From 1fb424950de6d3e81ea3859c94bf0305595dbc26 Mon Sep 17 00:00:00 2001
From: minghong <englefly@gmail.com>
Date: Mon, 14 Oct 2024 15:54:49 +0800
Subject: [PATCH] if ndv is invalid, disable join reorder

---
 .../doris/nereids/stats/StatsCalculator.java  | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/fe/fe-core/src/main/java/org/apache/doris/nereids/stats/StatsCalculator.java b/fe/fe-core/src/main/java/org/apache/doris/nereids/stats/StatsCalculator.java
index 3c70d4cd51868fe..633d0cc8542b0dc 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/nereids/stats/StatsCalculator.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/nereids/stats/StatsCalculator.java
@@ -219,7 +219,9 @@ public static void disableJoinReorderIfTableRowCountNotAvailable(
         StatsCalculator calculator = new StatsCalculator(context);
         for (LogicalOlapScan scan : scans) {
             double rowCount = calculator.getOlapTableRowCount(scan);
-            if (rowCount == -1 && ConnectContext.get() != null) {
+            // analyzed rowCount may be zero, but BE-reported rowCount could be positive.
+            // check ndv validation when reported rowCount > 0
+            if ((rowCount == -1 || !calculator.checkNdvValidation(scan, rowCount)) && ConnectContext.get() != null) {
                 try {
                     ConnectContext.get().getSessionVariable().disableNereidsJoinReorderOnce();
                     LOG.info("disable join reorder since row count not available: "
@@ -403,6 +405,21 @@ private double getOlapTableRowCount(OlapScan olapScan) {
         return rowCount;
     }
 
+    // check validation of ndv.
+    private boolean checkNdvValidation(OlapScan olapScan, double rowCount) {
+        for (Slot slot : ((Plan) olapScan).getOutput()) {
+            if (isVisibleSlotReference(slot)) {
+                ColumnStatistic cache = getColumnStatsFromTableCache((CatalogRelation) olapScan, (SlotReference) slot);
+                if (!cache.isUnKnown
+                        && cache.ndv == 0
+                        && (cache.minExpr != null || cache.maxExpr != null)) {
+                    return false;
+                }
+            }
+        }
+        return true;
+    }
+
     private Statistics computeOlapScan(OlapScan olapScan) {
         OlapTable olapTable = olapScan.getTable();
         double tableRowCount = getOlapTableRowCount(olapScan);