From 7ff72599b44d7993c3c68b0a670e2bc36a2ca35c Mon Sep 17 00:00:00 2001 From: Ryan Bogan Date: Fri, 22 Nov 2024 16:25:14 -0800 Subject: [PATCH 1/4] Add clarification footnote to exact search k-NN blog post Signed-off-by: Ryan Bogan --- _posts/2024-11-13-boosting-k-nn-exact-search.md | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/_posts/2024-11-13-boosting-k-nn-exact-search.md b/_posts/2024-11-13-boosting-k-nn-exact-search.md index 81a60d439..db6a7aa7f 100644 --- a/_posts/2024-11-13-boosting-k-nn-exact-search.md +++ b/_posts/2024-11-13-boosting-k-nn-exact-search.md @@ -110,3 +110,8 @@ The tests showed that OpenSearch's new SIMD support and optimized memory access ## What's next for exact k-NN search? Future OpenSearch versions will provide even more k-NN search flexibility. You'll be able to switch between exact and approximate search at query time. Additionally, future versions will provide the ability to specify which fields build indexes for exact and approximate search types. Stay tuned for these updates as we continue to improve OpenSearch's k-NN search capabilities. + + +#### Note: + +It is required to enable the setting `index.knn: true` to use the optimized Lucene format. With this setting in OpenSearch 2.17, the approximate nearest neighbors data structures were also built. In OpenSearch 2.18, we introduced a new setting `index.knn.advanced.approximate_threshold`. If you are only performing exact search, set this to either `-1` or `Integer.MAX - 2` in order to speed up indexing time. \ No newline at end of file From b2515e82fbe216460e2b2a07eabed49d7c3f1665 Mon Sep 17 00:00:00 2001 From: kolchfa-aws <105444904+kolchfa-aws@users.noreply.github.com> Date: Mon, 25 Nov 2024 12:01:27 -0500 Subject: [PATCH 2/4] Update _posts/2024-11-13-boosting-k-nn-exact-search.md Signed-off-by: kolchfa-aws <105444904+kolchfa-aws@users.noreply.github.com> --- _posts/2024-11-13-boosting-k-nn-exact-search.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/_posts/2024-11-13-boosting-k-nn-exact-search.md b/_posts/2024-11-13-boosting-k-nn-exact-search.md index db6a7aa7f..e1986324e 100644 --- a/_posts/2024-11-13-boosting-k-nn-exact-search.md +++ b/_posts/2024-11-13-boosting-k-nn-exact-search.md @@ -114,4 +114,4 @@ Future OpenSearch versions will provide even more k-NN search flexibility. You'l #### Note: -It is required to enable the setting `index.knn: true` to use the optimized Lucene format. With this setting in OpenSearch 2.17, the approximate nearest neighbors data structures were also built. In OpenSearch 2.18, we introduced a new setting `index.knn.advanced.approximate_threshold`. If you are only performing exact search, set this to either `-1` or `Integer.MAX - 2` in order to speed up indexing time. \ No newline at end of file +To use the optimized Lucene format in OpenSearch 2.17, set `index.knn` to `true` in order to build approximate nearest neighbor (ANN) data structures. In OpenSearch 2.18, a new `index.knn.advanced.approximate_threshold` setting is available. If you are performing only exact searches, set this value to `-1` or `Integer.MAX_VALUE - 2` to reduce indexing time. \ No newline at end of file From 0c08d31f0eafb4de6f4142bf20c6217f3e909f2b Mon Sep 17 00:00:00 2001 From: Ryan Bogan Date: Mon, 25 Nov 2024 09:10:03 -0800 Subject: [PATCH 3/4] Simplify recommended setting Signed-off-by: Ryan Bogan --- _posts/2024-11-13-boosting-k-nn-exact-search.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/_posts/2024-11-13-boosting-k-nn-exact-search.md b/_posts/2024-11-13-boosting-k-nn-exact-search.md index e1986324e..28d13ff17 100644 --- a/_posts/2024-11-13-boosting-k-nn-exact-search.md +++ b/_posts/2024-11-13-boosting-k-nn-exact-search.md @@ -114,4 +114,4 @@ Future OpenSearch versions will provide even more k-NN search flexibility. You'l #### Note: -To use the optimized Lucene format in OpenSearch 2.17, set `index.knn` to `true` in order to build approximate nearest neighbor (ANN) data structures. In OpenSearch 2.18, a new `index.knn.advanced.approximate_threshold` setting is available. If you are performing only exact searches, set this value to `-1` or `Integer.MAX_VALUE - 2` to reduce indexing time. \ No newline at end of file +To use the optimized Lucene format in OpenSearch 2.17, set `index.knn` to `true` in order to build approximate nearest neighbor (ANN) data structures. In OpenSearch 2.18, a new `index.knn.advanced.approximate_threshold` setting is available. If you are performing only exact searches, set this value to `-1` to reduce indexing time. \ No newline at end of file From d49d2bd8d68ce2a724f0a7c9278fbe4bbdb04b16 Mon Sep 17 00:00:00 2001 From: kolchfa-aws <105444904+kolchfa-aws@users.noreply.github.com> Date: Mon, 25 Nov 2024 12:28:36 -0500 Subject: [PATCH 4/4] Update _posts/2024-11-13-boosting-k-nn-exact-search.md Signed-off-by: kolchfa-aws <105444904+kolchfa-aws@users.noreply.github.com> --- _posts/2024-11-13-boosting-k-nn-exact-search.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/_posts/2024-11-13-boosting-k-nn-exact-search.md b/_posts/2024-11-13-boosting-k-nn-exact-search.md index 28d13ff17..530c38e28 100644 --- a/_posts/2024-11-13-boosting-k-nn-exact-search.md +++ b/_posts/2024-11-13-boosting-k-nn-exact-search.md @@ -112,6 +112,6 @@ The tests showed that OpenSearch's new SIMD support and optimized memory access Future OpenSearch versions will provide even more k-NN search flexibility. You'll be able to switch between exact and approximate search at query time. Additionally, future versions will provide the ability to specify which fields build indexes for exact and approximate search types. Stay tuned for these updates as we continue to improve OpenSearch's k-NN search capabilities. -#### Note: +#### Note To use the optimized Lucene format in OpenSearch 2.17, set `index.knn` to `true` in order to build approximate nearest neighbor (ANN) data structures. In OpenSearch 2.18, a new `index.knn.advanced.approximate_threshold` setting is available. If you are performing only exact searches, set this value to `-1` to reduce indexing time. \ No newline at end of file