Add new parameter to vector tracks to set vector index type (elastic#537

) * Add new parameter to vector tracks to set vector index type * Update README.md * Update index-vectors-only-mapping.json * Update index-vectors-with-text-mapping.json * Filter new parameter though tojson --------- Co-authored-by: Jason Bryan <[email protected]>
salvatore-campagna · Jan 25, 2024 · 8f081e2 · 8f081e2
1 parent b812ad8
commit 8f081e2
Show file tree

Hide file tree

Showing 10 changed files with 25 additions and 6 deletions.
diff --git a/cohere_vector/README.md b/cohere_vector/README.md
@@ -55,3 +55,4 @@ This track accepts the following parameters with Rally 0.8.0+ using `--track-par
  - post_ingest_sleep (default: false): Whether to pause after ingest and prior to subsequent operations.
  - post_ingest_sleep_duration (default: 30): Sleep duration in seconds.
  - standalone_search_iterations (default: 10000)
+ - vector_index_type (default: "hnsw"): The index kind for storing the vectors.
diff --git a/cohere_vector/index-vectors-only-mapping.json b/cohere_vector/index-vectors-only-mapping.json
@@ -19,7 +19,10 @@
         "element_type": "float",
         "dims": 768,
         "index": true,
-        "similarity": "max_inner_product"
+        "similarity": "max_inner_product",
+        "index_options": {
+          "type": {{ vector_index_type | default("hnsw") | tojson }}
+        }
       }
     }
   }

diff --git a/cohere_vector/index-vectors-with-text-mapping.json b/cohere_vector/index-vectors-with-text-mapping.json
@@ -24,7 +24,10 @@
         "element_type": "float",
         "dims": 768,
         "index": true,
-        "similarity": "max_inner_product"
+        "similarity": "max_inner_product",
+        "index_options": {
+          "type": {{ vector_index_type | default("hnsw") |tojson }}
+        }
       }
     }
   }

diff --git a/dense_vector/README.md b/dense_vector/README.md
@@ -33,3 +33,4 @@ This track accepts the following parameters with Rally 0.8.0+ using `--track-par
 * `number_of_replicas` (default: 0)
 * `post_ingest_sleep` (default: false): Whether to pause after ingest and prior to subsequent operations.
 * `post_ingest_sleep_duration` (default: 30): Sleep duration in seconds.
+* `vector_index_type` (default: "hnsw"): The index kind for storing the vectors.
diff --git a/dense_vector/index.json b/dense_vector/index.json
@@ -19,7 +19,7 @@
         "index" : true,
         "similarity": "dot_product",
         "index_options": {
-          "type": "hnsw",
+          "type": {{ vector_index_type | default("hnsw") | tojson }},
           "m": 32,
           "ef_construction": 100
         }

diff --git a/openai_vector/README.md b/openai_vector/README.md
@@ -50,6 +50,7 @@ This track accepts the following parameters with Rally 0.8.0+ using `--track-par
 - post_ingest_sleep_duration (default: 30): Sleep duration in seconds.
 - standalone_search_clients (default: 8)
 - standalone_search_iterations (default: 10000)
+- vector_index_type (default: "hnsw"): The index kind for storing the vectors.
 
 ### License
 

diff --git a/openai_vector/index-vectors-only-mapping.json b/openai_vector/index-vectors-only-mapping.json
@@ -19,7 +19,10 @@
         "element_type": "float",
         "dims": 1536,
         "index": true,
-        "similarity": "dot_product"
+        "similarity": "dot_product",
+        "index_options": {
+          "type": {{ vector_index_type | default("hnsw") | tojson }}
+        }
       }
     }
   }

diff --git a/openai_vector/index-vectors-with-text-mapping.json b/openai_vector/index-vectors-with-text-mapping.json
@@ -24,7 +24,10 @@
         "element_type": "float",
         "dims": 1536,
         "index": true,
-        "similarity": "dot_product"
+        "similarity": "dot_product",
+        "index_options": {
+          "type": {{ vector_index_type | default("hnsw") | tojson }}
+        }
       }
     }
   }

diff --git a/so_vector/README.md b/so_vector/README.md
@@ -68,6 +68,7 @@ This track accepts the following parameters with Rally 0.8.0+ using `--track-par
 * `max_num_segments` (default: 1)
 * `include_non_serverless_index_settings` (default: true for non-serverless clusters, false for serverless clusters): Whether to include non-serverless index settings.
 * `include_force_merge` (default: true for non-serverless clusters, false for serverless clusters): Whether to include force merge operation.
+* `vector_index_type` (default: "hnsw"): The index kind for storing the vectors.
 
 ### License
 We use the same license for the data as the original data: [CC-SA-4.0](http://creativecommons.org/licenses/by-sa/4.0/).

diff --git a/so_vector/index.json b/so_vector/index.json
@@ -33,7 +33,10 @@
         "type": "dense_vector",
         "dims" : 768,
         "index" : true,
-        "similarity": "dot_product"
+        "similarity": "dot_product",
+        "index_options": {
+          "type": {{ vector_index_type | default("hnsw") | tojson }}
+        }
       },
       "acceptedAnswerId": {
         "type": "keyword"