Add new benchmark config for query approximation (#16323)

We have guarded the experimental query approximation framework behind
a feature flag. In order to easily measure the impact of approximation
on big5 benchmarks, it would be nice to have a benchmark config.

Signed-off-by: Michael Froh <[email protected]>

.github/benchmark-configs.json

@@ -221,5 +221,23 @@
       "data_instance_config": "4vCPU, 32G Mem, 16G Heap"
     },
     "baseline_cluster_config": "x64-r5.xlarge-1-shard-0-replica-snapshot-baseline"
+  },
+  "id_14": {
+    "description": "Search only test-procedure for big5, uses snapshot to restore the data for OS-3.0.0. Enables range query approximation.",
+    "supported_major_versions": ["3"],
+    "cluster-benchmark-configs": {
+      "SINGLE_NODE_CLUSTER": "true",
+      "MIN_DISTRIBUTION": "true",
+      "TEST_WORKLOAD": "big5",
+      "ADDITIONAL_CONFIG": "opensearch.experimental.feature.approximate_point_range_query.enabled:true",
+      "WORKLOAD_PARAMS": "{\"snapshot_repo_name\":\"benchmark-workloads-repo-300\",\"snapshot_bucket_name\":\"benchmark-workload-snapshots\",\"snapshot_region\":\"us-east-1\",\"snapshot_base_path\":\"workload-snapshots-300\",\"snapshot_name\":\"big5_1_shard_ordered\"}",
+      "CAPTURE_NODE_STAT": "true",
+      "TEST_PROCEDURE": "restore-from-snapshot"
+    },
+    "cluster_configuration": {
+      "size": "Single-Node",
+      "data_instance_config": "4vCPU, 32G Mem, 16G Heap"
+    },
+    "baseline_cluster_config": "x64-r5.xlarge-1-shard-0-replica-snapshot-baseline"
   }
 }