From 79b4b4bec151224c2ddc4d5f094100911e348294 Mon Sep 17 00:00:00 2001
From: factosea <285808407@qq.com>
Date: Tue, 24 Dec 2024 09:53:11 +0800
Subject: [PATCH 1/5] fix: use limit with join

---
 source/libs/executor/src/mergejoin.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/source/libs/executor/src/mergejoin.c b/source/libs/executor/src/mergejoin.c
index 61809b99e0..adf1b4f0d1 100755
--- a/source/libs/executor/src/mergejoin.c
+++ b/source/libs/executor/src/mergejoin.c
@@ -46,7 +46,7 @@ static FORCE_INLINE bool mJoinBlkReachThreshold(SMJoinOperatorInfo* pInfo, int64
     return blkRows >= pInfo->ctx.mergeCtx.blkThreshold;
   }
   
-  return (pInfo->execInfo.resRows + blkRows) >= pInfo->ctx.mergeCtx.limit;
+  return (pInfo->execInfo.resRows + blkRows) >= pInfo->ctx.mergeCtx.limit || blkRows >= pInfo->ctx.mergeCtx.blkThreshold;
 }
 
 

From 4779f4a16fcb12c0af6a20363027152b69e367cf Mon Sep 17 00:00:00 2001
From: factosea <285808407@qq.com>
Date: Tue, 24 Dec 2024 17:22:39 +0800
Subject: [PATCH 2/5] test case

---
 tests/parallel_test/cases.task          |  1 +
 tests/pytest/util/cases.py              | 38 ++++++++++++++++++++
 tests/system-test/2-query/large_data.py | 46 +++++++++++++++++++++++++
 3 files changed, 85 insertions(+)
 create mode 100644 tests/system-test/2-query/large_data.py

diff --git a/tests/parallel_test/cases.task b/tests/parallel_test/cases.task
index c9d28e0623..9e4e6f9c71 100644
--- a/tests/parallel_test/cases.task
+++ b/tests/parallel_test/cases.task
@@ -247,6 +247,7 @@
 ,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/agg_group_NotReturnValue.py -Q 4
 
 ,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/td-32548.py
+,,n,system-test,python3 ./test.py -f 2-query/large_data.py
 
 ,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/stddev_test.py
 ,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/stddev_test.py -Q 2
diff --git a/tests/pytest/util/cases.py b/tests/pytest/util/cases.py
index 536b8f30d3..8d31cda78c 100644
--- a/tests/pytest/util/cases.py
+++ b/tests/pytest/util/cases.py
@@ -19,6 +19,7 @@ import inspect
 import importlib
 import traceback
 from util.log import *
+import platform
 
 
 class TDCase:
@@ -146,5 +147,42 @@ class TDCases:
 
         tdLog.notice("total %d Cluster test case(s) executed" % (runNum))
 
+    def getTaosBenchmarkPath(self, tool="taosBenchmark"):
+        if (platform.system().lower() == 'windows'):
+            tool = tool + ".exe"
+        selfPath = os.path.dirname(os.path.realpath(__file__))
 
+        if "community" in selfPath:
+            projPath = selfPath[: selfPath.find("community")]
+        else:
+            projPath = selfPath[: selfPath.find("tests")]
+
+        paths = []
+        for root, dirs, files in os.walk(projPath):
+            if (tool) in files:
+                rootRealPath = os.path.dirname(os.path.realpath(root))
+                if "packaging" not in rootRealPath:
+                    paths.append(os.path.join(root, tool))
+                    break
+        if len(paths) == 0:
+            tdLog.exit("taosBenchmark not found!")
+            return
+        else:
+            tdLog.info("taosBenchmark found in %s" % paths[0])
+            return paths[0]
+        
+    def taosBenchmarkExec(self, param):
+        buildPath = tdCases.getTaosBenchmarkPath()
+
+        if (platform.system().lower() == 'windows'):
+            cmdStr1 = ' mintty -h never %s %s '%(buildPath, param)
+            tdLog.info(cmdStr1)
+            os.system(cmdStr1)
+        else:
+            cmdStr1 = '%s %s &'%(buildPath, param)
+            tdLog.info(cmdStr1)
+            os.system(cmdStr1)
+            
+        time.sleep(15)
+       
 tdCases = TDCases()
diff --git a/tests/system-test/2-query/large_data.py b/tests/system-test/2-query/large_data.py
new file mode 100644
index 0000000000..6a5e3ca53c
--- /dev/null
+++ b/tests/system-test/2-query/large_data.py
@@ -0,0 +1,46 @@
+from util.log import *
+from util.sql import *
+from util.cases import *
+from util.dnodes import *
+
+class TDTestCase:
+
+    def init(self, conn, logSql, replicaVar=1):
+        self.replicaVar = int(replicaVar)
+        tdLog.debug(f"start to excute {__file__}")
+        tdSql.init(conn.cursor(), True)
+        
+    def prepare_data(self):
+        tdSql.execute("drop database if exists test;")
+        
+        tdCases.taosBenchmarkExec("-t 2 -n 1000000 -b int,float,nchar -y")
+
+    def ts5803(self):           
+        tdSql.query("select d0.ts,d0.c1,d0.c2 from test.d0 join test.d1 on d0.ts=d1.ts;")
+        num1 = tdSql.queryRows
+        
+        tdSql.query("select d0.ts,d0.c1,d0.c2 from test.d0 join test.d1 on d0.ts=d1.ts limit 1000000;")
+        tdSql.checkRows(num1)
+        
+        tdSql.query("select d0.ts from test.d0 join test.d1 on d0.ts=d1.ts limit 1000000;")
+        tdSql.checkRows(num1)
+        
+        tdSql.query("select d0.ts,d0.c1,d0.c2 from test.d0 left join test.d1 on d0.ts=d1.ts;")
+        num1 = tdSql.queryRows
+        
+        tdSql.query("select d0.ts,d0.c1,d0.c2 from test.d0 left join test.d1 on d0.ts=d1.ts limit 1000000;")
+        tdSql.checkRows(num1)
+        
+        tdSql.query("select d0.ts from test.d0 left join test.d1 on d0.ts=d1.ts limit 1000000;")
+        tdSql.checkRows(num1)
+
+    def run(self):
+        self.prepare_data()
+        self.ts5803()
+
+    def stop(self):
+        tdSql.close()
+        tdLog.success(f"{__file__} successfully executed")
+
+tdCases.addLinux(__file__, TDTestCase())
+tdCases.addWindows(__file__, TDTestCase())

From faaaa243afb4081e79c9591b781e467b48b9df4b Mon Sep 17 00:00:00 2001
From: factosea <285808407@qq.com>
Date: Wed, 25 Dec 2024 09:09:27 +0800
Subject: [PATCH 3/5] fix: test case

---
 tests/system-test/2-query/large_data.py | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/tests/system-test/2-query/large_data.py b/tests/system-test/2-query/large_data.py
index 6a5e3ca53c..69b15ce821 100644
--- a/tests/system-test/2-query/large_data.py
+++ b/tests/system-test/2-query/large_data.py
@@ -14,6 +14,15 @@ class TDTestCase:
         tdSql.execute("drop database if exists test;")
         
         tdCases.taosBenchmarkExec("-t 2 -n 1000000 -b int,float,nchar -y")
+        
+        while True:
+            tdSql.query("select d0.ts from test.d0;")
+            num1 = tdSql.queryRows
+            tdSql.query("select d0.ts from test.d1;")
+            num2 = tdSql.queryRows
+            if num1 == 1000000 and num2 == 1000000:
+                break
+            time.sleep(1)
 
     def ts5803(self):           
         tdSql.query("select d0.ts,d0.c1,d0.c2 from test.d0 join test.d1 on d0.ts=d1.ts;")

From f4180ec3d0c0ba166cab9daaeba6263b3946c0cd Mon Sep 17 00:00:00 2001
From: factosea <285808407@qq.com>
Date: Wed, 25 Dec 2024 09:11:54 +0800
Subject: [PATCH 4/5] fix: win system error desc

---
 source/util/src/terror.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/source/util/src/terror.c b/source/util/src/terror.c
index d2d551a539..b2a8c422f7 100644
--- a/source/util/src/terror.c
+++ b/source/util/src/terror.c
@@ -908,12 +908,6 @@ const char* tstrerror(int32_t err) {
   (void)taosThreadOnce(&tsErrorInit, tsSortError);
 
   // this is a system errno
-  if ((err & 0x00ff0000) == 0x00ff0000) {
-    int32_t code = err & 0x0000ffff;
-    // strerror can handle any invalid code
-    // invalid code return Unknown error
-    return strerror(code);
-  }
   #ifdef WINDOWS
   if ((err & 0x01ff0000) == 0x01ff0000) {
     snprintf(WinAPIErrDesc, 256, "windows api error, code: 0x%08x", err & 0x0000ffff);
@@ -923,6 +917,13 @@ const char* tstrerror(int32_t err) {
     return WinAPIErrDesc;
   }
   #endif
+  if ((err & 0x00ff0000) == 0x00ff0000) {
+    int32_t code = err & 0x0000ffff;
+    // strerror can handle any invalid code
+    // invalid code return Unknown error
+    return strerror(code);
+  }
+
   int32_t s = 0;
   int32_t e = sizeof(errors) / sizeof(errors[0]);
 

From 4a08d6915909a0f264b685bf1f705366d080d659 Mon Sep 17 00:00:00 2001
From: factosea <285808407@qq.com>
Date: Wed, 25 Dec 2024 09:18:57 +0800
Subject: [PATCH 5/5] fix: test case

---
 tests/pytest/util/cases.py              | 2 +-
 tests/system-test/2-query/large_data.py | 5 +++--
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/tests/pytest/util/cases.py b/tests/pytest/util/cases.py
index 8d31cda78c..eee8809ad0 100644
--- a/tests/pytest/util/cases.py
+++ b/tests/pytest/util/cases.py
@@ -183,6 +183,6 @@ class TDCases:
             tdLog.info(cmdStr1)
             os.system(cmdStr1)
             
-        time.sleep(15)
+        time.sleep(5)
        
 tdCases = TDCases()
diff --git a/tests/system-test/2-query/large_data.py b/tests/system-test/2-query/large_data.py
index 69b15ce821..2279c3ce70 100644
--- a/tests/system-test/2-query/large_data.py
+++ b/tests/system-test/2-query/large_data.py
@@ -16,12 +16,13 @@ class TDTestCase:
         tdCases.taosBenchmarkExec("-t 2 -n 1000000 -b int,float,nchar -y")
         
         while True:
-            tdSql.query("select d0.ts from test.d0;")
+            tdSql.query("select ts from test.d0;")
             num1 = tdSql.queryRows
-            tdSql.query("select d0.ts from test.d1;")
+            tdSql.query("select ts from test.d1;")
             num2 = tdSql.queryRows
             if num1 == 1000000 and num2 == 1000000:
                 break
+            tdLog.info(f"waiting for data ready, d0: {num1}, d1: {num2}")
             time.sleep(1)
 
     def ts5803(self):