fix

englefly · Nov 24, 2023 · 5a723e9 · 5a723e9
1 parent 7829873
commit 5a723e9
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 10 deletions.
diff --git a/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java b/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java
@@ -194,7 +194,7 @@ public Cost visitPhysicalDistribute(
             return CostV1.of(context.getSessionVariable(),
                     0,
                     0,
-                    intputRowCount * childStatistics.dataSizeFactor());
+                    intputRowCount * childStatistics.dataSizeFactor() / beNumber);
         }
 
         // replicate
@@ -205,7 +205,7 @@ public Cost visitPhysicalDistribute(
             return CostV1.of(context.getSessionVariable(),
                     0,
                     0,
-                    intputRowCount * childStatistics.dataSizeFactor() * beNumber);
+                    intputRowCount * childStatistics.dataSizeFactor());
 
         }
 
@@ -214,15 +214,15 @@ public Cost visitPhysicalDistribute(
             return CostV1.of(context.getSessionVariable(),
                     0,
                     0,
-                    intputRowCount * childStatistics.dataSizeFactor());
+                    intputRowCount * childStatistics.dataSizeFactor() / beNumber);
         }
 
         // any
         // cost of random shuffle is lower than hash shuffle.
         return CostV1.of(context.getSessionVariable(),
                 0,
                 0,
-                intputRowCount * childStatistics.dataSizeFactor() * RANDOM_SHUFFLE_TO_HASH_SHUFFLE_FACTOR);
+                intputRowCount * childStatistics.dataSizeFactor() * RANDOM_SHUFFLE_TO_HASH_SHUFFLE_FACTOR / beNumber);
     }
 
     @Override
@@ -291,10 +291,13 @@ public Cost visitPhysicalHashJoin(
             int totalInstanceNumber = parallelInstance * beNumber;
             if (buildSideFactor <= 1.0) {
                 // use totalInstanceNumber to the power of 2 as the default factor value
-                buildSideFactor = Math.pow(totalInstanceNumber, 0.5);
+                if (buildStats.computeSize() > 1024 * 1024) {
+                    buildSideFactor = Math.pow(totalInstanceNumber, 0.5);
+                }
             }
+            probeSideFactor = broadCastJoinBalancePenalty(probeStats, buildStats);
             return CostV1.of(context.getSessionVariable(),
-                    leftRowCount + rightRowCount * buildSideFactor + outputRowCount * probeSideFactor,
+                    leftRowCount * probeSideFactor + rightRowCount * buildSideFactor + outputRowCount,
                     rightRowCount,
                     0
             );

diff --git a/fe/fe-core/src/main/java/org/apache/doris/qe/SessionVariable.java b/fe/fe-core/src/main/java/org/apache/doris/qe/SessionVariable.java
@@ -460,7 +460,7 @@ public class SessionVariable implements Serializable, Writable {
     public static final String SHUFFLE_FACTOR_LOWER_BOUND = "shuffle_factor_lower_bound";
 
     @VariableMgr.VarAttr(name = SHUFFLE_FACTOR_LOWER_BOUND)
-    public double shuffleFactorLowerBound = 0.01;
+    public double shuffleFactorLowerBound = 0.05;
 
     public static final String ENABLE_STATS = "enable_stats";
     /**

diff --git a/fe/fe-core/src/main/java/org/apache/doris/statistics/Statistics.java b/fe/fe-core/src/main/java/org/apache/doris/statistics/Statistics.java
@@ -130,12 +130,11 @@ public double computeSize() {
     }
 
     public double dataSizeFactor() {
-        double lowerBound = 0.01;
+        double lowerBound = 0.05;
         if (ConnectContext.get() != null) {
             lowerBound = ConnectContext.get().getSessionVariable().shuffleFactorLowerBound;
         }
-        return computeTupleSize() * lowerBound;//0.001
-        // return Math.max(computeTupleSize() / K_BYTES, lowerBound);
+        return Math.min(Math.max(computeTupleSize() / K_BYTES, lowerBound), 0.1);
     }
 
     @Override