAINTAINERS_GUIDE.md and CONTRIBUTING.md by @IanHoang in #494
• Revise guidelines based on comments by @IanHoang in #496
• Always print test_execution_id by @cgchinmay in #492
• Support request-params for bulk operation by @chishui in #501
• Allow users to specify percentiles for compare subcommand (#503) by @IanHoang in #504
• Add request start and end parameter to sleep runner by @VijayanB in #491
• Loosen the dependency version restriction (#499) by @chishui in #509
• Increase 'mapping.total_fields.limit' to '2000' by @rishabh6788 in #510
• Added runners to register and deploy ml-model by @vpehkone in #497
• Update force merge polling by @VijayanB in #489

New Contributors

• @chishui made their first contribution in #501
• @vpehkone made their first contribution in #497

Full Changelog: 1.4.0...1.5.0