diff --git a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/HybridSearchIT.java b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/HybridSearchIT.java
index a00c54abd..f8e85778e 100644
--- a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/HybridSearchIT.java
+++ b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/HybridSearchIT.java
@@ -115,7 +115,7 @@ private void validateTestIndexOnUpgrade(
         );
         assertNotNull(searchResponseAsMap);
         int hits = getHitCount(searchResponseAsMap);
-        assertEquals(1, hits);
+        assertEquals(2, hits);
         List<Double> scoresList = getNormalizationScoreList(searchResponseAsMap);
         for (Double score : scoresList) {
             assertTrue(0 <= score && score <= 2);
diff --git a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/MultiModalSearchIT.java b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/MultiModalSearchIT.java
index 246695d69..9d7536197 100644
--- a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/MultiModalSearchIT.java
+++ b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/MultiModalSearchIT.java
@@ -59,7 +59,7 @@ public void testTextImageEmbeddingProcessor_E2EFlow() throws Exception {
             case UPGRADED:
                 try {
                     modelId = getModelId(getIngestionPipeline(PIPELINE_NAME), TEXT_IMAGE_EMBEDDING_PROCESSOR);
-                    int totalDocsCountUpgraded = 1 * NUM_DOCS_PER_ROUND;
+                    int totalDocsCountUpgraded = 2 * NUM_DOCS_PER_ROUND;
                     loadModel(modelId);
                     addDocument(getIndexNameForTest(), "1", TEST_FIELD, TEXT_UPGRADED, TEST_IMAGE_FIELD, TEST_IMAGE_TEXT_UPGRADED);
                     validateTestIndexOnUpgrade(totalDocsCountUpgraded, modelId, TEXT_UPGRADED, TEST_IMAGE_TEXT_UPGRADED);
diff --git a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/NeuralSparseSearchIT.java b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/NeuralSparseSearchIT.java
index aa2ecffb7..1c3c6365d 100644
--- a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/NeuralSparseSearchIT.java
+++ b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/NeuralSparseSearchIT.java
@@ -80,7 +80,7 @@ public void testSparseEncodingProcessor_E2EFlow() throws Exception {
             case UPGRADED:
                 try {
                     modelId = getModelId(getIngestionPipeline(PIPELINE_NAME), SPARSE_ENCODING_PROCESSOR);
-                    int totalDocsCountUpgraded = 1 * NUM_DOCS_PER_ROUND;
+                    int totalDocsCountUpgraded = 2 * NUM_DOCS_PER_ROUND;
                     loadModel(modelId);
                     addSparseEncodingDoc(
                         getIndexNameForTest(),
diff --git a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/SemanticSearchIT.java b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/SemanticSearchIT.java
index c8e1a9e0b..6c9c07148 100644
--- a/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/SemanticSearchIT.java
+++ b/qa/rolling-upgrade/src/test/java/org/opensearch/neuralsearch/bwc/SemanticSearchIT.java
@@ -53,7 +53,7 @@ public void testSemanticSearch_E2EFlow() throws Exception {
                 String modelId3 = null;
                 try {
                     modelId3 = getModelId(getIngestionPipeline(PIPELINE_NAME), TEXT_EMBEDDING_PROCESSOR);
-                    int totalDocsCountUpgraded = 1 * NUM_DOCS_PER_ROUND;
+                    int totalDocsCountUpgraded = 2 * NUM_DOCS_PER_ROUND;
                     loadModel(modelId3);
                     addDocument(getIndexNameForTest(), "1", TEST_FIELD, TEXT_UPGRADED, null, null);
                     validateTestIndexOnUpgrade(totalDocsCountUpgraded, modelId3, TEXT_UPGRADED);