Merge pull request #14900 from taosdata/test3.0/lihui

test:add test case for tmq

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

@@ -428,7 +428,7 @@ class TMQCom:
         pThread.start()
         return pThread
 
-    def checkFileContent(self, consumerId, queryString):
+    def checkFileContent(self, consumerId, queryString, skipRowsOfCons=0):
         buildPath = tdCom.getBuildPath()
         cfgPath = tdCom.getClientCfgPath()
         dstFile = '%s/../log/dstrows_%d.txt'%(cfgPath, consumerId)
@@ -445,6 +445,10 @@ class TMQCom:
         # skip first line for it is schema
         queryFile.readline()
         
+        # skip offset for consumer
+        for i in range(0,skipRowsOfCons):
+            consumeFile.readline()            
+        
         lines = 0
         while True:
             dst = queryFile.readline()

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

+
+import taos
+import sys
+import time
+import socket
+import os
+import threading
+from enum import Enum
+
+from util.log import *
+from util.sql import *
+from util.cases import *
+from util.dnodes import *
+sys.path.append("./7-tmq")
+from tmqCommon import *
+
+class TDTestCase:
+    def __init__(self):
+        self.snapshot   = 0
+        self.vgroups    = 4
+        self.ctbNum     = 1
+        self.rowsPerTbl = 10000
+        
+    def init(self, conn, logSql):
+        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':     1,
+                    'rowsPerTbl': 100000,
+                    'batchNum':   1200,
+                    'startTs':    1640966400000,  # 2022-01-01 00:00:00.000
+                    'pollDelay':  3,
+                    'showMsg':    1,
+                    'showRow':    1,
+                    'snapshot':   0}
+
+        paraDict['vgroups'] = self.vgroups
+        paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+        
+        tmqCom.initConsumerTable()
+        tdCom.create_database(tdSql, paraDict["dbName"],paraDict["dropFlag"], vgroups=paraDict["vgroups"],replica=1)
+        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'])
+        # tmqCom.insert_data_with_autoCreateTbl(tsql=tdSql,dbName=paraDict["dbName"],stbName=paraDict["stbName"],ctbPrefix="ctbx",
+        #                                       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")        
+        # tdSql.query("flush database %s"%(paraDict['dbName']))
+        return
+    
+    def delData(self,tsql,dbName,ctbPrefix,ctbNum,startTs=0,endTs=0,ctbStartIdx=0):
+        tdLog.debug("start to del data ............")
+        for i in range(ctbNum):
+            sql = "delete from %s.%s%d where _c0 >= %d and _c0 <= %d "%(dbName,ctbPrefix,i+ctbStartIdx,startTs,endTs)
+            tsql.execute(sql)
+
+        tdLog.debug("del data ............ [OK]")
+        return
+
+    def tmqCase1(self):
+        tdLog.printNoPrefix("======== test case 1: ")
+        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': 100000,
+                    'batchNum':   3000,
+                    'startTs':    1640966400000,  # 2022-01-01 00:00:00.000
+                    'endTs': 0,
+                    'pollDelay':  5,
+                    'showMsg':    1,
+                    'showRow':    1,
+                    'snapshot':   0}
+        paraDict['snapshot'] = self.snapshot
+        paraDict['vgroups'] = self.vgroups
+        paraDict['ctbNum'] = self.ctbNum
+        paraDict['rowsPerTbl'] = self.rowsPerTbl
+        
+        # del some data        
+        rowsOfDelete = int(paraDict["rowsPerTbl"] / 4)
+        paraDict["endTs"] = paraDict["startTs"] + rowsOfDelete - 1
+        self.delData(tsql=tdSql,dbName=paraDict["dbName"],ctbPrefix=paraDict["ctbPrefix"],ctbNum=paraDict["ctbNum"], 
+                     startTs=paraDict["startTs"], endTs=paraDict["endTs"],ctbStartIdx=paraDict['ctbStartIdx'])
+                
+        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     = 0
+        expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"])
+        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)
+
+        tdLog.info("start consume processor")
+        tmqCom.startTmqSimProcess(pollDelay=paraDict['pollDelay'],dbName=paraDict["dbName"],showMsg=paraDict['showMsg'], showRow=paraDict['showRow'],snapshot=paraDict['snapshot'])
+
+        tdLog.info("insert process end, and start to check consume result")
+        expectRows = 1
+        resultList = tmqCom.selectConsumeResult(expectRows)
+        totalConsumeRows = 0
+        for i in range(expectRows):
+            totalConsumeRows += resultList[i]
+
+        tdSql.query(queryString)
+        totalRowsInserted = tdSql.getRows()
+                
+        tdLog.info("act consume rows: %d, expect consume rows: %d, act insert rows: %d"%(totalConsumeRows, expectrowcnt, totalRowsInserted))
+        if totalConsumeRows != expectrowcnt:
+            tdLog.exit("tmq consume rows error!")
+            
+        tmqCom.checkFileContent(consumerId, queryString, rowsOfDelete) 
+
+        tdSql.query("drop topic %s"%topicFromStb1)
+        tdLog.printNoPrefix("======== test case 1 end ...... ")
+
+    def tmqCase2(self):
+        tdLog.printNoPrefix("======== test case 2: ")  
+        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':  5,
+                    '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']))
+        
+        # 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'])        
+        # 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'])     
+        
+        # del some data
+        rowsOfDelete = int(self.rowsPerTbl / 4 )
+        paraDict["endTs"] = paraDict["startTs"] + rowsOfDelete - 1
+        self.delData(tsql=tdSql,dbName=paraDict["dbName"],ctbPrefix=paraDict["ctbPrefix"],ctbNum=paraDict["ctbNum"], 
+                     startTs=paraDict["startTs"], endTs=paraDict["endTs"],ctbStartIdx=paraDict['ctbStartIdx'])
+        
+        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:
+            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1 + 1/4 + 3/4))
+        elif self.snapshot == 1:
+            expectrowcnt   = int(paraDict["rowsPerTbl"] * paraDict["ctbNum"] * (1 - 1/4 + 1/4 + 3/4 - 1/4))
+        
+        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)
+
+        tdLog.info("start consume processor")
+        tmqCom.startTmqSimProcess(pollDelay=paraDict['pollDelay'],dbName=paraDict["dbName"],showMsg=paraDict['showMsg'], showRow=paraDict['showRow'],snapshot=paraDict['snapshot'])
+        
+        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)
+        totalRowsInserted = tdSql.getRows()
+        
+        tdLog.info("act consume rows: %d, act insert rows: %d, expect consume rows: %d, "%(totalConsumeRows, totalRowsInserted, expectrowcnt))
+        
+        if totalConsumeRows != expectrowcnt:
+            tdLog.exit("tmq consume rows error!")
+            
+        # tmqCom.checkFileContent(consumerId, queryString)   
+
+        tdSql.query("drop topic %s"%topicFromStb1)
+
+        tdLog.printNoPrefix("======== test case 2 end ...... ")
+
+    def run(self):
+        tdSql.prepare()
+        self.prepareTestEnv()
+        tdLog.printNoPrefix("=============================================")
+        tdLog.printNoPrefix("======== snapshot is 0: only consume from wal")
+        self.tmqCase1()
+        self.tmqCase2()
+        
+        self.prepareTestEnv()
+        tdLog.printNoPrefix("====================================================================")
+        tdLog.printNoPrefix("======== snapshot is 1: firstly consume from tsbs, and then from wal")
+        self.snapshot = 1
+        self.tmqCase1()
+        self.tmqCase2()        
+
+    def stop(self):
+        tdSql.close()
+        tdLog.success(f"{__file__} successfully executed")
+
+event = threading.Event()
+
+tdCases.addLinux(__file__, TDTestCase())
+tdCases.addWindows(__file__, TDTestCase())

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

@@ -235,18 +235,18 @@ class TDTestCase:
 
     def run(self):
         tdSql.prepare()
-        # self.prepareTestEnv()
-        # tdLog.printNoPrefix("=============================================")
-        # tdLog.printNoPrefix("======== snapshot is 0: only consume from wal")
-        # self.tmqCase1()
-        # self.tmqCase2()
+        self.prepareTestEnv()
+        tdLog.printNoPrefix("=============================================")
+        tdLog.printNoPrefix("======== snapshot is 0: only consume from wal")
+        self.tmqCase1()
+        self.tmqCase2()
         
         self.prepareTestEnv()
         tdLog.printNoPrefix("====================================================================")
         tdLog.printNoPrefix("======== snapshot is 1: firstly consume from tsbs, and then from wal")
         self.snapshot = 1
         self.tmqCase1()
-        # self.tmqCase2()
+        self.tmqCase2()
         
 
     def stop(self):