join order good for rf

englefly · Jul 30, 2024 · 88268d6 · 88268d6
1 parent 1b17f99
commit 88268d6
Showing 1 changed file with 33 additions and 0 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
@@ -21,6 +21,7 @@
 import org.apache.doris.catalog.KeysType;
 import org.apache.doris.catalog.OlapTable;
 import org.apache.doris.nereids.PlanContext;
+import org.apache.doris.nereids.processor.post.RuntimeFilterGenerator;
 import org.apache.doris.nereids.properties.DistributionSpec;
 import org.apache.doris.nereids.properties.DistributionSpecGather;
 import org.apache.doris.nereids.properties.DistributionSpecHash;
@@ -55,6 +56,7 @@
 import org.apache.doris.nereids.trees.plans.visitor.PlanVisitor;
 import org.apache.doris.qe.ConnectContext;
 import org.apache.doris.qe.SessionVariable;
+import org.apache.doris.statistics.ColumnStatistic;
 import org.apache.doris.statistics.Statistics;
 
 import com.google.common.base.Preconditions;
@@ -371,6 +373,20 @@ public Cost visitPhysicalHashJoin(
                 leftRowCount += 1e-3;
             }
         }
+        if (!physicalHashJoin.getGroupExpression().get().getOwnerGroup().isStatsReliable()
+                && !(RuntimeFilterGenerator.DENIED_JOIN_TYPES.contains(physicalHashJoin.getJoinType()))
+                && !physicalHashJoin.isMarkJoin()
+                && buildStats.getWidthInJoinCluster() == 1) {
+            // we prefer join order: A-B-filter(C) to A-filter(C)-B,
+            // since filter(C) may generate effective RF to B, and RF(C->B) makes RF(B->A) more effective.
+            double filterFactor = computeFilterFactor(buildStats);
+            if (filterFactor > 1.0) {
+                double bonus = filterFactor * probeStats.getWidthInJoinCluster();
+                if (leftRowCount > bonus) {
+                    leftRowCount -= bonus;
+                }
+            }
+        }
 
         /*
         pattern1: L join1 (Agg1() join2 Agg2())
@@ -428,6 +444,23 @@ public Cost visitPhysicalHashJoin(
         );
     }
 
+    private double computeFilterFactor(Statistics stats) {
+        double factor = 1.0;
+        double maxBaseTableRowCount = 0.0;
+        for (Expression expr : stats.columnStatistics().keySet()) {
+            ColumnStatistic colStats = stats.findColumnStatistics(expr);
+            if (colStats.isUnKnown) {
+                maxBaseTableRowCount = Math.max(maxBaseTableRowCount, colStats.count);
+            } else {
+                break;
+            }
+        }
+        if (maxBaseTableRowCount != 0) {
+            factor = Math.max(1, Math.min(2, maxBaseTableRowCount / stats.getRowCount()));
+        }
+        return factor;
+    }
+
     /*
     in a join cluster graph, if a node has higher connectivity, it is more likely to be reduced
     by runtime filters, and it is also more likely to produce effective runtime filters.