Merge from master

snap/snapcraft.yaml

 name: tdengine
 base: core18
-
 version: '2.1.1.0'
 icon: snap/gui/t-dengine.svg
 summary: an open-source big data platform designed and optimized for IoT. 

src/client/src/tscSQLParser.c

@@ -7187,6 +7187,11 @@ int32_t validateSqlNode(SSqlObj* pSql, SSqlNode* pSqlNode, int32_t index) {
   const char* msg1 = "point interpolation query needs timestamp";
   const char* msg2 = "too many tables in from clause";
   const char* msg3 = "start(end) time of query range required or time range too large";
+  // const char* msg5 = "too many columns in selection clause";
+  // const char* msg6 = "too many tables in from clause";
+  // const char* msg7 = "invalid table alias name";
+  // const char* msg8 = "alias name too long";
+  const char* msg9 = "only tag query not compatible with normal column filter";
 
   int32_t code = TSDB_CODE_SUCCESS;
 
@@ -7326,6 +7331,20 @@ int32_t validateSqlNode(SSqlObj* pSql, SSqlNode* pSqlNode, int32_t index) {
       }
     }
 
+    if (tscQueryTags(pQueryInfo)) {
+      SExprInfo* pExpr1 = tscSqlExprGet(pQueryInfo, 0);
+
+      if (pExpr1->base.functionId != TSDB_FUNC_TID_TAG) {
+        int32_t numOfCols = (int32_t)taosArrayGetSize(pQueryInfo->colList);
+        for (int32_t i = 0; i < numOfCols; ++i) {
+          SColumn* pCols = taosArrayGetP(pQueryInfo->colList, i);
+          if (pCols->info.flist.numOfFilters > 0) {
+            return invalidSqlErrMsg(tscGetErrorMsgPayload(pCmd), msg9);
+          }
+        }
+      }
+    }
+
     // parse the having clause in the first place
     if (validateHavingClause(pQueryInfo, pSqlNode->pHaving, pCmd, pSqlNode->pSelNodeList, joinQuery, timeWindowQuery) !=
         TSDB_CODE_SUCCESS) {

src/client/src/tscServer.c

@@ -1928,8 +1928,9 @@ int tscProcessTableMetaRsp(SSqlObj *pSql) {
     }
   }
 
-  tscDebug("0x%"PRIx64" recv table meta, uid:%" PRIu64 ", tid:%d, name:%s", pSql->self, pTableMeta->id.uid, pTableMeta->id.tid,
-           tNameGetTableName(&pTableMetaInfo->name));
+  tscDebug("0x%"PRIx64" recv table meta, uid:%" PRIu64 ", tid:%d, name:%s, numOfCols:%d, numOfTags:%d", pSql->self,
+      pTableMeta->id.uid, pTableMeta->id.tid, tNameGetTableName(&pTableMetaInfo->name), pTableMeta->tableInfo.numOfColumns,
+      pTableMeta->tableInfo.numOfTags);
 
   free(pTableMeta);
   return TSDB_CODE_SUCCESS;
@@ -2072,7 +2073,7 @@ int tscProcessSTableVgroupRsp(SSqlObj *pSql) {
 
     pInfo->vgroupList->numOfVgroups = pVgroupMsg->numOfVgroups;
     if (pInfo->vgroupList->numOfVgroups <= 0) {
-      tscDebug("0x%"PRIx64" empty vgroup info, no corresponding tables for stable", pSql->self);
+      tscDebug("0x%" PRIx64 " empty vgroup info, no corresponding tables for stable", pSql->self);
     } else {
       for (int32_t j = 0; j < pInfo->vgroupList->numOfVgroups; ++j) {
         // just init, no need to lock

src/client/src/tscSub.c

@@ -215,7 +215,7 @@ static void tscProcessSubscriptionTimer(void *handle, void *tmrId) {
   taosTmrReset(tscProcessSubscriptionTimer, pSub->interval, pSub, tscTmr, &pSub->pTimer);
 }
 
-
+//TODO refactor: extract table list name not simply from the sql
 static SArray* getTableList( SSqlObj* pSql ) {
   const char* p = strstr( pSql->sqlstr, " from " );
   assert(p != NULL); // we are sure this is a 'select' statement
@@ -224,11 +224,11 @@ static SArray* getTableList( SSqlObj* pSql ) {
   
   SSqlObj* pNew = taos_query(pSql->pTscObj, sql);
   if (pNew == NULL) {
-    tscError("0x%"PRIx64"failed to retrieve table id: cannot create new sql object.", pSql->self);
+    tscError("0x%"PRIx64" failed to retrieve table id: cannot create new sql object.", pSql->self);
     return NULL;
 
   } else if (taos_errno(pNew) != TSDB_CODE_SUCCESS) {
-    tscError("0x%"PRIx64"failed to retrieve table id,error: %s", pSql->self, tstrerror(taos_errno(pNew)));
+    tscError("0x%"PRIx64" failed to retrieve table id,error: %s", pSql->self, tstrerror(taos_errno(pNew)));
     return NULL;
   }
 

go @ 8ce6d865

-Subproject commit 050667e5b4d0eafa5387e4283e713559b421203f
+Subproject commit 8ce6d86558afc8c0b50c10f990fd2b4270cf06fc

src/inc/taoserror.h

@@ -219,6 +219,7 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_VND_NO_WRITE_AUTH             TAOS_DEF_ERROR_CODE(0, 0x0512)  //"Database write operation denied")
 #define TSDB_CODE_VND_IS_SYNCING                TAOS_DEF_ERROR_CODE(0, 0x0513)  //"Database is syncing")
 #define TSDB_CODE_VND_INVALID_TSDB_STATE        TAOS_DEF_ERROR_CODE(0, 0x0514)  //"Invalid tsdb state")
+#define TSDB_CODE_VND_IS_CLOSING                TAOS_DEF_ERROR_CODE(0, 0x0515)  //"Database is closing")
 
 // tsdb
 #define TSDB_CODE_TDB_INVALID_TABLE_ID          TAOS_DEF_ERROR_CODE(0, 0x0600)  //"Invalid table ID")

src/kit/taosdemo/taosdemo.c

@@ -470,6 +470,7 @@ typedef struct SThreadInfo_S {
 
   // seq of query or subscribe
   uint64_t  querySeq;   // sequence number of sql command
+  TAOS_SUB*  tsub;
 
 } threadInfo;
 

src/mnode/src/mnodeSdb.c

@@ -719,13 +719,13 @@ static int32_t sdbProcessWrite(void *wparam, void *hparam, int32_t qtype, void *
   if (action == SDB_ACTION_INSERT) {
     return sdbPerformInsertAction(pHead, pTable);
   } else if (action == SDB_ACTION_DELETE) {
-    if (qtype == TAOS_QTYPE_FWD) {
+    //if (qtype == TAOS_QTYPE_FWD) {
       // Drop database/stable may take a long time and cause a timeout, so we confirm first then reput it into queue
-      sdbWriteFwdToQueue(1, hparam, TAOS_QTYPE_QUERY, unused);
-      return TSDB_CODE_SUCCESS;
-    } else {
+    //  sdbWriteFwdToQueue(1, hparam, TAOS_QTYPE_QUERY, unused);
+    //  return TSDB_CODE_SUCCESS;
+    //} else {
       return sdbPerformDeleteAction(pHead, pTable);
-    }
+    //}
   } else if (action == SDB_ACTION_UPDATE) {
     return sdbPerformUpdateAction(pHead, pTable);
   } else {

src/mnode/src/mnodeTable.c

@@ -1189,8 +1189,8 @@ static int32_t mnodeFindSuperTableTagIndex(SSTableObj *pStable, const char *tagN
 
 static int32_t mnodeAddSuperTableTagCb(SMnodeMsg *pMsg, int32_t code) {
   SSTableObj *pStable = (SSTableObj *)pMsg->pTable;
-  mLInfo("msg:%p, app:%p stable %s, add tag result:%s", pMsg, pMsg->rpcMsg.ahandle, pStable->info.tableId,
-          tstrerror(code));
+  mLInfo("msg:%p, app:%p stable %s, add tag result:%s, numOfTags:%d", pMsg, pMsg->rpcMsg.ahandle, pStable->info.tableId,
+          tstrerror(code), pStable->numOfTags);
 
   return code;
 }

src/sync/src/syncMain.c

@@ -709,7 +709,7 @@ static void syncChooseMaster(SSyncNode *pNode) {
 }
 
 static SSyncPeer *syncCheckMaster(SSyncNode *pNode) {
-  int32_t onlineNum = 0;
+  int32_t onlineNum = 0, arbOnlineNum = 0;
   int32_t masterIndex = -1;
   int32_t replica = pNode->replica;
 
@@ -723,13 +723,15 @@ static SSyncPeer *syncCheckMaster(SSyncNode *pNode) {
   SSyncPeer *pArb = pNode->peerInfo[TAOS_SYNC_MAX_REPLICA];
   if (pArb && pArb->role != TAOS_SYNC_ROLE_OFFLINE) {
     onlineNum++;
+    ++arbOnlineNum;
     replica = pNode->replica + 1;
   }
 
   if (onlineNum <= replica * 0.5) {
     if (nodeRole != TAOS_SYNC_ROLE_UNSYNCED) {
-       if (nodeRole == TAOS_SYNC_ROLE_MASTER && onlineNum == replica * 0.5 && onlineNum >= 1) {
+      if (nodeRole == TAOS_SYNC_ROLE_MASTER && onlineNum == replica * 0.5 && ((replica > 2 && onlineNum - arbOnlineNum > 1) || pNode->replica < 3)) {
          sInfo("vgId:%d, self keep work as master, online:%d replica:%d", pNode->vgId, onlineNum, replica);
+	 masterIndex = pNode->selfIndex;
        } else {
         nodeRole = TAOS_SYNC_ROLE_UNSYNCED;
         sInfo("vgId:%d, self change to unsynced state, online:%d replica:%d", pNode->vgId, onlineNum, replica);

src/util/src/tcache.c

@@ -613,7 +613,7 @@ void doCleanupDataCache(SCacheObj *pCacheObj) {
 
   // todo memory leak if there are object with refcount greater than 0 in hash table?
   taosHashCleanup(pCacheObj->pHashTable);
-  taosTrashcanEmpty(pCacheObj, true);
+  taosTrashcanEmpty(pCacheObj, false);
 
   __cache_lock_destroy(pCacheObj);
   

src/vnode/src/vnodeMain.c

@@ -454,7 +454,11 @@ void vnodeDestroy(SVnodeObj *pVnode) {
   }
 
   if (pVnode->tsdb) {
-    code = tsdbCloseRepo(pVnode->tsdb, 1);
+    // the deleted vnode does not need to commit, so as to speed up the deletion
+    int toCommit = 1;
+    if (pVnode->dropped) toCommit = 0;
+
+    code = tsdbCloseRepo(pVnode->tsdb, toCommit);
     pVnode->tsdb = NULL;
   }
 

src/vnode/src/vnodeSync.c

@@ -126,11 +126,16 @@ void vnodeStopSyncFile(int32_t vgId, uint64_t fversion) {
 }
 
 void vnodeConfirmForard(int32_t vgId, void *wparam, int32_t code) {
-  void *pVnode = vnodeAcquire(vgId);
+  SVnodeObj *pVnode = vnodeAcquire(vgId);
   if (pVnode == NULL) {
     vError("vgId:%d, vnode not found while confirm forward", vgId);
   }
 
+  if (code == TSDB_CODE_SYN_CONFIRM_EXPIRED && pVnode->status == TAOS_VN_STATUS_CLOSING) {
+    vDebug("vgId:%d, db:%s, vnode is closing while confirm forward", vgId, pVnode->db);
+    code = TSDB_CODE_VND_IS_CLOSING;
+  }
+
   dnodeSendRpcVWriteRsp(pVnode, wparam, code);
   vnodeRelease(pVnode);
 }

tests/mas/Jenkinsfile

+def pre_test(){
+   
+    sh '''
+    sudo rmtaos||echo 'no taosd installed'
+    '''
+    sh '''
+    cd ${WKC}
+    git reset --hard
+    git checkout $BRANCH_NAME
+    git pull
+    git submodule update
+    cd ${WK}
+    git reset --hard
+    git checkout $BRANCH_NAME
+    git pull
+    export TZ=Asia/Harbin
+    date
+    rm -rf ${WK}/debug
+    mkdir debug
+    cd debug
+    cmake .. > /dev/null
+    make > /dev/null
+    make install > /dev/null
+    pip3 install ${WKC}/src/connector/python/linux/python3/
+    '''
+    return 1
+}
+pipeline {
+  agent none
+  environment{
+    
+      WK = '/var/lib/jenkins/workspace/TDinternal'
+      WKC= '/var/lib/jenkins/workspace/TDinternal/community'
+  }
+
+  stages {
+      stage('Parallel test stage') {
+      parallel {
+        stage('pytest') {
+          agent{label 'slam1'}
+          steps {
+            pre_test()
+            sh '''
+            cd ${WKC}/tests
+            find pytest -name '*'sql|xargs rm -rf
+            ./test-all.sh pytest
+            date'''
+          }
+        }
+        stage('test_b1') {
+          agent{label 'slam2'}
+          steps {
+            pre_test()
+            
+            sh '''
+            cd ${WKC}/tests
+            ./test-all.sh b1
+            date'''
+            
+            
+          }
+        }
+
+        stage('test_crash_gen') {
+          agent{label "slam3"}
+          steps {
+            pre_test()
+            sh '''
+            cd ${WKC}/tests/pytest
+            '''
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/pytest
+                ./crash_gen.sh -a -p -t 4 -s 2000
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/pytest
+                rm -rf /var/lib/taos/*
+                rm -rf /var/log/taos/*
+                ./handle_crash_gen_val_log.sh
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/pytest
+                rm -rf /var/lib/taos/*
+                rm -rf /var/log/taos/*
+                ./handle_taosd_val_log.sh
+                '''
+            }
+            
+            sh'''
+            systemctl start taosd 
+            sleep 10
+            '''
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/gotest
+                bash batchtest.sh
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/examples/python/PYTHONConnectorChecker
+                python3 PythonChecker.py
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/examples/JDBC/JDBCDemo/
+                mvn clean package assembly:single -DskipTests >/dev/null 
+                java -jar target/JDBCDemo-SNAPSHOT-jar-with-dependencies.jar -host 127.0.0.1
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/src/connector/jdbc
+                mvn clean package -Dmaven.test.skip=true >/dev/null 
+                cd ${WKC}/tests/examples/JDBC/JDBCDemo/
+                java --class-path=../../../../src/connector/jdbc/target:$JAVA_HOME/jre/lib/ext -jar target/JDBCDemo-SNAPSHOT-jar-with-dependencies.jar -host 127.0.0.1
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cp -rf ${WKC}/tests/examples/nodejs ${JENKINS_HOME}/workspace/
+                cd ${JENKINS_HOME}/workspace/nodejs
+                node nodejsChecker.js host=localhost
+                '''
+            }
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${JENKINS_HOME}/workspace/C#NET/src/CheckC#
+                dotnet run
+                '''
+            }
+            sh '''
+            systemctl stop taosd 
+            cd ${WKC}/tests
+            ./test-all.sh b2
+            date
+            '''
+            sh '''
+            cd ${WKC}/tests
+            ./test-all.sh full unit
+            date'''
+          }
+        }
+
+        stage('test_valgrind') {
+          agent{label "slam4"}
+
+          steps {
+            pre_test()
+            catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                sh '''
+                cd ${WKC}/tests/pytest
+                nohup taosd >/dev/null &
+                sleep 10
+                python3 concurrent_inquiry.py -c 1
+                
+                '''
+            }
+            sh '''
+            cd ${WKC}/tests
+            ./test-all.sh full jdbc
+            date'''
+            sh '''
+            cd ${WKC}/tests/pytest
+            ./valgrind-test.sh 2>&1 > mem-error-out.log
+            ./handle_val_log.sh
+          
+            date
+            cd ${WKC}/tests
+            ./test-all.sh b3
+            date'''
+            sh '''
+            date
+            cd ${WKC}/tests
+            ./test-all.sh full example
+            date'''
+          }
+        }
+       
+       stage('arm64_build'){
+                agent{label 'arm64'}
+                steps{
+                    sh '''
+                    cd ${WK}
+                    git fetch
+                    git checkout develop
+                    git pull
+                    cd ${WKC}
+                    git fetch
+                    git checkout develop
+                    git pull
+                    git submodule update                    
+                    cd ${WKC}/packaging
+                    ./release.sh -v cluster -c aarch64 -n 2.0.0.0 -m 2.0.0.0
+                    
+                    '''
+                }
+            }
+            stage('arm32_build'){
+                agent{label 'arm32'}
+                steps{
+                    catchError(buildResult: 'SUCCESS', stageResult: 'FAILURE') {
+                        sh '''
+                        cd ${WK}
+                        git fetch
+                        git checkout develop
+                        git pull
+                        cd ${WKC}
+                        git fetch
+                        git checkout develop
+                        git pull
+                        git submodule update
+                        cd ${WKC}/packaging
+                        ./release.sh -v cluster -c aarch32 -n 2.0.0.0 -m 2.0.0.0
+                        
+                        '''
+                    }
+                    
+                }
+            }
+      }
+    }
+
+  }
+  post {             
+        success {
+            emailext (
+                subject: "SUCCESSFUL: Job '${env.JOB_NAME} [${env.BUILD_NUMBER}]'",
+                body: '''<!DOCTYPE html>
+                <html>
+                <head>
+                <meta charset="UTF-8">
+                </head>
+                <body leftmargin="8" marginwidth="0" topmargin="8" marginheight="4" offset="0">
+                    <table width="95%" cellpadding="0" cellspacing="0" style="font-size: 16pt; font-family: Tahoma, Arial, Helvetica, sans-serif">
+                        <tr>
+                            <td><br />
+                                <b><font color="#0B610B"><font size="6">构建信息</font></font></b>
+                                <hr size="2" width="100%" align="center" /></td>
+                        </tr>
+                        <tr>
+                            <td>
+                                <ul>
+                                <div style="font-size:18px">
+                                    <li>构建名称>>分支：${PROJECT_NAME}</li>
+                                    <li>构建结果：<span style="color:green"> Successful </span></li>
+                                    <li>构建编号：${BUILD_NUMBER}</li>
+                                    <li>触发用户：${CAUSE}</li>
+                                    <li>变更概要：${CHANGES}</li>
+                                    <li>构建地址：<a href=${BUILD_URL}>${BUILD_URL}</a></li>
+                                    <li>构建日志：<a href=${BUILD_URL}console>${BUILD_URL}console</a></li>
+                                    <li>变更集：${JELLY_SCRIPT}</li>
+                                </div>
+                                </ul>
+                            </td>
+                        </tr>
+                    </table></font>
+                </body>
+                </html>''',
+                to: "yqliu@taosdata.com,pxiao@taosdata.com",
+                from: "support@taosdata.com"
+            )
+        }
+        failure {
+            emailext (
+                subject: "FAILED: Job '${env.JOB_NAME} [${env.BUILD_NUMBER}]'",
+                body: '''<!DOCTYPE html>
+                <html>
+                <head>
+                <meta charset="UTF-8">
+                </head>
+                <body leftmargin="8" marginwidth="0" topmargin="8" marginheight="4" offset="0">
+                    <table width="95%" cellpadding="0" cellspacing="0" style="font-size: 16pt; font-family: Tahoma, Arial, Helvetica, sans-serif">
+                        <tr>
+                            <td><br />
+                                <b><font color="#0B610B"><font size="6">构建信息</font></font></b>
+                                <hr size="2" width="100%" align="center" /></td>
+                        </tr>
+                        <tr>
+                            <td>
+                                <ul>
+                                <div style="font-size:18px">
+                                    <li>构建名称>>分支：${PROJECT_NAME}</li>
+                                    <li>构建结果：<span style="color:green"> Successful </span></li>
+                                    <li>构建编号：${BUILD_NUMBER}</li>
+                                    <li>触发用户：${CAUSE}</li>
+                                    <li>变更概要：${CHANGES}</li>
+                                    <li>构建地址：<a href=${BUILD_URL}>${BUILD_URL}</a></li>
+                                    <li>构建日志：<a href=${BUILD_URL}console>${BUILD_URL}console</a></li>
+                                    <li>变更集：${JELLY_SCRIPT}</li>
+                                </div>
+                                </ul>
+                            </td>
+                        </tr>
+                    </table></font>
+                </body>
+                </html>''',
+                to: "yqliu@taosdata.com,pxiao@taosdata.com",
+                from: "support@taosdata.com"
+            )
+        }
+    }
+}
\ No newline at end of file

tests/pytest/crash_gen/service_manager.py

@@ -22,7 +22,7 @@ from queue import Queue, Empty
 from .shared.config import Config
 from .shared.db import DbTarget, DbConn
 from .shared.misc import Logging, Helper, CrashGenError, Status, Progress, Dice
-from .shared.types import DirPath
+from .shared.types import DirPath, IpcStream
 
 # from crash_gen.misc import CrashGenError, Dice, Helper, Logging, Progress, Status
 # from crash_gen.db import DbConn, DbTarget
@@ -177,13 +177,12 @@ quorum 2
         return "127.0.0.1"
 
     def getServiceCmdLine(self): # to start the instance
-        cmdLine = []
         if Config.getConfig().track_memory_leaks:
             Logging.info("Invoking VALGRIND on service...")
-            cmdLine = ['valgrind', '--leak-check=yes']
-        # TODO: move "exec -c" into Popen(), we can both "use shell" and NOT fork so ask to lose kill control
-        cmdLine += ["exec " + self.getExecFile(), '-c', self.getCfgDir()] # used in subproce.Popen()        
-        return cmdLine
+            return ['exec /usr/bin/valgrind', '--leak-check=yes', self.getExecFile(), '-c', self.getCfgDir()]
+        else:
+            # TODO: move "exec -c" into Popen(), we can both "use shell" and NOT fork so ask to lose kill control
+            return ["exec " + self.getExecFile(), '-c', self.getCfgDir()] # used in subproce.Popen()
     
     def _getDnodes(self, dbc):
         dbc.query("show dnodes")
@@ -281,16 +280,16 @@ class TdeSubProcess:
         return '[TdeSubProc: pid = {}, status = {}]'.format(
             self.getPid(), self.getStatus() )
 
-    def getStdOut(self) -> BinaryIO :
+    def getIpcStdOut(self) -> IpcStream :
         if self._popen.universal_newlines : # alias of text_mode
             raise CrashGenError("We need binary mode for STDOUT IPC")
         # Logging.info("Type of stdout is: {}".format(type(self._popen.stdout)))
-        return typing.cast(BinaryIO, self._popen.stdout)
+        return typing.cast(IpcStream, self._popen.stdout)
 
-    def getStdErr(self) -> BinaryIO :
+    def getIpcStdErr(self) -> IpcStream :
         if self._popen.universal_newlines : # alias of text_mode
             raise CrashGenError("We need binary mode for STDERR IPC")
-        return typing.cast(BinaryIO, self._popen.stderr)
+        return typing.cast(IpcStream, self._popen.stderr)
 
     # Now it's always running, since we matched the life cycle
     # def isRunning(self):
@@ -301,11 +300,6 @@ class TdeSubProcess:
 
     def _start(self, cmdLine) -> Popen :
         ON_POSIX = 'posix' in sys.builtin_module_names
-
-        # Sanity check
-        # if self.subProcess:  # already there
-        #     raise RuntimeError("Corrupt process state")
-
         
         # Prepare environment variables for coverage information
         # Ref: https://stackoverflow.com/questions/2231227/python-subprocess-popen-with-a-modified-environment
@@ -314,9 +308,8 @@ class TdeSubProcess:
 
         # print(myEnv)
         # print("Starting TDengine with env: ", myEnv.items())
-        # print("Starting TDengine via Shell: {}".format(cmdLineStr))
+        print("Starting TDengine: {}".format(cmdLine))
 
-        # useShell = True # Needed to pass environments into it
         return Popen(            
             ' '.join(cmdLine), # ' '.join(cmdLine) if useShell else cmdLine,
             shell=True, # Always use shell, since we need to pass ENV vars
@@ -732,19 +725,19 @@ class ServiceManagerThread:
         self._ipcQueue = Queue() # type: Queue
         self._thread = threading.Thread( # First thread captures server OUTPUT
             target=self.svcOutputReader,
-            args=(subProc.getStdOut(), self._ipcQueue, logDir))
+            args=(subProc.getIpcStdOut(), self._ipcQueue, logDir))
         self._thread.daemon = True  # thread dies with the program
         self._thread.start()
         time.sleep(0.01)
         if not self._thread.is_alive(): # What happened?
-            Logging.info("Failed to started process to monitor STDOUT")
+            Logging.info("Failed to start process to monitor STDOUT")
             self.stop()
             raise CrashGenError("Failed to start thread to monitor STDOUT")
         Logging.info("Successfully started process to monitor STDOUT")
 
         self._thread2 = threading.Thread( # 2nd thread captures server ERRORs
             target=self.svcErrorReader,
-            args=(subProc.getStdErr(), self._ipcQueue, logDir))
+            args=(subProc.getIpcStdErr(), self._ipcQueue, logDir))
         self._thread2.daemon = True  # thread dies with the program
         self._thread2.start()
         time.sleep(0.01)
@@ -887,14 +880,19 @@ class ServiceManagerThread:
             print("\nNon-UTF8 server output: {}\n".format(bChunk.decode('cp437')))
             return None
 
-    def _textChunkGenerator(self, streamIn: BinaryIO, logDir: str, logFile: str
+    def _textChunkGenerator(self, streamIn: IpcStream, logDir: str, logFile: str
             ) -> Generator[TextChunk, None, None]:
         '''
-        Take an input stream with binary data, produced a generator of decoded
-        "text chunks", and also save the original binary data in a log file.
+        Take an input stream with binary data (likely from Popen), produced a generator of decoded
+        "text chunks".
+        
+        Side effect: it also save the original binary data in a log file.
         '''
         os.makedirs(logDir, exist_ok=True)
         logF = open(os.path.join(logDir, logFile), 'wb')
+        if logF is None:
+            Logging.error("Failed to open log file (binary write): {}/{}".format(logDir, logFile))
+            return
         for bChunk in iter(streamIn.readline, b''):
             logF.write(bChunk) # Write to log file immediately
             tChunk = self._decodeBinaryChunk(bChunk) # decode
@@ -902,14 +900,14 @@ class ServiceManagerThread:
                 yield tChunk # TODO: split into actual text lines
 
         # At the end...
-        streamIn.close() # Close the stream
-        logF.close() # Close the output file
+        streamIn.close() # Close the incoming stream
+        logF.close() # Close the log file
 
-    def svcOutputReader(self, stdOut: BinaryIO, queue, logDir: str):
+    def svcOutputReader(self, ipcStdOut: IpcStream, queue, logDir: str):
         '''
         The infinite routine that processes the STDOUT stream for the sub process being managed.
 
-        :param stdOut: the IO stream object used to fetch the data from
+        :param ipcStdOut: the IO stream object used to fetch the data from
         :param queue: the queue where we dump the roughly parsed chunk-by-chunk text data
         :param logDir: where we should dump a verbatim output file
         '''
@@ -917,7 +915,7 @@ class ServiceManagerThread:
         # Important Reference: https://stackoverflow.com/questions/375427/non-blocking-read-on-a-subprocess-pipe-in-python
         # print("This is the svcOutput Reader...")
         # stdOut.readline() # Skip the first output? TODO: remove?
-        for tChunk in self._textChunkGenerator(stdOut, logDir, 'stdout.log') :
+        for tChunk in self._textChunkGenerator(ipcStdOut, logDir, 'stdout.log') :
             queue.put(tChunk) # tChunk garanteed not to be None
             self._printProgress("_i")
 
@@ -940,12 +938,12 @@ class ServiceManagerThread:
         Logging.info("EOF found TDengine STDOUT, marking the process as terminated")
         self.setStatus(Status.STATUS_STOPPED)
 
-    def svcErrorReader(self, stdErr: BinaryIO, queue, logDir: str):
+    def svcErrorReader(self, ipcStdErr: IpcStream, queue, logDir: str):
         # os.makedirs(logDir, exist_ok=True)
         # logFile = os.path.join(logDir,'stderr.log')
         # fErr = open(logFile, 'wb')
         # for line in iter(err.readline, b''):
-        for tChunk in self._textChunkGenerator(stdErr, logDir, 'stderr.log') :
+        for tChunk in self._textChunkGenerator(ipcStdErr, logDir, 'stderr.log') :
             queue.put(tChunk) # tChunk garanteed not to be None
             # fErr.write(line)
             Logging.info("TDengine STDERR: {}".format(tChunk))

tests/pytest/crash_gen/shared/types.py

-from typing import Any, List, Dict, NewType
+from typing import Any, BinaryIO, List, Dict, NewType
 from enum import Enum
 
 DirPath = NewType('DirPath', str)
@@ -26,3 +26,5 @@ class TdDataType(Enum):
 
 TdColumns = Dict[str, TdDataType]
 TdTags    = Dict[str, TdDataType]
+
+IpcStream = NewType('IpcStream', BinaryIO)
\ No newline at end of file

tests/script/general/parser/select_with_tags.sim

@@ -887,10 +887,16 @@ sql_error select tbname, t1 from select_tags_mt0 interval(1y);
 #valid sql: select first(c1), last(c2), count(*) from select_tags_mt0 group by tbname, t1;
 #valid sql: select first(c1), tbname, t1 from select_tags_mt0 group by t2;
 
+print ==================================>TD-4231
+sql_error select t1,tbname from select_tags_mt0 where c1<0
+sql_error select t1,tbname from select_tags_mt0 where c1<0 and tbname in ('select_tags_tb12')
+
+sql select tbname from select_tags_mt0 where tbname in ('select_tags_tb12');
+
 sql_error select first(c1), last(c2), t1 from select_tags_mt0 group by tbname;
 sql_error select first(c1), last(c2), tbname, t2 from select_tags_mt0 group by tbname;
 sql_error select first(c1), count(*), t2, t1, tbname from select_tags_mt0 group by tbname;
-# this sql is valid: select first(c1), t2 from select_tags_mt0 group by tbname;
+#valid sql: select first(c1), t2 from select_tags_mt0 group by tbname;
 
 #sql select first(ts), tbname from select_tags_mt0 group by tbname;
 #sql select count(c1) from select_tags_mt0 where c1=99 group by tbname;

tests/script/unique/arbitrator/dn3_mn1_vnode_nomaster.sim

@@ -158,7 +158,7 @@ if $dnode4Vtatus != offline then
   sleep 2000
   goto wait_dnode4_vgroup_offline
 endi
-if $dnode3Vtatus != master then
+if $dnode3Vtatus != unsynced then
   sleep 2000
   goto wait_dnode4_vgroup_offline
 endi