From 984b47d9b12b897f14c617a2020824e5e63de914 Mon Sep 17 00:00:00 2001
From: "github-actions[bot]"
 <41898282+github-actions[bot]@users.noreply.github.com>
Date: Mon, 18 Nov 2024 20:34:39 +0800
Subject: [PATCH] branch-3.0: [fix](external) fix count(*) performance fallback
 in external table query #44172 (#44193)

Cherry-picked from #44172

Co-authored-by: Mingyu Chen (Rayner) <morningman@163.com>
---
 .../org/apache/doris/datasource/FileQueryScanNode.java     | 3 +++
 .../main/java/org/apache/doris/planner/OlapScanNode.java   | 7 -------
 .../src/main/java/org/apache/doris/planner/ScanNode.java   | 7 ++++++-
 3 files changed, 9 insertions(+), 8 deletions(-)

diff --git a/fe/fe-core/src/main/java/org/apache/doris/datasource/FileQueryScanNode.java b/fe/fe-core/src/main/java/org/apache/doris/datasource/FileQueryScanNode.java
index e75e2d144eff2e..b1cd799d5c4168 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/datasource/FileQueryScanNode.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/datasource/FileQueryScanNode.java
@@ -297,6 +297,7 @@ public void createScanRangeLocations() throws UserException {
                 location.setServer(new TNetworkAddress(backend.getHost(), backend.getBePort()));
                 curLocations.addToLocations(location);
                 scanRangeLocations.add(curLocations);
+                scanBackendIds.add(backendId);
                 return;
             }
         }
@@ -346,6 +347,7 @@ public void createScanRangeLocations() throws UserException {
                 // However, even one ScanNode instance can provide maximum scanning concurrency.
                 scanRangeLocations.add(curLocations);
                 setLocationPropertiesIfNecessary(backend, locationType, locationProperties);
+                scanBackendIds.add(backend.getId());
             }
         } else {
             List<Split> inputSplits = getSplits();
@@ -363,6 +365,7 @@ public void createScanRangeLocations() throws UserException {
                     scanRangeLocations.add(splitToScanRange(backend, locationProperties, split, pathPartitionKeys));
                     totalFileSize += split.getLength();
                 }
+                scanBackendIds.add(backend.getId());
             }
         }
 
diff --git a/fe/fe-core/src/main/java/org/apache/doris/planner/OlapScanNode.java b/fe/fe-core/src/main/java/org/apache/doris/planner/OlapScanNode.java
index afdffc748c01a6..41c055062e3e9b 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/planner/OlapScanNode.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/planner/OlapScanNode.java
@@ -192,8 +192,6 @@ public class OlapScanNode extends ScanNode {
     private Set<Long> sampleTabletIds = Sets.newHashSet();
     private TableSample tableSample;
 
-    private HashSet<Long> scanBackendIds = new HashSet<>();
-
     private Map<Long, Integer> tabletId2BucketSeq = Maps.newHashMap();
     // a bucket seq may map to many tablets, and each tablet has a
     // TScanRangeLocations.
@@ -1919,9 +1917,4 @@ public boolean pushDownAggNoGroupingCheckCol(FunctionCallExpr aggExpr, Column co
     public int getScanRangeNum() {
         return getScanTabletIds().size();
     }
-
-    @Override
-    public int numScanBackends() {
-        return scanBackendIds.size();
-    }
 }
diff --git a/fe/fe-core/src/main/java/org/apache/doris/planner/ScanNode.java b/fe/fe-core/src/main/java/org/apache/doris/planner/ScanNode.java
index b392075be9b938..7a36fbb3b047aa 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/planner/ScanNode.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/planner/ScanNode.java
@@ -110,6 +110,11 @@ public abstract class ScanNode extends PlanNode implements SplitGenerator {
 
     protected TableSnapshot tableSnapshot;
 
+    // Save the id of backends which this scan node will be executed on.
+    // This is also important for local shuffle logic.
+    // Now only OlapScanNode and FileQueryScanNode implement this.
+    protected HashSet<Long> scanBackendIds = new HashSet<>();
+
     public ScanNode(PlanNodeId id, TupleDescriptor desc, String planNodeName, StatisticalType statisticalType) {
         super(id, desc.getId().asList(), planNodeName, statisticalType);
         this.desc = desc;
@@ -731,7 +736,7 @@ && getScanRangeNum()
     }
 
     public int numScanBackends() {
-        return 0;
+        return scanBackendIds.size();
     }
 
     public int getScanRangeNum() {