Merge branch 'fix/TD023101' of https://github.com/taosdata/TDengine into fix/TD023101

source/dnode/vnode/src/vnd/vnodeSnapshot.c

@@ -335,6 +335,7 @@ int32_t vnodeSnapWriterClose(SVSnapWriter *pWriter, int8_t rollback, SSnapshot *
 
   // commit json
   if (!rollback) {
+    pWriter->info.state.committed = pWriter->ever;
     pVnode->config = pWriter->info.config;
     pVnode->state = (SVState){.committed = pWriter->info.state.committed,
                               .applied = pWriter->info.state.committed,

source/libs/executor/src/projectoperator.c

@@ -161,10 +161,9 @@ static int32_t discardGroupDataBlock(SSDataBlock* pBlock, SLimitInfo* pLimitInfo
       if (pLimitInfo->remainGroupOffset > 0) {
         return PROJECT_RETRIEVE_CONTINUE;
       }
-    }
 
-    // set current group id of the project operator
-    pLimitInfo->currentGroupId = pBlock->info.id.groupId;
+      pLimitInfo->currentGroupId = 0;
+    }
   }
 
   return PROJECT_RETRIEVE_DONE;
@@ -175,19 +174,29 @@ static int32_t setInfoForNewGroup(SSDataBlock* pBlock, SLimitInfo* pLimitInfo, S
   // here check for a new group data, we need to handle the data of the previous group.
   ASSERT(pLimitInfo->remainGroupOffset == 0 || pLimitInfo->remainGroupOffset == -1);
 
-  if (pLimitInfo->currentGroupId != 0 && pLimitInfo->currentGroupId != pBlock->info.id.groupId) {
+  bool newGroup = false;
+  if (0 == pBlock->info.id.groupId) {
+    pLimitInfo->numOfOutputGroups = 1;
+  } else if (pLimitInfo->currentGroupId != pBlock->info.id.groupId) {
+    pLimitInfo->currentGroupId = pBlock->info.id.groupId;
     pLimitInfo->numOfOutputGroups += 1;
-    if ((pLimitInfo->slimit.limit > 0) && (pLimitInfo->slimit.limit <= pLimitInfo->numOfOutputGroups)) {
+    newGroup = true;
+  } else {
+    return PROJECT_RETRIEVE_CONTINUE;
+  }
+
+  if ((pLimitInfo->slimit.limit >= 0) && (pLimitInfo->slimit.limit < pLimitInfo->numOfOutputGroups)) {
     setOperatorCompleted(pOperator);
     return PROJECT_RETRIEVE_DONE;
   }
 
   // reset the value for a new group data
   // existing rows that belongs to previous group.
+  if (newGroup) {
     resetLimitInfoForNextGroup(pLimitInfo);
   }
   
-  return PROJECT_RETRIEVE_DONE;
+  return PROJECT_RETRIEVE_CONTINUE;
 }
 
 // todo refactor
@@ -199,7 +208,7 @@ static int32_t doIngroupLimitOffset(SLimitInfo* pLimitInfo, uint64_t groupId, SS
   if (pBlock->info.rows == 0) {
     return PROJECT_RETRIEVE_CONTINUE;
   } else {
-    if (limitReached && (pLimitInfo->slimit.limit > 0 && pLimitInfo->slimit.limit <= pLimitInfo->numOfOutputGroups)) {
+    if (limitReached && (pLimitInfo->slimit.limit >= 0 && pLimitInfo->slimit.limit <= pLimitInfo->numOfOutputGroups)) {
       setOperatorCompleted(pOperator);
     }
   }

tests/script/tsim/parser/limit1_stb.sim

@@ -484,6 +484,7 @@ if $rows != 2 then
   return -1
 endi
 
+print === select max(c1), min(c2), avg(c3), sum(c5), spread(c6), first(c7), last(c8), first(c9) from $stb where ts >= $ts0 and ts <= $tsu and t1 > 1 and t1 < 5 and c1 > 0 and c2 < 9 and c3 > 1 and c4 < 7 and c5 > 4 partition by t1 interval(5m) limit 1 offset 0
 sql select max(c1), min(c2), avg(c3), sum(c5), spread(c6), first(c7), last(c8), first(c9) from $stb where ts >= $ts0 and ts <= $tsu and t1 > 1 and t1 < 5 and c1 > 0 and c2 < 9 and c3 > 1 and c4 < 7 and c5 > 4 partition by t1 interval(5m) limit 1 offset 0
 if $rows != 3 then
   return -1

tests/script/tsim/parser/slimit1_query.sim

@@ -70,7 +70,7 @@ endi
 
 ### empty result set
 sql select count(*) from stb partition by t2,t1 order by t2 asc slimit 0 soffset 0
-if $rows != 9 then
+if $rows != 0 then
   return -1
 endi
 

tests/system-test/7-tmq/subscribeStb1.py

@@ -245,8 +245,8 @@ class TDTestCase:
         for i in range(expectRows):
             totalConsumeRows += resultList[i]
 
-        if totalConsumeRows != expectrowcnt/4:
-            tdLog.info("act consume rows: %d, expect consume rows: %d"%(totalConsumeRows, expectrowcnt/4))
+        tdLog.info("act consume rows: %d, expect consume rows greater than or equal to: %d"%(totalConsumeRows, expectrowcnt/4))
+        if totalConsumeRows < expectrowcnt/4:            
             tdLog.exit("tmq consume rows error!")
 
         self.initConsumerInfoTable()
@@ -267,8 +267,8 @@ class TDTestCase:
         for i in range(expectRows):
             totalConsumeRows += resultList[i]
 
-        if totalConsumeRows != expectrowcnt:
-            tdLog.info("act consume rows: %d, expect consume rows: %d"%(totalConsumeRows, expectrowcnt))
+        tdLog.info("act consume rows: %d, expect consume rows greater than or equal to: %d"%(totalConsumeRows, expectrowcnt))
+        if totalConsumeRows < expectrowcnt:            
             tdLog.exit("tmq consume rows error!")
 
         tdSql.query("drop topic %s"%topicFromStb1)

tests/system-test/7-tmq/tmqDelete-1ctb.py

@@ -323,7 +323,7 @@ class TDTestCase:
 
         if self.snapshot == 0:
             consumerId     = 4
-            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1 + 1/4 + 3/4))
+            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1/4 + 3/4))
         elif self.snapshot == 1:
             consumerId     = 5
             expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1 - 1/4 + 1/4 + 3/4))
@@ -369,9 +369,14 @@ class TDTestCase:
         tdLog.info("act consume rows: %d, act query rows: %d, expect consume rows: %d, "%(totalConsumeRows, totalRowsFromQuery, expectrowcnt))
 
         if self.snapshot == 0:
-            if totalConsumeRows != expectrowcnt:
+            # If data writing is completed before consumer get snapshot, will consume 7500 from wal;
+            # If data writing has not started before consumer get snapshot, will consume 10000 from wal;
+            minRows = int(expectrowcnt * (1 - 1/4)) # 7500
+            tdLog.info("consume rows should be between %d and %d, "%(minRows, expectrowcnt))
+            if not ((totalConsumeRows >= minRows) and (totalConsumeRows <= expectrowcnt)):
                 tdLog.exit("tmq consume rows error with snapshot = 0!")
         elif self.snapshot == 1:
+            tdLog.info("consume rows should be between %d and %d, "%(totalRowsFromQuery, expectrowcnt))
             if not ((totalConsumeRows >= totalRowsFromQuery) and (totalConsumeRows <= expectrowcnt)):
                 tdLog.exit("tmq consume rows error with snapshot = 1!")
 
@@ -381,6 +386,112 @@ class TDTestCase:
 
         tdLog.printNoPrefix("======== test case 3 end ...... ")
 
+    def tmqCase4(self):
+        tdLog.printNoPrefix("======== test case 4: ")
+        paraDict = {'dbName':     'dbt',
+                    'dropFlag':   1,
+                    'event':      '',
+                    'vgroups':    4,
+                    'stbName':    'stb',
+                    'colPrefix':  'c',
+                    'tagPrefix':  't',
+                    'colSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1},{'type': 'TIMESTAMP', 'count':1}],
+                    'tagSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1}],
+                    'ctbPrefix':  'ctb',
+                    'ctbStartIdx': 0,
+                    'ctbNum':     1,
+                    'rowsPerTbl': 10000,
+                    'batchNum':   5000,
+                    'startTs':    1640966400000,  # 2022-01-01 00:00:00.000
+                    'pollDelay':  15,
+                    'showMsg':    1,
+                    'showRow':    1,
+                    'snapshot':   0}
+
+        paraDict['snapshot'] = self.snapshot
+        paraDict['vgroups'] = self.vgroups
+        paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+
+        tdLog.info("restart taosd to ensure that the data falls into the disk")
+        tdSql.query("flush database %s"%(paraDict['dbName']))
+
+        tmqCom.initConsumerTable()
+        tdLog.info("create topics from stb1")
+        topicFromStb1 = 'topic_stb1'
+        queryString = "select ts, c1, c2 from %s.%s"%(paraDict['dbName'], paraDict['stbName'])
+        sqlString = "create topic %s as %s" %(topicFromStb1, queryString)
+        tdLog.info("create topic sql: %s"%sqlString)
+        tdSql.execute(sqlString)
+
+        # paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+        consumerId     = 1
+
+        if self.snapshot == 0:
+            consumerId     = 4
+            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"])
+        elif self.snapshot == 1:
+            consumerId     = 5
+            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1 - 1/4 + 1/4 + 3/4))
+
+        topicList      = topicFromStb1
+        ifcheckdata    = 1
+        ifManualCommit = 1
+        keyList        = 'group.id:cgrp1,\
+                        enable.auto.commit:true,\
+                        auto.commit.interval.ms:1000,\
+                        auto.offset.reset:earliest'
+        tmqCom.insertConsumerInfo(consumerId, expectrowcnt,topicList,keyList,ifcheckdata,ifManualCommit)
+
+        # del some data
+        rowsOfDelete = int(self.rowsPerTbl / 4 )
+        paraDict["endTs"] = paraDict["startTs"] + rowsOfDelete - 1
+        # pDeleteThread = self.asyncDeleteData(paraDict)
+        self.threadFunctionForDeletaData(paraDict)
+
+        tdLog.info("start consume processor")
+        tmqCom.startTmqSimProcess(pollDelay=paraDict['pollDelay'],dbName=paraDict["dbName"],showMsg=paraDict['showMsg'], showRow=paraDict['showRow'],snapshot=paraDict['snapshot'])
+
+        tmqCom.getStartConsumeNotifyFromTmqsim()
+        
+        # update to 1/4 rows and insert 3/4 new rows
+        paraDict['startTs'] = paraDict['startTs'] + int(self.rowsPerTbl * 3 / 4)
+        # paraDict['rowsPerTbl'] = self.rowsPerTbl
+        # tmqCom.insert_data_with_autoCreateTbl(tsql=tdSql,dbName=paraDict["dbName"],stbName=paraDict["stbName"],ctbPrefix=paraDict["ctbPrefix"],
+        #                                       ctbNum=paraDict["ctbNum"],rowsPerTbl=paraDict["rowsPerTbl"],batchNum=paraDict["batchNum"],
+        #                                       startTs=paraDict["startTs"],ctbStartIdx=paraDict['ctbStartIdx'])
+        pInsertThread = tmqCom.asyncInsertDataByInterlace(paraDict)
+
+        pInsertThread.join()
+
+        tdLog.info("start to check consume result")
+        expectRows = 1
+        resultList = tmqCom.selectConsumeResult(expectRows)
+        totalConsumeRows = 0
+        for i in range(expectRows):
+            totalConsumeRows += resultList[i]
+
+        tdSql.query(queryString)
+        totalRowsFromQuery = tdSql.getRows()
+
+        tdLog.info("act consume rows: %d, act query rows: %d, expect consume rows: %d, "%(totalConsumeRows, totalRowsFromQuery, expectrowcnt))
+
+        if self.snapshot == 0:
+            tdLog.info("consume rows should be %d"%(expectrowcnt))
+            if (totalConsumeRows != expectrowcnt):
+                tdLog.exit("tmq consume rows error with snapshot = 0!")
+        elif self.snapshot == 1:
+            # If data writing has not started before consumer get snapshot, will consume 10000 from wal, and consumer 7500 from tsdb;
+            tdLog.info("consume rows should be %d, "%(expectrowcnt))
+            if (totalConsumeRows != expectrowcnt):
+                tdLog.exit("tmq consume rows error with snapshot = 1!")
+
+        # tmqCom.checkFileContent(consumerId, queryString)
+
+        tdSql.query("drop topic %s"%topicFromStb1)
+
+        tdLog.printNoPrefix("======== test case 4 end ...... ")
         
     def run(self):
         # tdSql.prepare()
@@ -409,6 +520,17 @@ class TDTestCase:
         self.prepareTestEnv()
         self.tmqCase3()
 
+        tdLog.printNoPrefix("=============================================")
+        tdLog.printNoPrefix("======== snapshot is 0: only consume from wal")
+        self.snapshot = 0
+        self.prepareTestEnv()
+        self.tmqCase4()
+        tdLog.printNoPrefix("====================================================================")
+        tdLog.printNoPrefix("======== snapshot is 1: firstly consume from tsbs, and then from wal")
+        self.snapshot = 1
+        self.prepareTestEnv()
+        self.tmqCase4()
+
     def stop(self):
         tdSql.close()
         tdLog.success(f"{__file__} successfully executed")

tests/system-test/7-tmq/tmqDnodeRestart.py

@@ -136,7 +136,7 @@ class TDTestCase:
         tdLog.info("================= restart dnode ===========================")
         tdDnodes.stoptaosd(1)
         tdDnodes.starttaosd(1)
-        # time.sleep(3)
+        time.sleep(5)
 
         tdLog.info(" restart taosd end and wait to check consume result")
         expectRows = 1
@@ -220,7 +220,7 @@ class TDTestCase:
         tdLog.info("================= restart dnode ===========================")
         tdDnodes.stoptaosd(1)
         tdDnodes.starttaosd(1)
-        # time.sleep(3)
+        time.sleep(5)
 
         tdLog.info("create some new child table and insert data ")
         paraDict["batchNum"] = 100

tests/system-test/7-tmq/tmqMultiConsumer.py

+
+import taos
+import sys
+import time
+import socket
+import os
+import threading
+import math
+# from tests.pytest.util.common import TDCom
+# from tests.pytest.util.log import TDLog
+
+from util.log import *
+from util.sql import *
+from util.cases import *
+from util.dnodes import *
+from util.common import *
+sys.path.append("./7-tmq")
+from tmqCommon import *
+
+class TDTestCase:
+    def __init__(self):
+        self.vgroups    = 32
+        self.ctbNum     = 100
+        self.rowsPerTbl = 1000
+        self.snapshot   = 1
+        self.replicaVar = 3
+
+    def init(self, conn, logSql, replicaVar=1):
+        self.replicaVar = int(replicaVar)
+        tdLog.debug(f"start to excute {__file__}")
+        tdSql.init(conn.cursor(), False)
+
+    def prepareTestEnv(self):
+        tdLog.printNoPrefix("======== prepare test env include database, stable, ctables, and insert data: ")
+        paraDict = {'dbName':     'dbt',
+                    'dropFlag':   1,
+                    'event':      '',
+                    'vgroups':    4,
+                    'stbName':    'stb',
+                    'colPrefix':  'c',
+                    'tagPrefix':  't',
+                    'colSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1},{'type': 'TIMESTAMP', 'count':1}],
+                    'tagSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1}],
+                    'ctbPrefix':  'ctb',
+                    'ctbStartIdx': 0,
+                    'ctbNum':     10,
+                    'rowsPerTbl': 10000,
+                    'batchNum':   1000,
+                    'startTs':    1640966400000,  # 2022-01-01 00:00:00.000
+                    'pollDelay':  10,
+                    'showMsg':    1,
+                    'showRow':    1,
+                    'snapshot':   0}
+
+        paraDict['vgroups'] = self.vgroups
+        paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+
+        tmqCom.initConsumerTable()
+        tmqCom.create_database(tdSql, paraDict["dbName"],paraDict["dropFlag"], vgroups=paraDict["vgroups"],replica=self.replicaVar)
+        tdLog.info("create stb")
+        tmqCom.create_stable(tdSql, dbName=paraDict["dbName"],stbName=paraDict["stbName"])
+        tdLog.info("create ctb")
+        tmqCom.create_ctable(tdSql, dbName=paraDict["dbName"],stbName=paraDict["stbName"],ctbPrefix=paraDict['ctbPrefix'],
+                             ctbNum=paraDict["ctbNum"],ctbStartIdx=paraDict['ctbStartIdx'])
+        tdLog.info("insert data")
+        tmqCom.insert_data_interlaceByMultiTbl(tsql=tdSql,dbName=paraDict["dbName"],ctbPrefix=paraDict["ctbPrefix"],
+                                               ctbNum=paraDict["ctbNum"],rowsPerTbl=paraDict["rowsPerTbl"],batchNum=paraDict["batchNum"],
+                                               startTs=paraDict["startTs"],ctbStartIdx=paraDict['ctbStartIdx'])
+
+        # tdLog.info("restart taosd to ensure that the data falls into the disk")
+        # tdDnodes.stop(1)
+        # tdDnodes.start(1)
+        # tdSql.query("flush database %s"%(paraDict['dbName']))
+        return
+
+    def tmqCase1(self):
+        tdLog.printNoPrefix("======== test case 1: ")
+        paraDict = {'dbName':     'dbt',
+                    'dropFlag':   1,
+                    'event':      '',
+                    'vgroups':    1,
+                    'stbName':    'stb',
+                    'colPrefix':  'c',
+                    'tagPrefix':  't',
+                    'colSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1},{'type': 'TIMESTAMP', 'count':1}],
+                    'tagSchema':   [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'BINARY', 'len':32, 'count':1},{'type': 'NCHAR', 'len':32, 'count':1}],
+                    'ctbPrefix':  'ctb',
+                    'ctbStartIdx': 0,
+                    'ctbNum':     1,
+                    'rowsPerTbl': 10000,
+                    'batchNum':   1000,
+                    'startTs':    1640966400000,  # 2022-01-01 00:00:00.000
+                    'pollDelay':  20,
+                    'showMsg':    1,
+                    'showRow':    1,
+                    'snapshot':   1}
+
+        paraDict['vgroups'] = self.vgroups
+        paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+        paraDict['snapshot'] = self.snapshot
+
+        topicNameList = ['topic1', 'topic2']
+        tmqCom.initConsumerTable()
+
+        tdLog.info("create topics from stb with filter")
+        # queryString = "select ts, acos(c1), ceil(pow(c1,3)) from %s.%s where (sin(c2) >= 0) and (c1 %% 4 == 0) and (ts >= %d) and (t4 like 'shanghai')"%(paraDict['dbName'], paraDict['stbName'], paraDict["startTs"]+9379)
+        queryString = "select ts, acos(c1), ceil(pow(c1,3)) from %s.%s "%(paraDict['dbName'], paraDict['stbName'])
+        # sqlString = "create topic %s as stable %s" %(topicNameList[0], paraDict['stbName'])
+        sqlString = "create topic %s as %s" %(topicNameList[0], queryString)
+        tdLog.info("create topic sql: %s"%sqlString)
+        tdSql.execute(sqlString)
+        
+        sqlString = "create topic %s as %s" %(topicNameList[1], queryString)
+        tdLog.info("create topic sql: %s"%sqlString)
+        tdSql.execute(sqlString)
+        # tdSql.query(queryString)
+        # expectRowsList.append(tdSql.getRows())
+
+        # init consume info, and start tmq_sim, then check consume result
+        tdLog.info("insert consume info to consume processor")
+        consumerId   = 0
+        expectrowcnt = paraDict["rowsPerTbl"] * paraDict["ctbNum"] * 4
+        topicList    = topicNameList[0] + ',' + topicNameList[0] + ',' + topicNameList[1]
+        ifcheckdata  = 0
+        ifManualCommit = 1
+        keyList      = 'group.id:cgrp1, enable.auto.commit:true, auto.commit.interval.ms:1000, auto.offset.reset:earliest'
+        tmqCom.insertConsumerInfo(consumerId, expectrowcnt,topicList,keyList,ifcheckdata,ifManualCommit)
+        
+        consumerId   = 1
+        tmqCom.insertConsumerInfo(consumerId, expectrowcnt,topicList,keyList,ifcheckdata,ifManualCommit)
+
+        tdLog.info("start consume processor")
+        tmqCom.startTmqSimProcess(pollDelay=paraDict['pollDelay'],dbName=paraDict["dbName"],showMsg=paraDict['showMsg'], showRow=paraDict['showRow'],snapshot=paraDict['snapshot'])
+        tdLog.info("wait the consume result")
+        
+        # continue to insert new rows
+        paraDict['startTs'] = paraDict['startTs'] + int(self.rowsPerTbl)
+        pInsertThread = tmqCom.asyncInsertDataByInterlace(paraDict)
+        pInsertThread.join()         
+
+        expectRows = 2
+        resultList = tmqCom.selectConsumeResult(expectRows)
+        actConsumeTotalRows = resultList[0] + resultList[1]
+        
+        tdLog.info("act consume rows: %d, expect consume rows: %d"%(actConsumeTotalRows, expectrowcnt))
+        
+        if not ((expectrowcnt <= actConsumeTotalRows) or ((resultList[0] == 0) and (resultList[1] >= expectrowcnt)) or ((resultList[1] == 0) and (resultList[0] >= expectrowcnt))):
+            tdLog.info("expect consume rows: %d, act consume rows: %d"%(expectrowcnt, actConsumeTotalRows))
+            tdLog.exit("%d tmq consume rows error!"%consumerId)
+
+        # tmqCom.checkFileContent(consumerId, queryString)
+
+        time.sleep(10)
+        for i in range(len(topicNameList)):
+            tdSql.query("drop topic %s"%topicNameList[i])
+
+        tdLog.printNoPrefix("======== test case 1 end ...... ")
+
+
+    def run(self):
+        tdSql.prepare()
+        self.prepareTestEnv()
+        self.tmqCase1()        
+
+    def stop(self):
+        tdSql.close()
+        tdLog.success(f"{__file__} successfully executed")
+
+event = threading.Event()
+
+tdCases.addLinux(__file__, TDTestCase())
+tdCases.addWindows(__file__, TDTestCase())