Merge branch 'develop' into feature/python-test-no-sudo

.travis.yml

@@ -24,10 +24,11 @@ matrix:
           - python-setuptools
           - python3-pip
           - python3-setuptools
+          - valgrind
 
     before_install:
       - sudo apt update -y -qq
-      - sudo apt install -y net-tools python-pip python-setuptools python3-pip python3-setuptools
+      - sudo apt install -y net-tools python-pip python-setuptools python3-pip python3-setuptools valgrind
 
     before_script:
       - cd ${TRAVIS_BUILD_DIR}
@@ -43,16 +44,32 @@ matrix:
         case $TRAVIS_OS_NAME in
           linux)
           cd ${TRAVIS_BUILD_DIR}/debug
-          sudo make install || exit $?
+          sudo make install || travis_terminate $?
 
           pip install --user ${TRAVIS_BUILD_DIR}/src/connector/python/linux/python2/
           pip3 install --user ${TRAVIS_BUILD_DIR}/src/connector/python/linux/python3/
 
           cd ${TRAVIS_BUILD_DIR}/tests
-          bash ./test-all.sh
+          ./test-all.sh || travis_terminate $?
 
-          if [ "$?" -ne "0" ]; then
-            exit $?
+          cd ${TRAVIS_BUILD_DIR}/tests/pytest
+          ./simpletest.sh -g 2>&1 | tee mem-error-out.txt
+          sleep 1
+
+          # Color setting
+          RED='\033[0;31m'
+          GREEN='\033[1;32m'
+          GREEN_DARK='\033[0;32m'
+          GREEN_UNDERLINE='\033[4;32m'
+          NC='\033[0m'
+
+          memError=`grep -m 1 'ERROR SUMMARY' mem-error-out.txt | awk '{print $4}'`
+
+          if [ -n "$memError" ]; then
+            if [ "$memError" -gt 23 ]; then
+              echo -e "${RED} ## Memory errors number valgrind reports is $memError. More than our threshold! ## ${NC} "
+              travis_terminate $memError
+            fi
           fi
 
           ;;
@@ -74,20 +91,20 @@ matrix:
         # GitHub project metadata
         # ** specific to your project **
         project:
-          name: sangshuduo/TDengine
+          name: TDengine
           version: 2.x
-          description: sangshuduo/TDengine
+          description: taosdata/TDengine
 
         # Where email notification of build analysis results will be sent
-        notification_email: sangshuduo@gmail.com
+        notification_email: sdsang@taosdata.com
 
         # Commands to prepare for build_command
         # ** likely specific to your build **
-        build_command_prepend: cmake ..
+        build_command_prepend: cmake .
 
         # The command that will be added as an argument to "cov-build" to compile your project for analysis,
         # ** likely specific to your build **
-        build_command: cmake --build .
+        build_command: make
 
         # Pattern to match selecting branches that will run analysis. We recommend leaving this set to 'coverity_scan'.
         # Take care in resource usage, and consider the build frequency allowances per
@@ -132,17 +149,17 @@ matrix:
         case $TRAVIS_OS_NAME in
           linux)
           cd ${TRAVIS_BUILD_DIR}/debug
-          sudo make install || exit $?
+          sudo make install || travis_terminate $?
 
           pip install --user ${TRAVIS_BUILD_DIR}/src/connector/python/linux/python2/
           pip3 install --user ${TRAVIS_BUILD_DIR}/src/connector/python/linux/python3/
 
           cd ${TRAVIS_BUILD_DIR}/tests
 
-          bash ./test-all.sh
+          ./test-all.sh
 
           if [ "$?" -ne "0" ]; then
-            exit $?
+            travis_terminate $?
           fi
 
           sudo pkill taosd
@@ -150,7 +167,7 @@ matrix:
 
           cd ${TRAVIS_BUILD_DIR}
           lcov -d . --capture --rc lcov_branch_coverage=1 -o coverage.info
-          lcov -l --rc lcov_branch_coverage=1 coverage.info || exit $?
+          lcov -l --rc lcov_branch_coverage=1 coverage.info || travis_terminate $?
 
           gem install coveralls-lcov
 
@@ -166,7 +183,6 @@ matrix:
             echo -e "${GREEN} ## Uploaded to Coveralls.io! ## ${NC}"
           else
             echo -e "${RED} ## Coveralls.io not collect coverage report! ## ${NC} "
-            exit $?
           fi
 
           bash <(curl -s https://codecov.io/bash) -y .codecov.yml -f coverage.info
@@ -174,7 +190,6 @@ matrix:
             echo -e "${GREEN} ## Uploaded to Codecov! ## ${NC} "
           else
             echo -e "${RED} ## Codecov did not collect coverage report! ## ${NC} "
-            exit $?
           fi
 
           ;;

src/client/inc/tsclient.h

@@ -465,7 +465,7 @@ extern void *     tscQhandle;
 extern int        tscKeepConn[];
 extern int        tsInsertHeadSize;
 extern int        tscNumOfThreads;
-extern SRpcIpSet  tscMgmtIpList;
+extern SRpcIpSet  tscMgmtIpSet;
 
 typedef void (*__async_cb_func_t)(void *param, TAOS_RES *tres, int numOfRows);
 

src/client/src/tscServer.c

@@ -30,7 +30,7 @@
 
 #define TSC_MGMT_VNODE 999
 
-SRpcIpSet  tscMgmtIpList;
+SRpcIpSet  tscMgmtIpSet;
 SRpcIpSet  tscDnodeIpSet;
 
 int (*tscBuildMsg[TSDB_SQL_MAX])(SSqlObj *pSql, SSqlInfo *pInfo) = {0};
@@ -58,35 +58,40 @@ static void tscSetDnodeIpList(SSqlObj* pSql, STableMeta* pTableMeta) {
 }
 
 void tscPrintMgmtIp() {
-  if (tscMgmtIpList.numOfIps <= 0) {
-    tscError("invalid mgmt IP list:%d", tscMgmtIpList.numOfIps);
+  if (tscMgmtIpSet.numOfIps <= 0) {
+    tscError("invalid mgmt IP list:%d", tscMgmtIpSet.numOfIps);
   } else {
-    for (int i = 0; i < tscMgmtIpList.numOfIps; ++i) {
-      tscTrace("mgmt index:%d ip:%d", i, tscMgmtIpList.ip[i]);
+    for (int i = 0; i < tscMgmtIpSet.numOfIps; ++i) {
+      tscTrace("mgmt index:%d ip:%d", i, tscMgmtIpSet.ip[i]);
     }
   }
 }
 
 void tscSetMgmtIpListFromCluster(SRpcIpSet *pIpList) {
-  tscMgmtIpList.numOfIps = htons(pIpList->numOfIps);
-  tscMgmtIpList.inUse = htons(pIpList->inUse);
-  tscMgmtIpList.port = htons(pIpList->port);
-  for (int32_t i = 0; i <tscMgmtIpList.numOfIps; ++i) {
-    tscMgmtIpList.ip[i] = pIpList->ip[i];
+  tscMgmtIpSet.numOfIps = pIpList->numOfIps;
+  tscMgmtIpSet.inUse = pIpList->inUse;
+  tscMgmtIpSet.port = htons(pIpList->port);
+  for (int32_t i = 0; i < tscMgmtIpSet.numOfIps; ++i) {
+    tscMgmtIpSet.ip[i] = htonl(pIpList->ip[i]);
   }
 }
 
 void tscSetMgmtIpListFromEdge() {
-  if (tscMgmtIpList.numOfIps != 1) {
-    tscMgmtIpList.numOfIps = 1;
-    tscMgmtIpList.inUse = 0;
-    tscMgmtIpList.port = tsMnodeShellPort;
-    tscMgmtIpList.ip[0] = inet_addr(tsMasterIp);
+  if (tscMgmtIpSet.numOfIps != 1) {
+    tscMgmtIpSet.numOfIps = 1;
+    tscMgmtIpSet.inUse = 0;
+    tscMgmtIpSet.port = tsMnodeShellPort;
+    tscMgmtIpSet.ip[0] = inet_addr(tsMasterIp);
     tscTrace("edge mgmt IP list:");
     tscPrintMgmtIp();
   }
 }
 
+void tscUpdateIpSet(void *ahandle, SRpcIpSet *pIpSet) {
+  tscTrace("mgmt IP list is changed for ufp is called");
+  tscMgmtIpSet = *pIpSet;
+}
+
 void tscSetMgmtIpList(SRpcIpSet *pIpList) {
   /*
     * The iplist returned by the cluster edition is the current management nodes
@@ -109,7 +114,7 @@ void tscSetMgmtIpList(SRpcIpSet *pIpList) {
 UNUSED_FUNC
 static int32_t tscGetMgmtConnMaxRetryTimes() {
   int32_t factor = 2;
-  return tscMgmtIpList.numOfIps * factor;
+  return tscMgmtIpSet.numOfIps * factor;
 }
 
 void tscProcessHeartBeatRsp(void *param, TAOS_RES *tres, int code) {
@@ -204,7 +209,7 @@ int tscSendMsgToServer(SSqlObj *pSql) {
     };
     rpcSendRequest(pVnodeConn, &pSql->ipList, &rpcMsg);
   } else {
-    pSql->ipList = tscMgmtIpList;
+    pSql->ipList = tscMgmtIpSet;
     pSql->ipList.port = tsMnodeShellPort;
     
     tscTrace("%p msg:%s is sent to server %d", pSql, taosMsg[pSql->cmd.msgType], pSql->ipList.port);
@@ -425,7 +430,7 @@ int tscProcessSql(SSqlObj *pSql) {
       return pSql->res.code;
     }
   } else if (pSql->cmd.command < TSDB_SQL_LOCAL) {
-    pSql->ipList = tscMgmtIpList;
+    pSql->ipList = tscMgmtIpSet;
   } else {  // local handler
     return (*tscProcessMsgRsp[pCmd->command])(pSql);
   }
@@ -2224,10 +2229,7 @@ int tscProcessConnectRsp(SSqlObj *pSql) {
   assert(len <= tListLen(pObj->db));
   strncpy(pObj->db, temp, tListLen(pObj->db));
   
-//  SIpList *    pIpList;
-//  char *rsp = pRes->pRsp + sizeof(SCMConnectRsp);
-//  pIpList = (SIpList *)rsp;
-//  tscSetMgmtIpList(pIpList);
+  tscSetMgmtIpList(&pConnect->ipList);
 
   strcpy(pObj->sversion, pConnect->serverVersion);
   pObj->writeAuth = pConnect->writeAuth;

src/client/src/tscSql.c

@@ -72,23 +72,23 @@ STscObj *taosConnectImpl(const char *ip, const char *user, const char *pass, con
   }
 
   if (ip && ip[0]) {
-    tscMgmtIpList.inUse = 0;
-    tscMgmtIpList.port = tsMnodeShellPort;
-    tscMgmtIpList.numOfIps = 1;
-    tscMgmtIpList.ip[0] = inet_addr(ip);
+    tscMgmtIpSet.inUse = 0;
+    tscMgmtIpSet.port = tsMnodeShellPort;
+    tscMgmtIpSet.numOfIps = 1;
+    tscMgmtIpSet.ip[0] = inet_addr(ip);
 
     if (tsMasterIp[0] && strcmp(ip, tsMasterIp) != 0) {
-      tscMgmtIpList.numOfIps = 2;
-      tscMgmtIpList.ip[1] = inet_addr(tsMasterIp);
+      tscMgmtIpSet.numOfIps = 2;
+      tscMgmtIpSet.ip[1] = inet_addr(tsMasterIp);
     }
 
     if (tsSecondIp[0] && strcmp(tsSecondIp, tsMasterIp) != 0) {
-      tscMgmtIpList.numOfIps = 3;
-      tscMgmtIpList.ip[2] = inet_addr(tsSecondIp);
+      tscMgmtIpSet.numOfIps = 3;
+      tscMgmtIpSet.ip[2] = inet_addr(tsSecondIp);
     }
   }
 
-  tscMgmtIpList.port = port ? port : tsMnodeShellPort;
+  tscMgmtIpSet.port = port ? port : tsMnodeShellPort;
   
   STscObj *pObj = (STscObj *)calloc(1, sizeof(STscObj));
   if (NULL == pObj) {

src/client/src/tscSystem.c

@@ -42,11 +42,13 @@ void *  tscTmr;
 void *  tscQhandle;
 void *  tscCheckDiskUsageTmr;
 int     tsInsertHeadSize;
+char    tsLastUser[TSDB_USER_LEN + 1];
 
 int tscNumOfThreads;
 
 static pthread_once_t tscinit = PTHREAD_ONCE_INIT;
 void taosInitNote(int numOfNoteLines, int maxNotes, char* lable);
+void tscUpdateIpSet(void *ahandle, SRpcIpSet *pIpSet);
 
 void tscCheckDiskUsage(void *para, void *unused) {
   taosGetDisk();
@@ -65,6 +67,7 @@ int32_t tscInitRpc(const char *user, const char *secret) {
     rpcInit.label = "TSC-vnode";
     rpcInit.numOfThreads = tscNumOfThreads;
     rpcInit.cfp = tscProcessMsgFromServer;
+    rpcInit.ufp = tscUpdateIpSet;
     rpcInit.sessions = tsMaxVnodeConnections;
     rpcInit.connType = TAOS_CONN_CLIENT;
     rpcInit.user = (char*)user;
@@ -79,6 +82,13 @@ int32_t tscInitRpc(const char *user, const char *secret) {
     }
   }
 
+  // not stop service, switch users
+  if (strcmp(tsLastUser, user) != 0 && pTscMgmtConn != NULL) {
+    tscTrace("switch user from %s to %s", user, tsLastUser);
+    rpcClose(pTscMgmtConn);
+    pTscMgmtConn = NULL;
+  }
+
   if (pTscMgmtConn == NULL) {
     memset(&rpcInit, 0, sizeof(rpcInit));
     rpcInit.localIp = tsLocalIp;
@@ -92,6 +102,7 @@ int32_t tscInitRpc(const char *user, const char *secret) {
     rpcInit.user = (char*)user;
     rpcInit.ckey = "key";
     rpcInit.secret = secretEncrypt;
+    strcpy(tsLastUser, user);
 
     pTscMgmtConn = rpcOpen(&rpcInit);
     if (pTscMgmtConn == NULL) {
@@ -145,14 +156,14 @@ void taos_init_imp() {
     taosInitNote(tsNumOfLogLines / 10, 1, (char*)"tsc_note");
   }
 
-  tscMgmtIpList.inUse = 0;
-  tscMgmtIpList.port = tsMnodeShellPort;
-  tscMgmtIpList.numOfIps = 1;
-  tscMgmtIpList.ip[0] = inet_addr(tsMasterIp);
+  tscMgmtIpSet.inUse = 0;
+  tscMgmtIpSet.port = tsMnodeShellPort;
+  tscMgmtIpSet.numOfIps = 1;
+  tscMgmtIpSet.ip[0] = inet_addr(tsMasterIp);
 
   if (tsSecondIp[0] && strcmp(tsSecondIp, tsMasterIp) != 0) {
-    tscMgmtIpList.numOfIps = 2;
-    tscMgmtIpList.ip[1] = inet_addr(tsSecondIp);
+    tscMgmtIpSet.numOfIps = 2;
+    tscMgmtIpSet.ip[1] = inet_addr(tsSecondIp);
   }
 
   tscInitMsgsFp();

src/dnode/src/dnodeMClient.c

@@ -54,6 +54,11 @@ static SRpcIpSet     tsMnodeIpSet  = {0};
 static SDMMnodeInfos tsMnodeInfos = {0};
 static SDMDnodeCfg   tsDnodeCfg = {0};
 
+void dnodeUpdateIpSet(void *ahandle, SRpcIpSet *pIpSet) {
+  dTrace("mgmt IP list is changed for ufp is called");
+  tsMnodeIpSet = *pIpSet;
+}
+
 int32_t dnodeInitMClient() {
   dnodeReadDnodeCfg();
   tsRebootTime = taosGetTimestampSec();
@@ -90,6 +95,7 @@ int32_t dnodeInitMClient() {
   rpcInit.label        = "DND-MC";
   rpcInit.numOfThreads = 1;
   rpcInit.cfp          = dnodeProcessRspFromMnode;
+  rpcInit.ufp          = dnodeUpdateIpSet;
   rpcInit.sessions     = 100;
   rpcInit.connType     = TAOS_CONN_CLIENT;
   rpcInit.idleTime     = tsShellActivityTimer * 2000;
@@ -292,7 +298,7 @@ static bool dnodeReadMnodeInfos() {
     tsMnodeInfos.nodeInfos[i].syncPort = (uint16_t)syncPort->valueint;
 
     cJSON *nodeName = cJSON_GetObjectItem(nodeInfo, "nodeName");
-    if (!nodeIp || nodeName->type != cJSON_String || nodeName->valuestring == NULL) {
+    if (!nodeName || nodeName->type != cJSON_String || nodeName->valuestring == NULL) {
       dError("failed to read mnode mgmtIpList.json, nodeName not found");
       goto PARSE_OVER;
     }
@@ -304,7 +310,7 @@ static bool dnodeReadMnodeInfos() {
   dPrint("read mnode iplist successed, numOfIps:%d inUse:%d", tsMnodeInfos.nodeNum, tsMnodeInfos.inUse);
   for (int32_t i = 0; i < tsMnodeInfos.nodeNum; i++) {
     dPrint("mnode:%d, ip:%s:%u name:%s", tsMnodeInfos.nodeInfos[i].nodeId,
-            taosIpStr(tsMnodeInfos.nodeInfos[i].nodeId), tsMnodeInfos.nodeInfos[i].nodePort,
+            taosIpStr(tsMnodeInfos.nodeInfos[i].nodeIp), tsMnodeInfos.nodeInfos[i].nodePort,
             tsMnodeInfos.nodeInfos[i].nodeName);
   }
 

src/dnode/src/dnodeMgmt.c

@@ -33,7 +33,6 @@ static int32_t  dnodeOpenVnodes();
 static void     dnodeCloseVnodes();
 static int32_t  dnodeProcessCreateVnodeMsg(SRpcMsg *pMsg);
 static int32_t  dnodeProcessDropVnodeMsg(SRpcMsg *pMsg);
-static int32_t  dnodeProcessAlterVnodeMsg(SRpcMsg *pMsg);
 static int32_t  dnodeProcessAlterStreamMsg(SRpcMsg *pMsg);
 static int32_t  dnodeProcessConfigDnodeMsg(SRpcMsg *pMsg);
 static int32_t (*dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MAX])(SRpcMsg *pMsg);
@@ -41,7 +40,6 @@ static int32_t (*dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MAX])(SRpcMsg *pMsg);
 int32_t dnodeInitMgmt() {
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_CREATE_VNODE] = dnodeProcessCreateVnodeMsg;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_DROP_VNODE]   = dnodeProcessDropVnodeMsg;
-  dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_ALTER_VNODE]  = dnodeProcessAlterVnodeMsg;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_ALTER_STREAM] = dnodeProcessAlterStreamMsg;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_CONFIG_DNODE] = dnodeProcessConfigDnodeMsg;
 
@@ -130,24 +128,30 @@ static void dnodeCloseVnodes() {
 static int32_t dnodeProcessCreateVnodeMsg(SRpcMsg *rpcMsg) {
   SMDCreateVnodeMsg *pCreate = rpcMsg->pCont;
   pCreate->cfg.vgId                = htonl(pCreate->cfg.vgId);
-  pCreate->cfg.maxSessions     = htonl(pCreate->cfg.maxSessions);
-  pCreate->cfg.cacheBlockSize  = htonl(pCreate->cfg.cacheBlockSize);
+  pCreate->cfg.maxTables           = htonl(pCreate->cfg.maxTables);
+  pCreate->cfg.maxCacheSize        = htobe64(pCreate->cfg.maxCacheSize);
+  pCreate->cfg.minRowsPerFileBlock = htonl(pCreate->cfg.minRowsPerFileBlock);
+  pCreate->cfg.maxRowsPerFileBlock = htonl(pCreate->cfg.maxRowsPerFileBlock);
   pCreate->cfg.daysPerFile         = htonl(pCreate->cfg.daysPerFile);
   pCreate->cfg.daysToKeep1         = htonl(pCreate->cfg.daysToKeep1);
   pCreate->cfg.daysToKeep2         = htonl(pCreate->cfg.daysToKeep2);
   pCreate->cfg.daysToKeep          = htonl(pCreate->cfg.daysToKeep);
   pCreate->cfg.commitTime          = htonl(pCreate->cfg.commitTime);
-  pCreate->cfg.rowsInFileBlock = htonl(pCreate->cfg.rowsInFileBlock);
-  pCreate->cfg.blocksPerTable  = htons(pCreate->cfg.blocksPerTable);
-  pCreate->cfg.cacheNumOfBlocks.totalBlocks = htonl(pCreate->cfg.cacheNumOfBlocks.totalBlocks);
+  pCreate->cfg.arbitratorIp        = htonl(pCreate->cfg.arbitratorIp);
 
   for (int32_t j = 0; j < pCreate->cfg.replications; ++j) {
-    pCreate->vpeerDesc[j].vgId    = htonl(pCreate->vpeerDesc[j].vgId);
-    pCreate->vpeerDesc[j].dnodeId = htonl(pCreate->vpeerDesc[j].dnodeId);
-    pCreate->vpeerDesc[j].ip      = htonl(pCreate->vpeerDesc[j].ip);
+    pCreate->nodes[j].nodeId = htonl(pCreate->nodes[j].nodeId);
+    pCreate->nodes[j].nodeIp = htonl(pCreate->nodes[j].nodeIp);
   }
   
+  void *pVnode = vnodeAccquireVnode(pCreate->cfg.vgId);
+  if (pVnode != NULL) {
+    int32_t code = vnodeAlter(pVnode, pCreate);
+    vnodeRelease(pVnode);
+    return code;
+  } else {
     return vnodeCreate(pCreate);
+  }
 }
 
 static int32_t dnodeProcessDropVnodeMsg(SRpcMsg *rpcMsg) {
@@ -157,15 +161,6 @@ static int32_t dnodeProcessDropVnodeMsg(SRpcMsg *rpcMsg) {
   return vnodeDrop(pDrop->vgId);
 }
 
-static int32_t dnodeProcessAlterVnodeMsg(SRpcMsg *rpcMsg) {
-  SMDCreateVnodeMsg *pCreate = rpcMsg->pCont;
-  pCreate->cfg.vgId        = htonl(pCreate->cfg.vgId);
-  pCreate->cfg.maxSessions = htonl(pCreate->cfg.maxSessions);
-  pCreate->cfg.daysPerFile = htonl(pCreate->cfg.daysPerFile);
-
-  return 0;
-}
-
 static int32_t dnodeProcessAlterStreamMsg(SRpcMsg *pMsg) {
 //  SMDAlterStreamMsg *pStream = pCont;
 //  pStream->uid    = htobe64(pStream->uid);

src/dnode/src/dnodeMnode.c

@@ -33,7 +33,6 @@ int32_t dnodeInitMnode() {
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_DROP_STABLE]  = dnodeWrite;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_CREATE_VNODE] = dnodeMgmt;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_DROP_VNODE]   = dnodeMgmt;
-  dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_ALTER_VNODE]  = dnodeMgmt;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_ALTER_STREAM] = dnodeMgmt;
   dnodeProcessMgmtMsgFp[TSDB_MSG_TYPE_MD_CONFIG_DNODE] = dnodeMgmt;
 

src/inc/taoserror.h

@@ -75,14 +75,19 @@ TAOS_DEFINE_ERROR(TSDB_CODE_NOT_CONFIGURED,             0, 27, "not configured")
 TAOS_DEFINE_ERROR(TSDB_CODE_NODE_OFFLINE,               0, 28, "node offline")
 TAOS_DEFINE_ERROR(TSDB_CODE_NETWORK_UNAVAIL,            0, 29, "network unavailable")
 
-// db & user
+// db
 TAOS_DEFINE_ERROR(TSDB_CODE_DB_NOT_SELECTED,            0, 100, "db not selected")
 TAOS_DEFINE_ERROR(TSDB_CODE_DB_ALREADY_EXIST,           0, 101, "database aleady exist")
 TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_DB,                 0, 102, "invalid database")
 TAOS_DEFINE_ERROR(TSDB_CODE_MONITOR_DB_FORBIDDEN,       0, 103, "monitor db forbidden")
-TAOS_DEFINE_ERROR(TSDB_CODE_USER_ALREADY_EXIST,         0, 104, "user already exist")
-TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_USER,               0, 105, "invalid user")
-TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_PASS,               0, 106, "invalid password")
+
+// user
+TAOS_DEFINE_ERROR(TSDB_CODE_USER_ALREADY_EXIST,         0, 150, "user already exist")
+TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_USER,               0, 151, "invalid user")
+TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_PASS,               0, 152, "invalid password")
+TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_USER_FORMAT,        0, 153, "invalid user format")
+TAOS_DEFINE_ERROR(TSDB_CODE_INVALID_PASS_FORMAT,        0, 154, "invalid password format")
+TAOS_DEFINE_ERROR(TSDB_CODE_NO_USER_FROM_CONN,          0, 155, "can not get user from conn")
 
 // table
 TAOS_DEFINE_ERROR(TSDB_CODE_TABLE_ALREADY_EXIST,        0, 200, "table already exist")

src/inc/taosmsg.h

@@ -48,14 +48,12 @@ extern "C" {
 #define TSDB_MSG_TYPE_MD_CREATE_VNODE_RSP 16
 #define TSDB_MSG_TYPE_MD_DROP_VNODE       17
 #define TSDB_MSG_TYPE_MD_DROP_VNODE_RSP   18
-#define TSDB_MSG_TYPE_MD_ALTER_VNODE      19
-#define TSDB_MSG_TYPE_MD_ALTER_VNODE_RSP  20
-#define TSDB_MSG_TYPE_MD_DROP_STABLE      21
-#define TSDB_MSG_TYPE_MD_DROP_STABLE_RSP  22
-#define TSDB_MSG_TYPE_MD_ALTER_STREAM     23
-#define TSDB_MSG_TYPE_MD_ALTER_STREAM_RSP 24
-#define TSDB_MSG_TYPE_MD_CONFIG_DNODE     25
-#define TSDB_MSG_TYPE_MD_CONFIG_DNODE_RSP 26
+#define TSDB_MSG_TYPE_MD_DROP_STABLE      19
+#define TSDB_MSG_TYPE_MD_DROP_STABLE_RSP  20
+#define TSDB_MSG_TYPE_MD_ALTER_STREAM     21
+#define TSDB_MSG_TYPE_MD_ALTER_STREAM_RSP 22
+#define TSDB_MSG_TYPE_MD_CONFIG_DNODE     23
+#define TSDB_MSG_TYPE_MD_CONFIG_DNODE_RSP 24
 
 // message from client to mnode
 #define TSDB_MSG_TYPE_CM_CONNECT          31
@@ -245,12 +243,6 @@ typedef struct SSchema {
   int16_t bytes;
 } SSchema;
 
-typedef struct {
-  int32_t  vgId;
-  int32_t  dnodeId;
-  uint32_t ip;
-} SVnodeDesc;
-
 typedef struct {
   int32_t  contLen;
   int32_t  vgId;
@@ -518,12 +510,10 @@ typedef struct {
   uint8_t status;
   uint8_t role;
   uint8_t accessState;
+  uint8_t replica;
   uint8_t reserved[5];
 } SVnodeLoad;
 
-/*
- * NOTE: sizeof(SVnodeCfg) < TSDB_FILE_HEADER_LEN / 4
- */
 typedef struct {
   char     acct[TSDB_USER_LEN + 1];
   char     db[TSDB_DB_NAME_LEN + 1];
@@ -548,7 +538,7 @@ typedef struct {
   int8_t  loadLatest;  // load into mem or not
   uint8_t precision;   // time resolution
   int8_t  reserved[16];
-} SVnodeCfg, SDbCfg, SCMCreateDbMsg, SCMAlterDbMsg;
+} SDbCfg, SCMCreateDbMsg, SCMAlterDbMsg;
 
 typedef struct {
   char    db[TSDB_TABLE_ID_LEN + 1];
@@ -614,8 +604,35 @@ typedef struct {
 } SDMStatusRsp;
 
 typedef struct {
-  SVnodeCfg  cfg;
-  SVnodeDesc vpeerDesc[TSDB_MAX_MPEERS];
+  uint32_t vgId;
+  int32_t  maxTables;
+  int64_t  maxCacheSize;
+  int32_t  minRowsPerFileBlock;
+  int32_t  maxRowsPerFileBlock;
+  int32_t  daysPerFile;
+  int32_t  daysToKeep;
+  int32_t  daysToKeep1;
+  int32_t  daysToKeep2;
+  int32_t  commitTime;
+  uint8_t  precision;  // time resolution
+  int8_t   compression;
+  int8_t   wals;
+  int8_t   commitLog;
+  int8_t   replications;
+  int8_t   quorum;
+  uint32_t arbitratorIp;
+  int8_t   reserved[16];
+} SMDVnodeCfg;
+
+typedef struct {
+  int32_t  nodeId;
+  uint32_t nodeIp;
+  char     nodeName[TSDB_NODE_NAME_LEN + 1];
+} SMDVnodeDesc;
+
+typedef struct {
+  SMDVnodeCfg  cfg;
+  SMDVnodeDesc nodes[TSDB_MAX_MPEERS];
 } SMDCreateVnodeMsg;
 
 typedef struct {
@@ -673,9 +690,16 @@ typedef struct {
   int32_t metaElem[TSDB_MAX_JOIN_TABLE_NUM];
 } SSuperTableMetaMsg;
 
+typedef struct {
+  int32_t  nodeId;
+  uint32_t nodeIp;
+  uint16_t nodePort;
+} SVnodeDesc;
+
 typedef struct {
   SVnodeDesc vpeerDesc[TSDB_REPLICA_MAX_NUM];
   int16_t    index;  // used locally
+  int32_t    vgId;
   int32_t    numOfSids;
   int32_t    pSidExtInfoList[];  // offset value of STableIdInfo
 } SVnodeSidList;

src/inc/vnode.h

@@ -38,6 +38,7 @@ typedef struct {
 int32_t vnodeCreate(SMDCreateVnodeMsg *pVnodeCfg);
 int32_t vnodeDrop(int32_t vgId);
 int32_t vnodeOpen(int32_t vgId, char *rootDir);
+int32_t vnodeAlter(void *pVnode, SMDCreateVnodeMsg *pVnodeCfg);
 int32_t vnodeClose(int32_t vgId);
 
 void    vnodeRelease(void *pVnode);

src/mnode/inc/mgmtAcct.h

@@ -24,10 +24,10 @@ extern "C" {
 
 int32_t mgmtInitAccts();
 void    mgmtCleanUpAccts();
-void   *mgmtGetAcct(char *acctName);
+void *  mgmtGetAcct(char *acctName);
+void *  mgmtGetNextAcct(void *pNode, SAcctObj **pAcct);
 void    mgmtIncAcctRef(SAcctObj *pAcct);
 void    mgmtDecAcctRef(SAcctObj *pAcct);
-
 void    mgmtAddDbToAcct(SAcctObj *pAcct, SDbObj *pDb);
 void    mgmtDropDbFromAcct(SAcctObj *pAcct, SDbObj *pDb);
 void    mgmtAddUserToAcct(SAcctObj *pAcct, SUserObj *pUser);

src/mnode/inc/mgmtDef.h

@@ -51,7 +51,6 @@ typedef struct SDnodeObj {
   int8_t     reserved[15];
   int8_t     updateEnd[1];
   int32_t    refCount;
-  SVnodeLoad vload[TSDB_MAX_VNODES];
   uint32_t   moduleStatus;
   uint32_t   lastReboot;       // time stamp for last reboot
   float      score;          // calc in balance function
@@ -72,13 +71,6 @@ typedef struct SMnodeObj {
   SDnodeObj *pDnode;
 } SMnodeObj;
 
-
-typedef struct {
-  int32_t  dnodeId;
-  uint32_t privateIp;
-  uint32_t publicIp;
-} SVnodeGid;
-
 typedef struct {
   char   tableId[TSDB_TABLE_ID_LEN + 1];
   int8_t type;
@@ -120,6 +112,13 @@ typedef struct {
   SSuperTableObj *superTable;
 } SChildTableObj;
 
+typedef struct {
+  int32_t    dnodeId;
+  int8_t     role;
+  int8_t     reserved[3];
+  SDnodeObj* pDnode;
+} SVnodeGid;
+
 typedef struct SVgObj {
   uint32_t       vgId;
   char           dbName[TSDB_DB_NAME_LEN + 1];
@@ -136,8 +135,11 @@ typedef struct SVgObj {
   struct SVgObj *prev, *next;
   struct SDbObj *pDb;
   int32_t        numOfTables;
+  int64_t        totalStorage;
+  int64_t        compStorage;
+  int64_t        pointsWritten;
   void *         idPool;
-  SChildTableObj ** tableList;
+  SChildTableObj **tableList;
 } SVgObj;
 
 typedef struct SDbObj {

src/mnode/inc/mgmtDnode.h

@@ -35,12 +35,15 @@ void    mgmtMonitorDnodeModule();
 
 int32_t mgmtGetDnodesNum();
 void *  mgmtGetNextDnode(void *pNode, SDnodeObj **pDnode);
-void    mgmtReleaseDnode(SDnodeObj *pDnode);
+void    mgmtIncDnodeRef(SDnodeObj *pDnode);
+void    mgmtDecDnodeRef(SDnodeObj *pDnode);
 void *  mgmtGetDnode(int32_t dnodeId);
 void *  mgmtGetDnodeByIp(uint32_t ip);
 void    mgmtUpdateDnode(SDnodeObj *pDnode);
 int32_t mgmtDropDnode(SDnodeObj *pDnode);
 
+extern int32_t tsAccessSquence;
+
 #ifdef __cplusplus
 }
 #endif

src/mnode/inc/mgmtMnode.h

@@ -40,7 +40,7 @@ void *  mgmtGetNextMnode(void *pNode, struct SMnodeObj **pMnode);
 void    mgmtReleaseMnode(struct SMnodeObj *pMnode);
 
 char *  mgmtGetMnodeRoleStr();
-void    mgmtGetMnodeIpList(SRpcIpSet *ipSet, bool usePublicIp);
+void    mgmtGetMnodeIpSet(SRpcIpSet *ipSet, bool usePublicIp);
 void    mgmtGetMnodeInfos(void *mnodes);
 
 #ifdef __cplusplus

src/mnode/inc/mgmtUser.h

@@ -24,7 +24,9 @@ extern "C" {
 int32_t   mgmtInitUsers();
 void      mgmtCleanUpUsers();
 SUserObj *mgmtGetUser(char *name);
-void      mgmtReleaseUser(SUserObj *pUser);
+void *    mgmtGetNextUser(void *pNode, SUserObj **pUser);
+void      mgmtIncUserRef(SUserObj *pUser);
+void      mgmtDecUserRef(SUserObj *pUser);
 SUserObj *mgmtGetUserFromConn(void *pConn, bool *usePublicIp);
 int32_t   mgmtCreateUser(SAcctObj *pAcct, char *name, char *pass);
 void      mgmtDropAllUsers(SAcctObj *pAcct);

src/mnode/inc/mgmtVgroup.h

@@ -30,12 +30,13 @@ enum _TSDB_VG_STATUS {
 int32_t mgmtInitVgroups();
 void    mgmtCleanUpVgroups();
 SVgObj *mgmtGetVgroup(int32_t vgId);
-void    mgmtReleaseVgroup(SVgObj *pVgroup);
+void    mgmtIncVgroupRef(SVgObj *pVgroup);
+void    mgmtDecVgroupRef(SVgObj *pVgroup);
 void    mgmtDropAllVgroups(SDbObj *pDropDb);
 
 void *  mgmtGetNextVgroup(void *pNode, SVgObj **pVgroup);
 void    mgmtUpdateVgroup(SVgObj *pVgroup);
-void    mgmtUpdateVgroupStatus(SVgObj *pVgroup, int32_t dnodeId, SVnodeLoad *pVload);
+void    mgmtUpdateVgroupStatus(SVgObj *pVgroup, SDnodeObj *dnodeId, SVnodeLoad *pVload);
 
 void    mgmtCreateVgroup(SQueuedMsg *pMsg, SDbObj *pDb);
 void    mgmtDropVgroup(SVgObj *pVgroup, void *ahandle);
@@ -46,6 +47,7 @@ void    mgmtAddTableIntoVgroup(SVgObj *pVgroup, SChildTableObj *pTable);
 void    mgmtRemoveTableFromVgroup(SVgObj *pVgroup, SChildTableObj *pTable);
 void    mgmtSendCreateVnodeMsg(SVgObj *pVgroup, SRpcIpSet *ipSet, void *ahandle);
 void    mgmtSendDropVnodeMsg(int32_t vgId, SRpcIpSet *ipSet, void *ahandle);
+void    mgmtSendCreateVgroupMsg(SVgObj *pVgroup, void *ahandle);
 
 SRpcIpSet mgmtGetIpSetFromVgroup(SVgObj *pVgroup);
 SRpcIpSet mgmtGetIpSetFromIp(uint32_t ip);

src/mnode/src/mgmtAcct.c

@@ -58,6 +58,7 @@ static int32_t mgmtActionAcctUpdate(SSdbOper *pOper) {
     memcpy(pSaved, pAcct, tsAcctUpdateSize);
     free(pAcct);
   }
+  mgmtDecAcctRef(pSaved);
   return TSDB_CODE_SUCCESS;
 }
 
@@ -106,11 +107,11 @@ int32_t mgmtInitAccts() {
 
   tsAcctSdb = sdbOpenTable(&tableDesc);
   if (tsAcctSdb == NULL) {
-    mError("failed to init acct data");
+    mError("table:%s, failed to create hash", tableDesc.tableName);
     return -1;
   }
 
-  mTrace("table:accounts table is created");
+  mTrace("table:%s, hash is created", tableDesc.tableName);
   return acctInit();
 }
 
@@ -123,6 +124,10 @@ void *mgmtGetAcct(char *name) {
   return sdbGetRow(tsAcctSdb, name);
 }
 
+void *mgmtGetNextAcct(void *pNode, SAcctObj **pAcct) {
+  return sdbFetchRow(tsAcctSdb, pNode, (void **)pAcct); 
+}
+
 void mgmtIncAcctRef(SAcctObj *pAcct) {
   sdbIncRef(tsAcctSdb, pAcct);
 }

src/mnode/src/mgmtBalance.c

@@ -47,7 +47,7 @@ int32_t balanceAllocVnodes(SVgObj *pVgroup) {
         vnodeUsage = usage;
       }
     }
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   if (pSelDnode == NULL) {
@@ -56,8 +56,7 @@ int32_t balanceAllocVnodes(SVgObj *pVgroup) {
   }
 
   pVgroup->vnodeGid[0].dnodeId = pSelDnode->dnodeId;
-  pVgroup->vnodeGid[0].privateIp = pSelDnode->privateIp;
-  pVgroup->vnodeGid[0].publicIp = pSelDnode->publicIp;
+  pVgroup->vnodeGid[0].pDnode = pSelDnode;
 
   mTrace("dnode:%d, alloc one vnode to vgroup, openVnodes:%d", pSelDnode->dnodeId, pSelDnode->openVnodes);
   return TSDB_CODE_SUCCESS;

src/mnode/src/mgmtDServer.c

@@ -28,8 +28,10 @@
 #include "mgmtLog.h"
 #include "mgmtDb.h"
 #include "mgmtDServer.h"
+#include "mgmtMnode.h"
 #include "mgmtProfile.h"
 #include "mgmtShell.h"
+#include "mgmtSdb.h"
 #include "mgmtTable.h"
 #include "mgmtVgroup.h"
 
@@ -100,6 +102,18 @@ static void mgmtProcessMsgFromDnode(SRpcMsg *rpcMsg) {
     return;
   }
 
+  if (!sdbIsMaster()) {
+    SRpcConnInfo connInfo;
+    rpcGetConnInfo(rpcMsg->handle, &connInfo);
+    bool usePublicIp = false;
+    
+    SRpcIpSet ipSet = {0};
+    mgmtGetMnodeIpSet(&ipSet, usePublicIp);
+    mTrace("conn from dnode ip:%s redirect msg", taosIpStr(connInfo.clientIp));
+    rpcSendRedirectRsp(rpcMsg->handle, &ipSet);
+    return;
+  }
+  
   if (mgmtProcessDnodeMsgFp[rpcMsg->msgType]) {
     SRpcMsg *pMsg = malloc(sizeof(SRpcMsg));
     memcpy(pMsg, rpcMsg, sizeof(SRpcMsg));

src/mnode/src/mgmtDb.c

@@ -63,6 +63,7 @@ static int32_t mgmtDbActionInsert(SSdbOper *pOper) {
 
   if (pAcct != NULL) {
     mgmtAddDbToAcct(pAcct, pDb);
+    mgmtDecAcctRef(pAcct);
   }
   else {
     mError("db:%s, acct:%s info not exist in sdb", pDb->name, pDb->cfg.acct);
@@ -80,6 +81,7 @@ static int32_t mgmtDbActionDelete(SSdbOper *pOper) {
   mgmtDropAllChildTables(pDb);
   mgmtDropAllSuperTables(pDb);
   mgmtDropAllVgroups(pDb);
+  mgmtDecAcctRef(pAcct);
   
   return TSDB_CODE_SUCCESS;
 }
@@ -527,7 +529,7 @@ static int32_t mgmtGetDbMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn)
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
   pShow->numOfRows = pUser->pAcct->acctInfo.numOfDbs;
 
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
   return 0;
 }
 
@@ -647,7 +649,7 @@ static int32_t mgmtRetrieveDbs(SShowObj *pShow, char *data, int32_t rows, void *
   }
 
   pShow->numOfReads += numOfRows;
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
   return numOfRows;
 }
 

src/mnode/src/mgmtDnode.c

@@ -23,6 +23,7 @@
 #include "tutil.h"
 #include "tsocket.h"
 #include "tbalance.h"
+#include "tsync.h"
 #include "dnode.h"
 #include "mgmtDef.h"
 #include "mgmtLog.h"
@@ -37,6 +38,8 @@
 
 void   *tsDnodeSdb = NULL;
 int32_t tsDnodeUpdateSize = 0;
+int32_t tsAccessSquence = 0;
+extern void *  tsMnodeSdb;
 extern void *  tsVgroupSdb;
 
 static int32_t mgmtCreateDnode(uint32_t ip);
@@ -99,7 +102,13 @@ static int32_t mgmtDnodeActionDelete(SSdbOper *pOper) {
     }
   }
 
-  mgmtDropMnode(pDnode->dnodeId);
+  SMnodeObj *pMnode = mgmtGetMnode(pDnode->dnodeId);
+  if (pMnode != NULL) {
+    SSdbOper oper = {.type = SDB_OPER_LOCAL, .table = tsMnodeSdb, .pObj = pMnode};
+    sdbDeleteRow(&oper);
+    mgmtReleaseMnode(pMnode);
+  }
+
   balanceNotify();
 
   mTrace("dnode:%d, all vgroups:%d is dropped from sdb", pDnode->dnodeId, numOfVgroups);
@@ -139,7 +148,7 @@ static int32_t mgmtDnodeActionRestored() {
     mgmtCreateDnode(ip);
     SDnodeObj *pDnode = mgmtGetDnodeByIp(ip);
     mgmtAddMnode(pDnode->dnodeId);
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   return TSDB_CODE_SUCCESS;
@@ -215,13 +224,17 @@ void *mgmtGetDnodeByIp(uint32_t ip) {
     if (ip == pDnode->privateIp) {
       return pDnode;
     }
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   return NULL;
 }
 
-void mgmtReleaseDnode(SDnodeObj *pDnode) {
+void mgmtIncDnodeRef(SDnodeObj *pDnode) {
+  sdbIncRef(tsDnodeSdb, pDnode);
+}
+
+void mgmtDecDnodeRef(SDnodeObj *pDnode) {
   sdbDecRef(tsDnodeSdb, pDnode);
 }
 
@@ -318,27 +331,27 @@ void mgmtProcessDnodeStatusMsg(SRpcMsg *rpcMsg) {
   pDnode->alternativeRole  = pStatus->alternativeRole;
   pDnode->totalVnodes      = pStatus->numOfTotalVnodes; 
   pDnode->moduleStatus     = pStatus->moduleStatus;
+  pDnode->lastAccess       = tsAccessSquence;
   
   if (pStatus->dnodeId == 0) {
     mTrace("dnode:%d, first access, privateIp:%s, name:%s", pDnode->dnodeId, taosIpStr(pDnode->privateIp), pDnode->dnodeName);
+  } else {
+    //mTrace("dnode:%d, status received, access times %d", pDnode->dnodeId, pDnode->lastAccess);
   }
  
   int32_t openVnodes = htons(pStatus->openVnodes);
   for (int32_t j = 0; j < openVnodes; ++j) {
     SVnodeLoad *pVload = &pStatus->load[j];
-    pDnode->vload[j].vgId          = htonl(pVload->vgId);
-    pDnode->vload[j].totalStorage  = htobe64(pVload->totalStorage);
-    pDnode->vload[j].compStorage   = htobe64(pVload->compStorage);
-    pDnode->vload[j].pointsWritten = htobe64(pVload->pointsWritten);
+    pVload->vgId = htonl(pVload->vgId);
 
-    SVgObj *pVgroup = mgmtGetVgroup(pDnode->vload[j].vgId);
+    SVgObj *pVgroup = mgmtGetVgroup(pVload->vgId);
     if (pVgroup == NULL) {
       SRpcIpSet ipSet = mgmtGetIpSetFromIp(pDnode->privateIp);
-      mPrint("dnode:%d, vgroup:%d not exist in mnode, drop it", pDnode->dnodeId, pDnode->vload[j].vgId);
-      mgmtSendDropVnodeMsg(pDnode->vload[j].vgId, &ipSet, NULL);
+      mPrint("dnode:%d, vgroup:%d not exist in mnode, drop it", pDnode->dnodeId, pVload->vgId);
+      mgmtSendDropVnodeMsg(pVload->vgId, &ipSet, NULL);
     } else {
-      mgmtUpdateVgroupStatus(pVgroup, pDnode->dnodeId, pVload);
-      mgmtReleaseVgroup(pVgroup);
+      mgmtUpdateVgroupStatus(pVgroup, pDnode, pVload);
+      mgmtDecVgroupRef(pVgroup);
     }
   }
 
@@ -348,7 +361,7 @@ void mgmtProcessDnodeStatusMsg(SRpcMsg *rpcMsg) {
     balanceNotify();
   }
 
-  mgmtReleaseDnode(pDnode);
+  mgmtDecDnodeRef(pDnode);
 
   int32_t contLen = sizeof(SDMStatusRsp) + TSDB_MAX_VNODES * sizeof(SDMVgroupAccess);
   SDMStatusRsp *pRsp = rpcMallocCont(contLen);
@@ -444,7 +457,7 @@ static int32_t mgmtDropDnodeByIp(uint32_t ip) {
     return TSDB_CODE_NO_REMOVE_MASTER;
   }
 
-#ifndef _VPEER
+#ifndef _SYNC
   return mgmtDropDnode(pDnode);
 #else
   return balanceDropDnode(pDnode);
@@ -495,7 +508,10 @@ static int32_t mgmtGetDnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) return 0;
 
-  if (strcmp(pUser->pAcct->user, "root") != 0) return TSDB_CODE_NO_RIGHTS;
+  if (strcmp(pUser->pAcct->user, "root") != 0) {
+    mgmtDecUserRef(pUser);
+    return TSDB_CODE_NO_RIGHTS;
+  }
 
   int32_t  cols = 0;
   SSchema *pSchema = pMeta->schema;
@@ -554,7 +570,7 @@ static int32_t mgmtGetDnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
   pShow->pNode = NULL;
 
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 
   return 0;
 }
@@ -604,7 +620,7 @@ static int32_t mgmtRetrieveDnodes(SShowObj *pShow, char *data, int32_t rows, voi
 
  
     numOfRows++;
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   pShow->numOfReads += numOfRows;
@@ -622,7 +638,10 @@ static int32_t mgmtGetModuleMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pC
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) return 0;
 
-  if (strcmp(pUser->user, "root") != 0) return TSDB_CODE_NO_RIGHTS;
+  if (strcmp(pUser->user, "root") != 0)  {
+    mgmtDecUserRef(pUser);
+    return TSDB_CODE_NO_RIGHTS;
+  }
 
   SSchema *pSchema = pMeta->schema;
 
@@ -661,7 +680,7 @@ static int32_t mgmtGetModuleMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pC
   pShow->numOfRows = mgmtGetDnodesNum() * TSDB_MOD_MAX;
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
   pShow->pNode = NULL;
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 
   return 0;
 }
@@ -712,7 +731,7 @@ int32_t mgmtRetrieveModules(SShowObj *pShow, char *data, int32_t rows, void *pCo
       numOfRows++;
     }
 
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   pShow->numOfReads += numOfRows;
@@ -731,7 +750,10 @@ static int32_t mgmtGetConfigMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pC
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) return 0;
 
-  if (strcmp(pUser->user, "root") != 0) return TSDB_CODE_NO_RIGHTS;
+  if (strcmp(pUser->user, "root") != 0)  {
+    mgmtDecUserRef(pUser);
+    return TSDB_CODE_NO_RIGHTS;
+  }
 
   SSchema *pSchema = pMeta->schema;
 
@@ -762,7 +784,7 @@ static int32_t mgmtGetConfigMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pC
 
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
   pShow->pNode = NULL;
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 
   return 0;
 }
@@ -814,7 +836,11 @@ static int32_t mgmtGetVnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   int32_t cols = 0;
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) return 0;
-  if (strcmp(pUser->user, "root") != 0) return TSDB_CODE_NO_RIGHTS;
+  
+  if (strcmp(pUser->user, "root") != 0)  {
+    mgmtDecUserRef(pUser);
+    return TSDB_CODE_NO_RIGHTS;
+  }
 
   SSchema *pSchema = pMeta->schema;
 
@@ -840,35 +866,18 @@ static int32_t mgmtGetVnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   if (pShow->payloadLen > 0 ) {
     uint32_t ip = ip2uint(pShow->payload);
     pDnode = mgmtGetDnodeByIp(ip);
-    if (NULL == pDnode) {
-      return TSDB_CODE_NODE_OFFLINE;
-    }
-
-    SVnodeLoad* pVnode;
-    pShow->numOfRows = 0;
-    for (int32_t i = 0 ; i < TSDB_MAX_VNODES; i++) {
-      pVnode = &pDnode->vload[i];
-      if (0 != pVnode->vgId) {
-        pShow->numOfRows++;
-      }
-    }
-    
-    pShow->pNode = pDnode;
   } else {
-    while (true) {
-      pShow->pNode = mgmtGetNextDnode(pShow->pNode, (SDnodeObj **)&pDnode);
-      if (pDnode == NULL) break;
-      pShow->numOfRows += pDnode->openVnodes;
-
-      if (0 == pShow->numOfRows) return TSDB_CODE_NODE_OFFLINE;      
+    mgmtGetNextDnode(NULL, (SDnodeObj **)&pDnode);
   }
 
-    pShow->pNode = NULL;
+  if (pDnode != NULL) {
+    pShow->numOfRows += pDnode->openVnodes;
+    mgmtDecDnodeRef(pDnode);
   }
 
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
-  mgmtReleaseDnode(pDnode);
-  mgmtReleaseUser(pUser);
+  pShow->pNode = pDnode;
+  mgmtDecUserRef(pUser);
 
   return 0;
 }
@@ -881,32 +890,32 @@ static int32_t mgmtRetrieveVnodes(SShowObj *pShow, char *data, int32_t rows, voi
 
   if (0 == rows) return 0;
 
-  if (pShow->payloadLen) {
-    // output the vnodes info of the designated dnode. And output all vnodes of this dnode, instead of rows (max 100)
   pDnode = (SDnodeObj *)(pShow->pNode);
   if (pDnode != NULL) {
-      SVnodeLoad* pVnode;
-      for (int32_t i = 0 ; i < TSDB_MAX_VNODES; i++) {
-        pVnode = &pDnode->vload[i];
-        if (0 == pVnode->vgId) {
-          continue;
-        }
+    void *pNode = NULL;
+    SVgObj *pVgroup;
+    while (1) {
+      pNode = mgmtGetNextVgroup(pNode, &pVgroup);
+      if (pVgroup == NULL) break;
 
+      for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
+        SVnodeGid *pVgid = &pVgroup->vnodeGid[i];
+        if (pVgid->pDnode == pDnode) {
           cols = 0;
 
           pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
-        *(uint32_t *)pWrite = pVnode->vgId;
+          *(uint32_t *)pWrite = pVgroup->vgId;
           cols++;
 
           pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
-        strcpy(pWrite, pVnode->status ? "ready" : "offline");
+          strcpy(pWrite, mgmtGetMnodeRoleStr(pVgid->role));
           cols++;
-        
-        numOfRows++;
         }
       }
+
+      mgmtDecVgroupRef(pVgroup);
+    }
   } else {
-    // TODO: output all vnodes of all dnodes
     numOfRows = 0;
   }
 

src/mnode/src/mgmtMain.c

@@ -149,12 +149,12 @@ void mgmtCleanUpSystem() {
   mgmtCleanUpShell();
   mgmtCleanupDClient();
   mgmtCleanupDServer();
+  mgmtCleanUpAccts();
   mgmtCleanUpTables();
   mgmtCleanUpVgroups();
   mgmtCleanUpDbs();
   mgmtCleanupDnodes();
   mgmtCleanUpUsers();
-  mgmtCleanUpAccts();
   sdbCleanUp();
   taosTmrCleanUp(tsMgmtTmr);
   tsMgmtIsRunning = false;

src/mnode/src/mgmtMnode.c

@@ -30,7 +30,7 @@
 #include "mgmtShell.h"
 #include "mgmtUser.h"
 
-static void *  tsMnodeSdb = NULL;
+void * tsMnodeSdb = NULL;
 static int32_t tsMnodeUpdateSize = 0;
 static int32_t mgmtGetMnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn);
 static int32_t mgmtRetrieveMnodes(SShowObj *pShow, char *data, int32_t rows, void *pConn);
@@ -47,7 +47,7 @@ static int32_t mgmtMnodeActionInsert(SSdbOper *pOper) {
 
   pMnode->pDnode = pDnode;
   pDnode->isMgmt = true;
-  mgmtReleaseDnode(pDnode);
+  mgmtDecDnodeRef(pDnode);
   
   return TSDB_CODE_SUCCESS;
 }
@@ -58,7 +58,7 @@ static int32_t mgmtMnodeActionDelete(SSdbOper *pOper) {
   SDnodeObj *pDnode = mgmtGetDnode(pMnode->mnodeId);
   if (pDnode == NULL) return TSDB_CODE_DNODE_NOT_EXIST;
   pDnode->isMgmt = false;
-  mgmtReleaseDnode(pDnode);
+  mgmtDecDnodeRef(pDnode);
 
   mTrace("mnode:%d, is dropped from sdb", pMnode->mnodeId);
   return TSDB_CODE_SUCCESS;
@@ -171,7 +171,7 @@ char *mgmtGetMnodeRoleStr(int32_t role) {
   }
 }
 
-void mgmtGetMnodeIpList(SRpcIpSet *ipSet, bool usePublicIp) {
+void mgmtGetMnodeIpSet(SRpcIpSet *ipSet, bool usePublicIp) {
   void *pNode = NULL;
   while (1) {
     SMnodeObj *pMnode = NULL;
@@ -268,7 +268,10 @@ static int32_t mgmtGetMnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) return 0;
 
-  if (strcmp(pUser->pAcct->user, "root") != 0) return TSDB_CODE_NO_RIGHTS;
+  if (strcmp(pUser->pAcct->user, "root") != 0)  {
+    mgmtDecUserRef(pUser);
+    return TSDB_CODE_NO_RIGHTS;
+  }
 
   int32_t  cols = 0;
   SSchema *pSchema = pMeta->schema;
@@ -314,7 +317,7 @@ static int32_t mgmtGetMnodeMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCo
   pShow->numOfRows = mgmtGetMnodesNum();
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
   pShow->pNode = NULL;
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 
   return 0;
 }

src/mnode/src/mgmtProfile.c

@@ -704,6 +704,7 @@ void mgmtProcessKillQueryMsg(SQueuedMsg *pMsg) {
 
   rpcRsp.code = code;
   rpcSendResponse(&rpcRsp);
+  mgmtDecUserRef(pUser);
 }
 
 void mgmtProcessKillStreamMsg(SQueuedMsg *pMsg) {
@@ -727,6 +728,7 @@ void mgmtProcessKillStreamMsg(SQueuedMsg *pMsg) {
 
   rpcRsp.code = code;
   rpcSendResponse(&rpcRsp);
+  mgmtDecUserRef(pUser);
 }
 
 void mgmtProcessKillConnectionMsg(SQueuedMsg *pMsg) {
@@ -750,6 +752,7 @@ void mgmtProcessKillConnectionMsg(SQueuedMsg *pMsg) {
 
   rpcRsp.code = code;
   rpcSendResponse(&rpcRsp);
+  mgmtDecUserRef(pUser);
 }
 
 int32_t mgmtInitProfile() {
@@ -790,12 +793,12 @@ void *mgmtMallocQueuedMsg(SRpcMsg *rpcMsg) {
 void mgmtFreeQueuedMsg(SQueuedMsg *pMsg) {
   if (pMsg != NULL) {
     rpcFreeCont(pMsg->pCont);
-    if (pMsg->pUser) mgmtReleaseUser(pMsg->pUser);
+    if (pMsg->pUser) mgmtDecUserRef(pMsg->pUser);
     if (pMsg->pDb) mgmtDecDbRef(pMsg->pDb);
-    if (pMsg->pVgroup) mgmtReleaseVgroup(pMsg->pVgroup);
+    if (pMsg->pVgroup) mgmtDecVgroupRef(pMsg->pVgroup);
     if (pMsg->pTable) mgmtDecTableRef(pMsg->pTable);
     if (pMsg->pAcct) mgmtDecAcctRef(pMsg->pAcct);
-    if (pMsg->pDnode) mgmtReleaseDnode(pMsg->pDnode);
+    if (pMsg->pDnode) mgmtDecDnodeRef(pMsg->pDnode);
     free(pMsg);
   }
 }

src/mnode/src/mgmtShell.c

@@ -141,13 +141,19 @@ void mgmtDealyedAddToShellQueue(SQueuedMsg *queuedMsg) {
 
 static void mgmtProcessMsgFromShell(SRpcMsg *rpcMsg) {
   if (rpcMsg == NULL || rpcMsg->pCont == NULL) {
+    mgmtSendSimpleResp(rpcMsg->handle, TSDB_CODE_INVALID_MSG_LEN);
     return;
   }
 
   if (!sdbIsMaster()) {
-    // rpcSendRedirectRsp(rpcMsg->handle, mgmtGetMnodeIpListForRedirect());
-    mgmtSendSimpleResp(rpcMsg->handle, TSDB_CODE_NO_MASTER);
-    rpcFreeCont(rpcMsg->pCont);
+    SRpcConnInfo connInfo;
+    rpcGetConnInfo(rpcMsg->handle, &connInfo);
+    bool usePublicIp = (connInfo.serverIp == tsPublicIpInt);
+    
+    SRpcIpSet ipSet = {0};
+    mgmtGetMnodeIpSet(&ipSet, usePublicIp);
+    mTrace("conn from ip:%s user:%s redirect msg", taosIpStr(connInfo.clientIp), connInfo.user);
+    rpcSendRedirectRsp(rpcMsg->handle, &ipSet);
     return;
   }
 
@@ -332,7 +338,7 @@ static void mgmtProcessHeartBeatMsg(SQueuedMsg *pMsg) {
     return;
   }
 
-  mgmtGetMnodeIpList(&pHBRsp->ipList, pMsg->usePublicIp);
+  mgmtGetMnodeIpSet(&pHBRsp->ipList, pMsg->usePublicIp);
   
   /*
    * TODO
@@ -357,14 +363,18 @@ static int mgmtShellRetriveAuth(char *user, char *spi, char *encrypt, char *secr
   *encrypt = 0;
   *ckey = 0;
 
+  if (!sdbIsMaster()) {
+    *secret = 0;
+    return TSDB_CODE_SUCCESS;
+  }
+
   SUserObj *pUser = mgmtGetUser(user);
   if (pUser == NULL) {
     *secret = 0;
-    mgmtReleaseUser(pUser);
     return TSDB_CODE_INVALID_USER;
   } else {
     memcpy(secret, pUser->pass, TSDB_KEY_LEN);
-    mgmtReleaseUser(pUser);
+    mgmtDecUserRef(pUser);
     return TSDB_CODE_SUCCESS;
   }
 }
@@ -414,7 +424,7 @@ static void mgmtProcessConnectMsg(SQueuedMsg *pMsg) {
   pConnectRsp->writeAuth = pUser->writeAuth;
   pConnectRsp->superAuth = pUser->superAuth;
 
-  mgmtGetMnodeIpList(&pConnectRsp->ipList, pMsg->usePublicIp);
+  mgmtGetMnodeIpSet(&pConnectRsp->ipList, pMsg->usePublicIp);
   
 connect_over:
   rpcRsp.code = code;

src/mnode/src/mgmtTable.c

@@ -97,7 +97,7 @@ static int32_t mgmtChildTableActionInsert(SSdbOper *pOper) {
     mError("ctable:%s, not in vgroup:%d", pTable->info.tableId, pTable->vgId);
     return TSDB_CODE_INVALID_VGROUP_ID;
   }
-  mgmtReleaseVgroup(pVgroup);
+  mgmtDecVgroupRef(pVgroup);
 
   SDbObj *pDb = mgmtGetDb(pVgroup->dbName);
   if (pDb == NULL) {
@@ -108,7 +108,7 @@ static int32_t mgmtChildTableActionInsert(SSdbOper *pOper) {
 
   SAcctObj *pAcct = mgmtGetAcct(pDb->cfg.acct);
   if (pAcct == NULL) {
-    mError("ctable:%s, account:%s not exists", pTable->info.tableId, pDb->cfg.acct);
+    mError("ctable:%s, acct:%s not exists", pTable->info.tableId, pDb->cfg.acct);
     return TSDB_CODE_INVALID_ACCT;
   }
   mgmtDecAcctRef(pAcct);
@@ -139,7 +139,7 @@ static int32_t mgmtChildTableActionDelete(SSdbOper *pOper) {
   if (pVgroup == NULL) {
     return TSDB_CODE_INVALID_VGROUP_ID;
   }
-  mgmtReleaseVgroup(pVgroup);
+  mgmtDecVgroupRef(pVgroup);
 
   SDbObj *pDb = mgmtGetDb(pVgroup->dbName);
   if (pDb == NULL) {
@@ -150,7 +150,7 @@ static int32_t mgmtChildTableActionDelete(SSdbOper *pOper) {
 
   SAcctObj *pAcct = mgmtGetAcct(pDb->cfg.acct);
   if (pAcct == NULL) {
-    mError("ctable:%s, account:%s not exists", pTable->info.tableId, pDb->cfg.acct);
+    mError("ctable:%s, acct:%s not exists", pTable->info.tableId, pDb->cfg.acct);
     return TSDB_CODE_INVALID_ACCT;
   }
   mgmtDecAcctRef(pAcct);
@@ -275,7 +275,7 @@ static int32_t mgmtChildTableActionRestored() {
       pNode = pLastNode;
       continue;
     }
-    mgmtReleaseVgroup(pVgroup);
+    mgmtDecVgroupRef(pVgroup);
 
     if (strcmp(pVgroup->dbName, pDb->name) != 0) {
       mError("ctable:%s, db:%s not match with vgroup:%d db:%s sid:%d, discard it",
@@ -1194,17 +1194,15 @@ static void mgmtProcessSuperTableVgroupMsg(SQueuedMsg *pMsg) {
 
     pRsp->vgroups[vg].vgId = htonl(vgId);
     for (int32_t vn = 0; vn < pVgroup->numOfVnodes; ++vn) {
-      SDnodeObj *pDnode = mgmtGetDnode(pVgroup->vnodeGid[vn].dnodeId);
+      SDnodeObj *pDnode = pVgroup->vnodeGid[vn].pDnode;
       if (pDnode == NULL) break;
 
       pRsp->vgroups[vg].ipAddr[vn].ip = htonl(pDnode->privateIp);
       pRsp->vgroups[vg].ipAddr[vn].port = htons(tsDnodeShellPort);
       pRsp->vgroups[vg].numOfIps++;
-
-      mgmtReleaseDnode(pDnode);
     }
 
-    mgmtReleaseVgroup(pVgroup);
+    mgmtDecVgroupRef(pVgroup);
   }
   pRsp->numOfVgroups = htonl(vg);
 
@@ -1613,7 +1611,7 @@ static int32_t mgmtDoGetChildTableMeta(SQueuedMsg *pMsg, STableMetaMsg *pMeta) {
       pMeta->vgroup.ipAddr[i].port = htonl(tsDnodeShellPort);
     }
     pMeta->vgroup.numOfIps++;
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
   pMeta->vgroup.vgId = htonl(pVgroup->vgId);
 
@@ -1742,7 +1740,7 @@ static SChildTableObj* mgmtGetTableByPos(uint32_t dnodeId, int32_t vnode, int32_
 
   SChildTableObj *pTable = pVgroup->tableList[sid];
   mgmtIncTableRef((STableObj *)pTable);
-  mgmtReleaseVgroup(pVgroup);
+  mgmtDecVgroupRef(pVgroup);
   return pTable;
 }
 

src/mnode/src/mgmtUser.c

@@ -19,17 +19,17 @@
 #include "ttime.h"
 #include "tutil.h"
 #include "tglobal.h"
+#include "tgrant.h"
 #include "dnode.h"
 #include "mgmtDef.h"
 #include "mgmtLog.h"
 #include "mgmtAcct.h"
-#include "tgrant.h"
 #include "mgmtMnode.h"
 #include "mgmtSdb.h"
 #include "mgmtShell.h"
 #include "mgmtUser.h"
 
-void *  tsUserSdb = NULL;
+static void *  tsUserSdb = NULL;
 static int32_t tsUserUpdateSize = 0;
 static int32_t mgmtGetUserMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn);
 static int32_t mgmtRetrieveUsers(SShowObj *pShow, char *data, int32_t rows, void *pConn);
@@ -48,8 +48,8 @@ static int32_t mgmtUserActionInsert(SSdbOper *pOper) {
 
   if (pAcct != NULL) {
     mgmtAddUserToAcct(pAcct, pUser);
-  }
-  else {
+    mgmtDecAcctRef(pAcct);
+  } else {
     mError("user:%s, acct:%s info not exist in sdb", pUser->user, pUser->acct);
     return TSDB_CODE_INVALID_ACCT;
   }
@@ -63,6 +63,7 @@ static int32_t mgmtUserActionDelete(SSdbOper *pOper) {
 
   if (pAcct != NULL) {
     mgmtDropUserFromAcct(pAcct, pUser);
+    mgmtDecAcctRef(pAcct);
   }
 
   return TSDB_CODE_SUCCESS;
@@ -72,9 +73,10 @@ static int32_t mgmtUserActionUpdate(SSdbOper *pOper) {
   SUserObj *pUser = pOper->pObj;
   SUserObj *pSaved = mgmtGetUser(pUser->user);
   if (pUser != pSaved) {
-    memcpy(pSaved, pUser, pOper->rowSize);
+    memcpy(pSaved, pUser, tsUserUpdateSize);
     free(pUser);
   }
+  mgmtDecUserRef(pSaved);
   return TSDB_CODE_SUCCESS;
 }
 
@@ -86,7 +88,7 @@ static int32_t mgmtUserActionEncode(SSdbOper *pOper) {
 }
 
 static int32_t mgmtUserActionDecode(SSdbOper *pOper) {
-  SUserObj *pUser = (SUserObj *) calloc(1, sizeof(SUserObj));
+  SUserObj *pUser = (SUserObj *)calloc(1, sizeof(SUserObj));
   if (pUser == NULL) return TSDB_CODE_SERV_OUT_OF_MEMORY;
 
   memcpy(pUser, pOper->rowData, tsUserUpdateSize);
@@ -103,7 +105,7 @@ static int32_t mgmtUserActionRestored() {
     mgmtDecAcctRef(pAcct);
   }
 
-  return 0;
+  return TSDB_CODE_SUCCESS;
 }
 
 int32_t mgmtInitUsers() {
@@ -128,7 +130,7 @@ int32_t mgmtInitUsers() {
 
   tsUserSdb = sdbOpenTable(&tableDesc);
   if (tsUserSdb == NULL) {
-    mError("failed to init user data");
+    mError("table:%s, failed to create hash", tableDesc.tableName);
     return -1;
   }
 
@@ -138,7 +140,7 @@ int32_t mgmtInitUsers() {
   mgmtAddShellShowMetaHandle(TSDB_MGMT_TABLE_USER, mgmtGetUserMeta);
   mgmtAddShellShowRetrieveHandle(TSDB_MGMT_TABLE_USER, mgmtRetrieveUsers);
   
-  mTrace("table:users table is created");
+  mTrace("table:%s, hash is created", tableDesc.tableName);
   return 0;
 }
 
@@ -150,7 +152,15 @@ SUserObj *mgmtGetUser(char *name) {
   return (SUserObj *)sdbGetRow(tsUserSdb, name);
 }
 
-void mgmtReleaseUser(SUserObj *pUser) { 
+void *mgmtGetNextUser(void *pNode, SUserObj **pUser) { 
+  return sdbFetchRow(tsUserSdb, pNode, (void **)pUser); 
+}
+
+void mgmtIncUserRef(SUserObj *pUser) { 
+  return sdbIncRef(tsUserSdb, pUser); 
+}
+
+void mgmtDecUserRef(SUserObj *pUser) { 
   return sdbDecRef(tsUserSdb, pUser); 
 }
 
@@ -172,18 +182,22 @@ static int32_t mgmtUpdateUser(SUserObj *pUser) {
 
 int32_t mgmtCreateUser(SAcctObj *pAcct, char *name, char *pass) {
   int32_t code = acctCheck(pAcct, ACCT_GRANT_USER);
-  if (code != 0) {
+  if (code != TSDB_CODE_SUCCESS) {
     return code;
   }
 
-  if (name[0] == 0 || pass[0] == 0) {
-    return TSDB_CODE_INVALID_MSG_CONTENT;
+  if (name[0] == 0) {
+    return TSDB_CODE_INVALID_USER_FORMAT;
+  }
+
+  if (pass[0] == 0) {
+    return TSDB_CODE_INVALID_PASS_FORMAT;
   }
 
   SUserObj *pUser = mgmtGetUser(name);
   if (pUser != NULL) {
-    mTrace("user:%s is already there", name);
-    mgmtReleaseUser(pUser);
+    mTrace("user:%s, is already there", name);
+    mgmtDecUserRef(pUser);
     return TSDB_CODE_USER_ALREADY_EXIST;
   }
 
@@ -237,7 +251,7 @@ static int32_t mgmtDropUser(SUserObj *pUser) {
 static int32_t mgmtGetUserMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn) {
   SUserObj *pUser = mgmtGetUserFromConn(pConn, NULL);
   if (pUser == NULL) {
-    return TSDB_CODE_INVALID_USER;
+    return TSDB_CODE_NO_USER_FROM_CONN;
   }
 
   int32_t cols = 0;
@@ -273,7 +287,7 @@ static int32_t mgmtGetUserMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pCon
   pShow->numOfRows = pUser->pAcct->acctInfo.numOfUsers;
   pShow->rowSize = pShow->offset[cols - 1] + pShow->bytes[cols - 1];
 
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
   return 0;
 }
 
@@ -308,8 +322,9 @@ static int32_t mgmtRetrieveUsers(SShowObj *pShow, char *data, int32_t rows, void
     cols++;
 
     numOfRows++;
-    mgmtReleaseUser(pUser);
+    mgmtDecUserRef(pUser);
   }
+
   pShow->numOfReads += numOfRows;
   return numOfRows;
 }
@@ -321,20 +336,21 @@ SUserObj *mgmtGetUserFromConn(void *pConn, bool *usePublicIp) {
       *usePublicIp = (connInfo.serverIp == tsPublicIpInt);
     }
     return mgmtGetUser(connInfo.user);
-  }
-
+  } else {
+    mError("can not get user from conn:%p", pConn);
     return NULL;
+  }
 }
 
 static void mgmtProcessCreateUserMsg(SQueuedMsg *pMsg) {
   int32_t code;
-  SUserObj *pUser = pMsg->pUser;
+  SUserObj *pOperUser = pMsg->pUser;
   
-  if (pUser->superAuth) {
+  if (pOperUser->superAuth) {
     SCMCreateUserMsg *pCreate = pMsg->pCont;
-    code = mgmtCreateUser(pUser->pAcct, pCreate->user, pCreate->pass);
+    code = mgmtCreateUser(pOperUser->pAcct, pCreate->user, pCreate->pass);
     if (code == TSDB_CODE_SUCCESS) {
-      mLPrint("user:%s is created by %s", pCreate->user, pUser->user);
+      mLPrint("user:%s, is created by %s", pCreate->user, pOperUser->user);
     }
   } else {
     code = TSDB_CODE_NO_RIGHTS;
@@ -356,7 +372,7 @@ static void mgmtProcessAlterUserMsg(SQueuedMsg *pMsg) {
 
   if (strcmp(pUser->user, "monitor") == 0 || (strcmp(pUser->user + 1, pUser->acct) == 0 && pUser->user[0] == '_')) {
     mgmtSendSimpleResp(pMsg->thandle, TSDB_CODE_NO_RIGHTS);
-    mgmtReleaseUser(pUser);
+    mgmtDecUserRef(pUser);
     return;
   }
 
@@ -380,7 +396,7 @@ static void mgmtProcessAlterUserMsg(SQueuedMsg *pMsg) {
       memset(pUser->pass, 0, sizeof(pUser->pass));
       taosEncryptPass((uint8_t*)pAlter->pass, strlen(pAlter->pass), pUser->pass);
       code = mgmtUpdateUser(pUser);
-      mLPrint("user:%s password is altered by %s, result:%d", pUser->user, pOperUser->user, tstrerror(code));
+      mLPrint("user:%s, password is altered by %s, result:%s", pUser->user, pOperUser->user, tstrerror(code));
     } else {
       code = TSDB_CODE_NO_RIGHTS;
     }
@@ -422,7 +438,7 @@ static void mgmtProcessAlterUserMsg(SQueuedMsg *pMsg) {
       }
 
       code = mgmtUpdateUser(pUser);
-      mLPrint("user:%s privilege is altered by %s, result:%d", pUser->user, pOperUser->user, tstrerror(code));
+      mLPrint("user:%s, privilege is altered by %s, result:%s", pUser->user, pOperUser->user, tstrerror(code));
     } else {
       code = TSDB_CODE_NO_RIGHTS;
     }
@@ -432,7 +448,7 @@ static void mgmtProcessAlterUserMsg(SQueuedMsg *pMsg) {
     mgmtSendSimpleResp(pMsg->thandle, TSDB_CODE_NO_RIGHTS);
   }
 
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 }
 
 static void mgmtProcessDropUserMsg(SQueuedMsg *pMsg) {
@@ -443,13 +459,13 @@ static void mgmtProcessDropUserMsg(SQueuedMsg *pMsg) {
   SUserObj *pUser = mgmtGetUser(pDrop->user);
   if (pUser == NULL) {
     mgmtSendSimpleResp(pMsg->thandle, TSDB_CODE_INVALID_USER);
-    return ;
+    return;
   }
 
   if (strcmp(pUser->user, "monitor") == 0 || strcmp(pUser->user, pUser->acct) == 0 ||
     (strcmp(pUser->user + 1, pUser->acct) == 0 && pUser->user[0] == '_')) {
     mgmtSendSimpleResp(pMsg->thandle, TSDB_CODE_NO_RIGHTS);
-    mgmtReleaseUser(pUser);
+    mgmtDecUserRef(pUser);
     return ;
   }
 
@@ -471,14 +487,14 @@ static void mgmtProcessDropUserMsg(SQueuedMsg *pMsg) {
   if (hasRight) {
     code = mgmtDropUser(pUser);
     if (code == TSDB_CODE_SUCCESS) {
-       mLPrint("user:%s is dropped by %s, result:%s", pUser->user, pOperUser->user, tstrerror(code));
+      mLPrint("user:%s, is dropped by %s, result:%s", pUser->user, pOperUser->user, tstrerror(code));
     }
   } else {
     code = TSDB_CODE_NO_RIGHTS;
   }
 
   mgmtSendSimpleResp(pMsg->thandle, code);
-  mgmtReleaseUser(pUser);
+  mgmtDecUserRef(pUser);
 }
 
 void  mgmtDropAllUsers(SAcctObj *pAcct)  {
@@ -504,7 +520,7 @@ void  mgmtDropAllUsers(SAcctObj *pAcct)  {
       numOfUsers++;
     }
 
-    mgmtReleaseUser(pUser);
+    mgmtDecUserRef(pUser);
   }
 
   mTrace("acct:%s, all users:%d is dropped from sdb", pAcct->user, numOfUsers);

src/mnode/src/mgmtVgroup.c

@@ -44,9 +44,7 @@ static int32_t mgmtRetrieveVgroups(SShowObj *pShow, char *data, int32_t rows, vo
 static void    mgmtProcessCreateVnodeRsp(SRpcMsg *rpcMsg);
 static void    mgmtProcessDropVnodeRsp(SRpcMsg *rpcMsg);
 static void    mgmtProcessVnodeCfgMsg(SRpcMsg *rpcMsg) ;
-
 static void    mgmtSendDropVgroupMsg(SVgObj *pVgroup, void *ahandle);
-static void mgmtSendCreateVgroupMsg(SVgObj *pVgroup, void *ahandle);
 
 static int32_t mgmtVgroupActionDestroy(SSdbOper *pOper) {
   SVgObj *pVgroup = pOper->pObj;
@@ -68,7 +66,6 @@ static int32_t mgmtVgroupActionInsert(SSdbOper *pOper) {
   if (pDb == NULL) {
     return TSDB_CODE_INVALID_DB;
   }
-  mgmtDecDbRef(pDb);
 
   pVgroup->pDb = pDb;
   pVgroup->prev = NULL;
@@ -91,15 +88,13 @@ static int32_t mgmtVgroupActionInsert(SSdbOper *pOper) {
   for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
     SDnodeObj *pDnode = mgmtGetDnode(pVgroup->vnodeGid[i].dnodeId);
     if (pDnode != NULL) {
-      pVgroup->vnodeGid[i].privateIp = pDnode->privateIp;
-      pVgroup->vnodeGid[i].publicIp = pDnode->publicIp;
+      pVgroup->vnodeGid[i].pDnode = pDnode;
       atomic_add_fetch_32(&pDnode->openVnodes, 1);
-      mgmtReleaseDnode(pDnode);
+      mgmtDecDnodeRef(pDnode);
     }
   }
 
   mgmtAddVgroupIntoDb(pVgroup);
-  mgmtIncDbRef(pVgroup->pDb);
 
   return TSDB_CODE_SUCCESS;
 }
@@ -115,10 +110,10 @@ static int32_t mgmtVgroupActionDelete(SSdbOper *pOper) {
 
   for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
     SDnodeObj *pDnode = mgmtGetDnode(pVgroup->vnodeGid[i].dnodeId);
-    if (pDnode) {
+    if (pDnode != NULL) {
       atomic_sub_fetch_32(&pDnode->openVnodes, 1);
     }
-    mgmtReleaseDnode(pDnode);
+    mgmtDecDnodeRef(pDnode);
   }
 
   return TSDB_CODE_SUCCESS;
@@ -127,9 +122,25 @@ static int32_t mgmtVgroupActionDelete(SSdbOper *pOper) {
 static int32_t mgmtVgroupActionUpdate(SSdbOper *pOper) {
   SVgObj *pNew = pOper->pObj;
   SVgObj *pVgroup = mgmtGetVgroup(pNew->vgId);
+
   if (pVgroup != pNew) {
+    for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
+      SDnodeObj *pDnode = pVgroup->vnodeGid[i].pDnode;
+      if (pDnode != NULL) {
+        atomic_sub_fetch_32(&pDnode->openVnodes, 1);
+      }
+    }
+
     memcpy(pVgroup, pNew, pOper->rowSize);
     free(pNew);
+
+    for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
+      SDnodeObj *pDnode = mgmtGetDnode(pVgroup->vnodeGid[i].dnodeId);
+      pVgroup->vnodeGid[i].pDnode = pDnode;
+      if (pDnode != NULL) {
+        atomic_add_fetch_32(&pDnode->openVnodes, 1);
+      }
+    }
   }
 
   int32_t oldTables = taosIdPoolMaxSize(pVgroup->idPool);
@@ -150,6 +161,12 @@ static int32_t mgmtVgroupActionUpdate(SSdbOper *pOper) {
 static int32_t mgmtVgroupActionEncode(SSdbOper *pOper) {
   SVgObj *pVgroup = pOper->pObj;
   memcpy(pOper->rowData, pVgroup, tsVgUpdateSize);
+  SVgObj *pTmpVgroup = pOper->rowData;
+  for (int32_t i = 0; i < TSDB_VNODES_SUPPORT; ++i) {
+    pTmpVgroup->vnodeGid[i].pDnode = NULL;
+    pTmpVgroup->vnodeGid[i].role = 0;
+  }
+
   pOper->rowSize = tsVgUpdateSize;
   return TSDB_CODE_SUCCESS;
 }
@@ -204,7 +221,11 @@ int32_t mgmtInitVgroups() {
   return 0;
 }
 
-void mgmtReleaseVgroup(SVgObj *pVgroup) { 
+void mgmtIncVgroupRef(SVgObj *pVgroup) {
+  return sdbIncRef(tsVgroupSdb, pVgroup); 
+}
+
+void mgmtDecVgroupRef(SVgObj *pVgroup) { 
   return sdbDecRef(tsVgroupSdb, pVgroup); 
 }
 
@@ -224,15 +245,37 @@ void mgmtUpdateVgroup(SVgObj *pVgroup) {
   mgmtSendCreateVgroupMsg(pVgroup, NULL);
 }
 
-void mgmtUpdateVgroupStatus(SVgObj *pVgroup, int32_t dnodeId, SVnodeLoad *pVload) {
-  if (pVload->role == TAOS_SYNC_ROLE_MASTER) {
+void mgmtUpdateVgroupStatus(SVgObj *pVgroup, SDnodeObj *pDnode, SVnodeLoad *pVload) {
+  bool dnodeExist = false;
   for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
     SVnodeGid *pVgid = &pVgroup->vnodeGid[i];
-      if (pVgid->dnodeId == dnodeId) {
+    if (pVgid->pDnode == pDnode) {
+      pVgid->role = pVload->role;
+      if (pVload->role == TAOS_SYNC_ROLE_MASTER) {
         pVgroup->inUse = i;
+      }
+      dnodeExist = true;
       break;
     }
   }
+
+  if (!dnodeExist) {
+    SRpcIpSet ipSet = mgmtGetIpSetFromIp(pDnode->privateIp);
+    mError("vgroup:%d, dnode:%d not exist in mnode, drop it", pVload->vgId, pDnode->dnodeId);
+    mgmtSendDropVnodeMsg(pVload->vgId, &ipSet, NULL);
+    return;
+  }
+
+  if (pVload->role == TAOS_SYNC_ROLE_MASTER) {
+    pVgroup->totalStorage = htobe64(pVload->totalStorage);
+    pVgroup->compStorage = htobe64(pVload->compStorage);
+    pVgroup->pointsWritten = htobe64(pVload->pointsWritten);
+  }
+
+  if (pVload->replica != pVgroup->numOfVnodes) {
+    mError("dnode:%d, vgroup:%d replica:%d not match with mgmt:%d", pDnode->dnodeId, pVload->vgId, pVload->replica,
+           pVgroup->numOfVnodes);
+    mgmtSendCreateVgroupMsg(pVgroup, NULL);
   }
 }
 
@@ -340,7 +383,7 @@ int32_t mgmtGetVgroupMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn) {
     mgmtDecTableRef(pTable);
     pVgroup = mgmtGetVgroup(((SChildTableObj*)pTable)->vgId);
     if (NULL == pVgroup) return TSDB_CODE_INVALID_TABLE_ID;
-    mgmtReleaseVgroup(pVgroup);
+    mgmtDecVgroupRef(pVgroup);
     maxReplica = pVgroup->numOfVnodes > maxReplica ? pVgroup->numOfVnodes : maxReplica;
   } else {
     SVgObj *pVgroup = pDb->pHead;
@@ -391,27 +434,6 @@ int32_t mgmtGetVgroupMeta(STableMetaMsg *pMeta, SShowObj *pShow, void *pConn) {
   return 0;
 }
 
-char *mgmtGetVnodeStatus(SVgObj *pVgroup, SVnodeGid *pVnode) {
-  SDnodeObj *pDnode = mgmtGetDnode(pVnode->dnodeId);
-  if (pDnode == NULL) {
-    mError("vgroup:%d, not exist in dnode:%d", pVgroup->vgId, pDnode->dnodeId);
-    return "null";
-  }
-  mgmtReleaseDnode(pDnode);
-
-  if (pDnode->status == TAOS_DN_STATUS_OFFLINE) {
-    return "offline";
-  }
-
-  for (int i = 0; i < pDnode->openVnodes; ++i) {
-    if (pDnode->vload[i].vgId == pVgroup->vgId) {
-       return pDnode->vload[i].status ? "ready" : "offline";
-    }
-  }
-  
-  return "null";
-}
-
 int32_t mgmtRetrieveVgroups(SShowObj *pShow, char *data, int32_t rows, void *pConn) {
   int32_t numOfRows = 0;
   SVgObj *pVgroup = NULL;
@@ -453,20 +475,25 @@ int32_t mgmtRetrieveVgroups(SShowObj *pShow, char *data, int32_t rows, void *pCo
       *(int16_t *) pWrite = pVgroup->vnodeGid[i].dnodeId;
       cols++;
 
-      tinet_ntoa(ipstr, pVgroup->vnodeGid[i].privateIp);
+      SDnodeObj *pDnode = pVgroup->vnodeGid[i].pDnode;
+
+      if (pDnode != NULL) {
+        tinet_ntoa(ipstr, pDnode->privateIp);
         pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
         strcpy(pWrite, ipstr);
         cols++;
-
         pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
-      if (pVgroup->vnodeGid[i].dnodeId != 0) {
-        char *vnodeStatus = mgmtGetVnodeStatus(pVgroup, pVgroup->vnodeGid + i);
-        strcpy(pWrite, vnodeStatus);
+        strcpy(pWrite, mgmtGetMnodeRoleStr(pVgroup->vnodeGid[i].role));
+        cols++;
       } else {
+        pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
+        strcpy(pWrite, "null");
+        cols++;
+        pWrite = data + pShow->offset[cols] * rows + pShow->bytes[cols] * numOfRows;
         strcpy(pWrite, "null");
-      }
         cols++;
       }
+    }
 
     numOfRows++;
   }
@@ -506,27 +533,38 @@ SMDCreateVnodeMsg *mgmtBuildCreateVnodeMsg(SVgObj *pVgroup) {
   SMDCreateVnodeMsg *pVnode = rpcMallocCont(sizeof(SMDCreateVnodeMsg));
   if (pVnode == NULL) return NULL;
 
-  pVnode->cfg = pDb->cfg;
-
-  SVnodeCfg *pCfg = &pVnode->cfg;
+  SMDVnodeCfg *pCfg = &pVnode->cfg;
   pCfg->vgId                = htonl(pVgroup->vgId);
-  pCfg->maxSessions                  = htonl(pCfg->maxSessions);
-  pCfg->cacheBlockSize               = htonl(pCfg->cacheBlockSize);
-  pCfg->cacheNumOfBlocks.totalBlocks = htonl(pCfg->cacheNumOfBlocks.totalBlocks);
-  pCfg->daysPerFile                  = htonl(pCfg->daysPerFile);
-  pCfg->daysToKeep1                  = htonl(pCfg->daysToKeep1);
-  pCfg->daysToKeep2                  = htonl(pCfg->daysToKeep2);
-  pCfg->daysToKeep                   = htonl(pCfg->daysToKeep);
-  pCfg->commitTime                   = htonl(pCfg->commitTime);
-  pCfg->rowsInFileBlock              = htonl(pCfg->rowsInFileBlock);
-  pCfg->blocksPerTable               = htons(pCfg->blocksPerTable);
+  pCfg->maxTables           = htonl(pDb->cfg.maxSessions);
+  pCfg->maxCacheSize        = htobe64((int64_t)pDb->cfg.cacheBlockSize * pDb->cfg.cacheNumOfBlocks.totalBlocks);
+  pCfg->maxCacheSize        = htobe64(-1);
+  pCfg->minRowsPerFileBlock = htonl(-1);
+  pCfg->maxRowsPerFileBlock = htonl(-1);
+  pCfg->daysPerFile         = htonl(pDb->cfg.daysPerFile);
+  pCfg->daysToKeep1         = htonl(pDb->cfg.daysToKeep1);
+  pCfg->daysToKeep2         = htonl(pDb->cfg.daysToKeep2);
+  pCfg->daysToKeep          = htonl(pDb->cfg.daysToKeep);
+  pCfg->daysToKeep          = htonl(-1);
+  pCfg->commitTime          = htonl(pDb->cfg.commitTime);
+  pCfg->precision           = pDb->cfg.precision;
+  pCfg->compression         = pDb->cfg.compression;
+  pCfg->compression         = -1;
+  pCfg->wals                = 3;
+  pCfg->commitLog           = pDb->cfg.commitLog;
   pCfg->replications        = (int8_t) pVgroup->numOfVnodes;
+  pCfg->quorum              = 1;
   
-  SVnodeDesc *vpeerDesc = pVnode->vpeerDesc;
+  SMDVnodeDesc *pNodes = pVnode->nodes;
   for (int32_t j = 0; j < pVgroup->numOfVnodes; ++j) {
-    vpeerDesc[j].vgId    = htonl(pVgroup->vgId);
-    vpeerDesc[j].dnodeId = htonl(pVgroup->vnodeGid[j].dnodeId);
-    vpeerDesc[j].ip      = htonl(pVgroup->vnodeGid[j].privateIp);
+    SDnodeObj *pDnode = pVgroup->vnodeGid[j].pDnode;
+    if (pDnode != NULL) {
+      pNodes[j].nodeId = htonl(pDnode->dnodeId);
+      pNodes[j].nodeIp = htonl(pDnode->privateIp);
+      strcpy(pNodes[j].nodeName, pDnode->dnodeName);
+      if (j == 0) {
+        pCfg->arbitratorIp = htonl(pDnode->privateIp);
+      }
+    }
   }
 
   return pVnode;
@@ -539,7 +577,7 @@ SRpcIpSet mgmtGetIpSetFromVgroup(SVgObj *pVgroup) {
     .port = tsDnodeMnodePort
   };
   for (int i = 0; i < pVgroup->numOfVnodes; ++i) {
-    ipSet.ip[i] = pVgroup->vnodeGid[i].privateIp;
+    ipSet.ip[i] = pVgroup->vnodeGid[i].pDnode->privateIp;
   }
   return ipSet;
 }
@@ -570,7 +608,7 @@ void mgmtSendCreateVnodeMsg(SVgObj *pVgroup, SRpcIpSet *ipSet, void *ahandle) {
 void mgmtSendCreateVgroupMsg(SVgObj *pVgroup, void *ahandle) {
   mTrace("vgroup:%d, send create all vnodes msg, ahandle:%p", pVgroup->vgId, ahandle);
   for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
-    SRpcIpSet ipSet = mgmtGetIpSetFromIp(pVgroup->vnodeGid[i].privateIp);
+    SRpcIpSet ipSet = mgmtGetIpSetFromIp(pVgroup->vnodeGid[i].pDnode->privateIp);
     mgmtSendCreateVnodeMsg(pVgroup, &ipSet, ahandle);
   }
 }
@@ -636,7 +674,7 @@ void mgmtSendDropVnodeMsg(int32_t vgId, SRpcIpSet *ipSet, void *ahandle) {
 static void mgmtSendDropVgroupMsg(SVgObj *pVgroup, void *ahandle) {
   mTrace("vgroup:%d, send drop all vnodes msg, ahandle:%p", pVgroup->vgId, ahandle);
   for (int32_t i = 0; i < pVgroup->numOfVnodes; ++i) {
-    SRpcIpSet ipSet = mgmtGetIpSetFromIp(pVgroup->vnodeGid[i].privateIp);
+    SRpcIpSet ipSet = mgmtGetIpSetFromIp(pVgroup->vnodeGid[i].pDnode->privateIp);
     mgmtSendDropVnodeMsg(pVgroup->vgId, &ipSet, ahandle);
   }
 }
@@ -687,7 +725,7 @@ static void mgmtProcessVnodeCfgMsg(SRpcMsg *rpcMsg) {
     mgmtSendSimpleResp(rpcMsg->handle, TSDB_CODE_NOT_ACTIVE_VNODE);
     return;
   }
-  mgmtReleaseDnode(pDnode);
+  mgmtDecDnodeRef(pDnode);
 
   SVgObj *pVgroup = mgmtGetVgroup(pCfg->vgId);
   if (pVgroup == NULL) {
@@ -695,7 +733,7 @@ static void mgmtProcessVnodeCfgMsg(SRpcMsg *rpcMsg) {
     mgmtSendSimpleResp(rpcMsg->handle, TSDB_CODE_NOT_ACTIVE_VNODE);
     return;
   }
-  mgmtReleaseVgroup(pVgroup);
+  mgmtDecVgroupRef(pVgroup);
 
   mgmtSendSimpleResp(rpcMsg->handle, TSDB_CODE_SUCCESS);
 
@@ -711,7 +749,7 @@ void mgmtDropAllVgroups(SDbObj *pDropDb) {
   SVgObj *pVgroup = NULL;
 
   while (1) {
-    mgmtReleaseVgroup(pVgroup);
+    mgmtDecVgroupRef(pVgroup);
     pNode = sdbFetchRow(tsVgroupSdb, pNode, (void **)&pVgroup);
     if (pVgroup == NULL) break;
 

src/plugins/monitor/src/monitorSystem.c

@@ -397,6 +397,7 @@ void monitorSaveAcctLog(char *acctId, int64_t currentPointsPerSecond, int64_t ma
                         int64_t totalOutbound, int64_t maxOutbound, int64_t totalDbs, int64_t maxDbs,
                         int64_t totalUsers, int64_t maxUsers, int64_t totalStreams, int64_t maxStreams,
                         int64_t totalConns, int64_t maxConns, int8_t accessState) {
+  if (monitor == NULL) return;
   if (monitor->state != MONITOR_STATE_INITIALIZED) return;
 
   char sql[1024] = {0};

src/rpc/src/rpcTcp.c

@@ -466,8 +466,8 @@ static void taosFreeFdObj(SFdObj *pFdObj) {
   }
 
   pFdObj->signature = NULL;
-  close(pFdObj->fd);
   epoll_ctl(pThreadObj->pollFd, EPOLL_CTL_DEL, pFdObj->fd, NULL);
+  close(pFdObj->fd);
 
   pThreadObj->numOfFds--;
 

src/tsdb/src/tsdbMain.c

@@ -11,7 +11,7 @@
 #define IS_VALID_COMPRESSION(compression) (((compression) >= NO_COMPRESSION) && ((compression) <= TWO_STAGE_COMP))
 #define TSDB_MIN_ID 0
 #define TSDB_MAX_ID INT_MAX
-#define TSDB_MIN_TABLES 10
+#define TSDB_MIN_TABLES 4
 #define TSDB_MAX_TABLES 100000
 #define TSDB_DEFAULT_TABLES 1000
 #define TSDB_DEFAULT_DAYS_PER_FILE 10
@@ -283,6 +283,8 @@ int32_t tsdbConfigRepo(TsdbRepoT *repo, STsdbCfg *pCfg) {
 int32_t tsdbTriggerCommit(TsdbRepoT *repo) {
   STsdbRepo *pRepo = (STsdbRepo *)repo;
 
+  if (pRepo->appH.walCallBack) pRepo->appH.walCallBack(pRepo->appH.appH);
+  
   tsdbLockRepo(repo);
   if (pRepo->commit) {
     tsdbUnLockRepo(repo);
@@ -387,7 +389,7 @@ int tsdbInitTableCfg(STableCfg *config, ETableType type, int64_t uid, int32_t ti
   config->superUid = TSDB_INVALID_SUPER_TABLE_ID;
   config->tableId.uid = uid;
   config->tableId.tid = tid;
-  config->name = strdup("test1");
+  config->name = NULL;
   return 0;
 }
 
@@ -854,8 +856,6 @@ static void *tsdbCommitData(void *arg) {
   SRWHelper   whelper = {0};
   if (pCache->imem == NULL) return NULL;
 
-  if (pRepo->appH.walCallBack) pRepo->appH.walCallBack(pRepo->appH.appH);
-
   // Create the iterator to read from cache
   SSkipListIterator **iters = tsdbCreateTableIters(pMeta, pCfg->maxTables);
   if (iters == NULL) {
@@ -880,6 +880,7 @@ static void *tsdbCommitData(void *arg) {
 _exit:
   tdFreeDataCols(pDataCols);
   tsdbDestroyTableIters(iters, pCfg->maxTables);
+  tsdbDestroyHelper(&whelper);
 
   tsdbLockRepo(arg);
   tdListMove(pCache->imem->list, pCache->pool.memPool);

src/tsdb/src/tsdbRWHelper.c

@@ -403,6 +403,7 @@ int tsdbWriteCompInfo(SRWHelper *pHelper) {
   } else {
     pHelper->pCompInfo->delimiter = TSDB_FILE_DELIMITER;
     pHelper->pCompInfo->uid = pHelper->tableInfo.uid;
+    pHelper->pCompInfo->checksum = 0;
     ASSERT((pIdx->len - sizeof(SCompInfo) - sizeof(TSCKSUM)) % sizeof(SCompBlock) == 0);
     taosCalcChecksumAppend(0, (uint8_t *)pHelper->pCompInfo, pIdx->len);
     pIdx->offset = lseek(pHelper->files.nHeadF.fd, 0, SEEK_END);

src/tsdb/tests/tsdbTests.cpp

@@ -4,6 +4,7 @@
 
 #include "tdataformat.h"
 #include "tsdbMain.h"
+#include "tskiplist.h"
 
 static double getCurTime() {
   struct timeval tv;
@@ -141,6 +142,7 @@ TEST(TsdbTest, createRepo) {
   STableCfg tCfg;
   ASSERT_EQ(tsdbInitTableCfg(&tCfg, TSDB_SUPER_TABLE, 987607499877672L, 0), -1);
   ASSERT_EQ(tsdbInitTableCfg(&tCfg, TSDB_NORMAL_TABLE, 987607499877672L, 0), 0);
+  tsdbTableSetName(&tCfg, "test", false);
 
   int       nCols = 5;
   STSchema *schema = tdNewSchema(nCols);
@@ -167,7 +169,7 @@ TEST(TsdbTest, createRepo) {
     .sversion = tCfg.sversion,
     .startTime = 1584081000000,
     .interval = 1000,
-    .totalRows = 5000000,
+    .totalRows = 10000000,
     .rowsPerSubmit = 1,
     .pSchema = schema
   };
@@ -263,3 +265,46 @@ TEST(TsdbTest, DISABLED_createFileGroup) {
 
   int k = 0;
 }
+
+static char *getTKey(const void *data) {
+  return (char *)data;
+}
+
+static void insertSkipList(bool isAscend) {
+  TSKEY start_time = 1587393453000;
+  TSKEY interval = 1000;
+
+  SSkipList *pList = tSkipListCreate(5, TSDB_DATA_TYPE_TIMESTAMP, sizeof(TSKEY), 0, 0, 1, getTKey);
+  ASSERT_NE(pList, nullptr);
+
+  for (size_t i = 0; i < 20000000; i++)
+  {
+    TSKEY time = isAscend ? (start_time + i * interval) : (start_time - i * interval);
+    int32_t level = 0;
+    int32_t headSize = 0;
+
+    tSkipListNewNodeInfo(pList, &level, &headSize);
+    SSkipListNode *pNode = (SSkipListNode *)malloc(headSize + sizeof(TSKEY));
+    ASSERT_NE(pNode, nullptr);
+    pNode->level = level;
+    *(TSKEY *)((char *)pNode + headSize) = time;
+    tSkipListPut(pList, pNode);
+  }
+
+  tSkipListDestroy(pList);
+}
+
+TEST(TsdbTest, DISABLED_testSkipList) {
+// TEST(TsdbTest, testSkipList) {
+  double stime = getCurTime();
+  insertSkipList(true);
+  double etime = getCurTime();
+
+  printf("Time used to insert 100000000 records takes %f seconds\n", etime-stime);
+
+  stime = getCurTime();
+  insertSkipList(false);
+  etime = getCurTime();
+
+  printf("Time used to insert 100000000 records takes %f seconds\n", etime-stime);
+}
\ No newline at end of file

src/util/inc/tutil.h

@@ -170,6 +170,8 @@ char *taosIpStr(uint32_t ipInt);
 
 uint32_t ip2uint(const char *const ip_addr);
 
+void taosRemoveDir(char *rootDir);
+
 #define TAOS_ALLOC_MODE_DEFAULT 0
 #define TAOS_ALLOC_MODE_RANDOM_FAIL 1
 #define TAOS_ALLOC_MODE_DETECT_LEAK 2

src/util/src/tutil.c

@@ -663,3 +663,27 @@ void tzfree(void *ptr) {
     free((void *)((char *)ptr - sizeof(size_t)));
   }
 }
+
+void taosRemoveDir(char *rootDir) {
+  DIR *dir = opendir(rootDir);
+  if (dir == NULL) return;
+
+  struct dirent *de = NULL;
+  while ((de = readdir(dir)) != NULL) {
+    if (strcmp(de->d_name, ".") == 0 || strcmp(de->d_name, "..") == 0) continue;
+     
+    char filename[1024];
+    snprintf(filename, 1023, "%s/%s", rootDir, de->d_name);
+    if (de->d_type & DT_DIR) {
+      taosRemoveDir(filename);
+    } else {
+      remove(filename);
+      uPrint("file:%s is removed", filename);
+    }
+  }
+
+  closedir(dir);
+  rmdir(rootDir);
+
+  uPrint("dir:%s is removed", rootDir);
+}
\ No newline at end of file

src/vnode/CMakeLists.txt

@@ -9,6 +9,7 @@ IF ((TD_LINUX_64) OR (TD_LINUX_32 AND TD_ARM))
   INCLUDE_DIRECTORIES(${TD_COMMUNITY_DIR}/src/tsdb/inc)
   INCLUDE_DIRECTORIES(${TD_COMMUNITY_DIR}/src/common/inc)
   INCLUDE_DIRECTORIES(${TD_COMMUNITY_DIR}/src/dnode/inc)
+  INCLUDE_DIRECTORIES(${TD_COMMUNITY_DIR}/deps/cJson/inc)
   INCLUDE_DIRECTORIES(${TD_ENTERPRISE_DIR}/src/inc)
   INCLUDE_DIRECTORIES(inc)
   AUX_SOURCE_DIRECTORY(src SRC)

src/vnode/src/vnodeMain.c

@@ -18,10 +18,12 @@
 #include "ihash.h"
 #include "taoserror.h"
 #include "taosmsg.h"
+#include "tutil.h"
 #include "trpc.h"
 #include "tsdb.h"
 #include "ttime.h"
 #include "ttimer.h"
+#include "cJSON.h"
 #include "twal.h"
 #include "tglobal.h"
 #include "dnode.h"
@@ -36,6 +38,8 @@ static void     vnodeBuildVloadMsg(char *pNode, void * param);
 static int      vnodeWalCallback(void *arg);
 static int32_t  vnodeSaveCfg(SMDCreateVnodeMsg *pVnodeCfg);
 static int32_t  vnodeReadCfg(SVnodeObj *pVnode);
+static int32_t  vnodeSaveVersion(SVnodeObj *pVnode);
+static int32_t  vnodeReadVersion(SVnodeObj *pVnode);
 static int      vnodeWalCallback(void *arg);
 static uint32_t vnodeGetFileInfo(void *ahandle, char *name, uint32_t *index, int32_t *size);
 static int      vnodeGetWalInfo(void *ahandle, char *name, uint32_t *index);
@@ -93,21 +97,21 @@ int32_t vnodeCreate(SMDCreateVnodeMsg *pVnodeCfg) {
 
   STsdbCfg tsdbCfg = {0};
   tsdbCfg.precision           = pVnodeCfg->cfg.precision;
-  tsdbCfg.compression         = -1;
+  tsdbCfg.compression         = pVnodeCfg->cfg.compression;;
   tsdbCfg.tsdbId              = pVnodeCfg->cfg.vgId;
-  tsdbCfg.maxTables           = pVnodeCfg->cfg.maxSessions;
+  tsdbCfg.maxTables           = pVnodeCfg->cfg.maxTables;
   tsdbCfg.daysPerFile         = pVnodeCfg->cfg.daysPerFile;
-  tsdbCfg.minRowsPerFileBlock = -1;
-  tsdbCfg.maxRowsPerFileBlock = -1;
-  tsdbCfg.keep                = -1;
-  tsdbCfg.maxCacheSize        = -1;
+  tsdbCfg.minRowsPerFileBlock = pVnodeCfg->cfg.minRowsPerFileBlock;
+  tsdbCfg.maxRowsPerFileBlock = pVnodeCfg->cfg.maxRowsPerFileBlock;
+  tsdbCfg.keep                = pVnodeCfg->cfg.daysToKeep;
+  tsdbCfg.maxCacheSize        = pVnodeCfg->cfg.maxCacheSize;
 
   char tsdbDir[TSDB_FILENAME_LEN] = {0};
   sprintf(tsdbDir, "%s/vnode%d/tsdb", tsVnodeDir, pVnodeCfg->cfg.vgId);
   code = tsdbCreateRepo(tsdbDir, &tsdbCfg, NULL);
   if (code != TSDB_CODE_SUCCESS) {
-    dError("vgId:%d, failed to create tsdb in vnode, reason:%s", pVnodeCfg->cfg.vgId, tstrerror(terrno));
-    return terrno;
+    dError("vgId:%d, failed to create tsdb in vnode, reason:%s", pVnodeCfg->cfg.vgId, tstrerror(code));
+    return TSDB_CODE_VG_INIT_FAILED;
   }
 
   dPrint("vgId:%d, vnode is created, clog:%d", pVnodeCfg->cfg.vgId, pVnodeCfg->cfg.commitLog);
@@ -131,6 +135,39 @@ int32_t vnodeDrop(int32_t vgId) {
   return TSDB_CODE_SUCCESS;
 }
 
+int32_t vnodeAlter(void *param, SMDCreateVnodeMsg *pVnodeCfg) {
+  SVnodeObj *pVnode = param;
+  int32_t code = vnodeSaveCfg(pVnodeCfg);
+  if (code != TSDB_CODE_SUCCESS) {
+    dError("vgId:%d, failed to save vnode cfg, reason:%s", pVnodeCfg->cfg.vgId, tstrerror(code));
+    return code;
+  }
+
+  code = vnodeReadCfg(pVnode);
+  if (code != TSDB_CODE_SUCCESS) {
+    dError("pVnode:%p vgId:%d, failed to read cfg file", pVnode, pVnode->vgId);
+    taosDeleteIntHash(tsDnodeVnodesHash, pVnode->vgId);
+    return code;
+  }
+
+  code = syncReconfig(pVnode->sync, &pVnode->syncCfg);
+  if (code != TSDB_CODE_SUCCESS) {
+    dTrace("pVnode:%p vgId:%d, failed to alter vnode, canot reconfig sync, result:%s", pVnode, pVnode->vgId,
+           tstrerror(code));
+    return code;
+  }
+
+  code = tsdbConfigRepo(pVnode->tsdb, &pVnode->tsdbCfg);
+  if (code != TSDB_CODE_SUCCESS) {
+    dTrace("pVnode:%p vgId:%d, failed to alter vnode, canot reconfig tsdb, result:%s", pVnode, pVnode->vgId,
+           tstrerror(code));
+    return code;
+  }
+
+  dTrace("pVnode:%p vgId:%d, vnode is altered", pVnode, pVnode->vgId);
+  return TSDB_CODE_SUCCESS;
+}
+
 int32_t vnodeOpen(int32_t vnode, char *rootDir) {
   char temp[TSDB_FILENAME_LEN];
   pthread_once(&vnodeModuleInit, vnodeInit);
@@ -149,6 +186,8 @@ int32_t vnodeOpen(int32_t vnode, char *rootDir) {
     return code;
   }
 
+  vnodeReadVersion(pVnode);
+  
   pVnode->wqueue = dnodeAllocateWqueue(pVnode);
   pVnode->rqueue = dnodeAllocateRqueue(pVnode);
 
@@ -227,7 +266,9 @@ void vnodeRelease(void *pVnodeRaw) {
   pVnode->wqueue = NULL;
 
   if (pVnode->status == TAOS_VN_STATUS_DELETING) {
-    // remove the whole directory
+    char rootDir[TSDB_FILENAME_LEN] = {0};
+    sprintf(rootDir, "%s/vnode%d", tsVnodeDir, vgId);
+    taosRemoveDir(rootDir);
   }
 
   free(pVnode);
@@ -246,7 +287,8 @@ void *vnodeGetVnode(int32_t vgId) {
   SVnodeObj **ppVnode = (SVnodeObj **)taosGetIntHashData(tsDnodeVnodesHash, vgId);
   if (ppVnode == NULL || *ppVnode == NULL) {
     terrno = TSDB_CODE_INVALID_VGROUP_ID;
-    assert(false);
+    dError("vgId:%d not exist");
+    return NULL;
   }
 
   return *ppVnode;
@@ -292,6 +334,7 @@ static void vnodeBuildVloadMsg(char *pNode, void * param) {
   pLoad->vgId = htonl(pVnode->vgId);
   pLoad->status = pVnode->status;
   pLoad->role = pVnode->role;
+  pLoad->replica = pVnode->syncCfg.replica;
 }
 
 static void vnodeCleanUp(SVnodeObj *pVnode) {
@@ -301,6 +344,7 @@ static void vnodeCleanUp(SVnodeObj *pVnode) {
   //syncStop(pVnode->sync);
   tsdbCloseRepo(pVnode->tsdb);
   walClose(pVnode->wal);
+  vnodeSaveVersion(pVnode);
 
   vnodeRelease(pVnode);
 }
@@ -328,88 +372,306 @@ static void vnodeNotifyRole(void *ahandle, int8_t role) {
 }
 
 static int32_t vnodeSaveCfg(SMDCreateVnodeMsg *pVnodeCfg) {
-  char cfgFile[TSDB_FILENAME_LEN * 2] = {0};
-  sprintf(cfgFile, "%s/vnode%d/config", tsVnodeDir, pVnodeCfg->cfg.vgId);
-
+  char cfgFile[TSDB_FILENAME_LEN + 30] = {0};
+  sprintf(cfgFile, "%s/vnode%d/config.json", tsVnodeDir, pVnodeCfg->cfg.vgId);
   FILE *fp = fopen(cfgFile, "w");
-  if (!fp) return errno;
+  if (!fp) {
+    dError("vgId:%d, failed to open vnode cfg file for write, error:%s", pVnodeCfg->cfg.vgId, strerror(errno));
+    return errno;
+  }
+
+  char    ipStr[20];
+  int32_t len = 0;
+  int32_t maxLen = 1000;
+  char *  content = calloc(1, maxLen + 1);
 
-  fprintf(fp, "commitLog %d\n", pVnodeCfg->cfg.commitLog);
-  fprintf(fp, "wals %d\n", 3);
-  fprintf(fp, "arbitratorIp %d\n", pVnodeCfg->vpeerDesc[0].ip);
-  fprintf(fp, "quorum %d\n", 1);
-  fprintf(fp, "replica %d\n", pVnodeCfg->cfg.replications);
+  len += snprintf(content + len, maxLen - len, "{\n");
+
+  len += snprintf(content + len, maxLen - len, "  \"precision\": %d,\n", pVnodeCfg->cfg.precision);
+  len += snprintf(content + len, maxLen - len, "  \"compression\": %d,\n", pVnodeCfg->cfg.compression);
+  len += snprintf(content + len, maxLen - len, "  \"maxTables\": %d,\n", pVnodeCfg->cfg.maxTables);
+  len += snprintf(content + len, maxLen - len, "  \"daysPerFile\": %d,\n", pVnodeCfg->cfg.daysPerFile);
+  len += snprintf(content + len, maxLen - len, "  \"minRowsPerFileBlock\": %d,\n", pVnodeCfg->cfg.minRowsPerFileBlock);
+  len += snprintf(content + len, maxLen - len, "  \"maxRowsPerFileBlock\": %d,\n", pVnodeCfg->cfg.maxRowsPerFileBlock);
+  len += snprintf(content + len, maxLen - len, "  \"daysToKeep\": %d,\n", pVnodeCfg->cfg.daysToKeep);
+
+  len += snprintf(content + len, maxLen - len, "  \"maxCacheSize\": %" PRId64 ",\n", pVnodeCfg->cfg.maxCacheSize);
+  
+  len += snprintf(content + len, maxLen - len, "  \"commitLog\": %d,\n", pVnodeCfg->cfg.commitLog);
+  len += snprintf(content + len, maxLen - len, "  \"wals\": %d,\n", pVnodeCfg->cfg.wals);
+
+  uint32_t ipInt =  pVnodeCfg->cfg.arbitratorIp;
+  sprintf(ipStr, "%u.%u.%u.%u", ipInt & 0xFF, (ipInt >> 8) & 0xFF, (ipInt >> 16) & 0xFF, (uint8_t)(ipInt >> 24));
+  len += snprintf(content + len, maxLen - len, "  \"arbitratorIp\": \"%s\",\n", ipStr);
+
+  len += snprintf(content + len, maxLen - len, "  \"quorum\": %d,\n", pVnodeCfg->cfg.quorum);
+  len += snprintf(content + len, maxLen - len, "  \"replica\": %d,\n", pVnodeCfg->cfg.replications);
+
+  len += snprintf(content + len, maxLen - len, "  \"nodeInfos\": [{\n");
   for (int32_t i = 0; i < pVnodeCfg->cfg.replications; i++) {
-    fprintf(fp, "index%d nodeId %d nodeIp %u name n%d\n", i, pVnodeCfg->vpeerDesc[i].dnodeId, pVnodeCfg->vpeerDesc[i].ip, pVnodeCfg->vpeerDesc[i].dnodeId);
+    len += snprintf(content + len, maxLen - len, "    \"nodeId\": %d,\n", pVnodeCfg->nodes[i].nodeId);
+
+    uint32_t ipInt = pVnodeCfg->nodes[i].nodeIp;
+    sprintf(ipStr, "%u.%u.%u.%u", ipInt & 0xFF, (ipInt >> 8) & 0xFF, (ipInt >> 16) & 0xFF, (uint8_t)(ipInt >> 24));
+    len += snprintf(content + len, maxLen - len, "    \"nodeIp\": \"%s\",\n", ipStr);
+
+    len += snprintf(content + len, maxLen - len, "    \"nodeName\": \"%s\"\n", pVnodeCfg->nodes[i].nodeName);
+
+    if (i < pVnodeCfg->cfg.replications - 1) {
+      len += snprintf(content + len, maxLen - len, "  },{\n");
+    } else {
+      len += snprintf(content + len, maxLen - len, "  }]\n");
     }
+  }
+  len += snprintf(content + len, maxLen - len, "}\n");
 
+  fwrite(content, 1, len, fp);
   fclose(fp);
-  dTrace("vgId:%d, save vnode cfg successed", pVnodeCfg->cfg.vgId);
+  free(content);
 
-  return TSDB_CODE_SUCCESS;
+  dPrint("vgId:%d, save vnode cfg successed", pVnodeCfg->cfg.vgId);
+
+  return 0;
 }
 
-// TODO: this is a simple implement
 static int32_t vnodeReadCfg(SVnodeObj *pVnode) {
-  char option[5][16] = {0};
-  char cfgFile[TSDB_FILENAME_LEN * 2] = {0};
-  sprintf(cfgFile, "%s/vnode%d/config", tsVnodeDir, pVnode->vgId);
-
+  char cfgFile[TSDB_FILENAME_LEN + 30] = {0};
+  sprintf(cfgFile, "%s/vnode%d/config.json", tsVnodeDir, pVnode->vgId);
   FILE *fp = fopen(cfgFile, "r");
-  if (!fp) return errno;
-
-  int32_t commitLog = -1;
-  int32_t num = fscanf(fp, "%s %d", option[0], &commitLog);
-  if (num != 2) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (strcmp(option[0], "commitLog") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (commitLog == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-  pVnode->walCfg.commitLog = (int8_t)commitLog;
-
-  int32_t wals = -1;
-  num = fscanf(fp, "%s %d", option[0], &wals);
-  if (num != 2) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (strcmp(option[0], "wals") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (wals == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-  pVnode->walCfg.wals = (int8_t)wals;
+  if (!fp) {
+    dError("pVnode:%p vgId:%d, failed to open vnode cfg file for read, error:%s", pVnode, pVnode->vgId, strerror(errno));
+    return errno;
+  }
+
+  int   ret = TSDB_CODE_OTHERS;
+  int   maxLen = 1000;
+  char *content = calloc(1, maxLen + 1);
+  int   len = fread(content, 1, maxLen, fp);
+  if (len <= 0) {
+    free(content);
+    fclose(fp);
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, content is null", pVnode, pVnode->vgId);
+    return false;
+  }
+
+  cJSON *root = cJSON_Parse(content);
+  if (root == NULL) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, invalid json format", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+
+  cJSON *precision = cJSON_GetObjectItem(root, "precision");
+  if (!precision || precision->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, precision not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.precision = (int8_t)precision->valueint;
+
+  cJSON *compression = cJSON_GetObjectItem(root, "compression");
+  if (!compression || compression->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, compression not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.compression = (int8_t)compression->valueint;
+
+  cJSON *maxTables = cJSON_GetObjectItem(root, "maxTables");
+  if (!maxTables || maxTables->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, maxTables not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.maxTables = maxTables->valueint;
+
+  cJSON *daysPerFile = cJSON_GetObjectItem(root, "daysPerFile");
+  if (!daysPerFile || daysPerFile->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, daysPerFile not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.daysPerFile = daysPerFile->valueint;
+
+  cJSON *minRowsPerFileBlock = cJSON_GetObjectItem(root, "minRowsPerFileBlock");
+  if (!minRowsPerFileBlock || minRowsPerFileBlock->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, minRowsPerFileBlock not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.minRowsPerFileBlock = minRowsPerFileBlock->valueint;
+
+  cJSON *maxRowsPerFileBlock = cJSON_GetObjectItem(root, "maxRowsPerFileBlock");
+  if (!maxRowsPerFileBlock || maxRowsPerFileBlock->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, maxRowsPerFileBlock not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.maxRowsPerFileBlock = maxRowsPerFileBlock->valueint;
+
+  cJSON *daysToKeep = cJSON_GetObjectItem(root, "daysToKeep");
+  if (!daysToKeep || daysToKeep->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, daysToKeep not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.keep = daysToKeep->valueint;
+
+  cJSON *maxCacheSize = cJSON_GetObjectItem(root, "maxCacheSize");
+  if (!maxCacheSize || maxCacheSize->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, maxCacheSize not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->tsdbCfg.maxCacheSize = maxCacheSize->valueint;
+
+  cJSON *commitLog = cJSON_GetObjectItem(root, "commitLog");
+  if (!commitLog || commitLog->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, commitLog not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->walCfg.commitLog = (int8_t)commitLog->valueint;
+
+  cJSON *wals = cJSON_GetObjectItem(root, "wals");
+  if (!wals || wals->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, wals not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->walCfg.wals = (int8_t)wals->valueint;
   pVnode->walCfg.keep = 0;
 
-  int32_t arbitratorIp = -1;
-  num = fscanf(fp, "%s %u", option[0], &arbitratorIp);
-  if (num != 2) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (strcmp(option[0], "arbitratorIp") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (arbitratorIp == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-  pVnode->syncCfg.arbitratorIp = arbitratorIp;
-
-  int32_t quorum = -1;
-  num = fscanf(fp, "%s %d", option[0], &quorum);
-  if (num != 2) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (strcmp(option[0], "quorum") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (quorum == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-  pVnode->syncCfg.quorum = (int8_t)quorum;
-
-  int32_t replica = -1;
-  num = fscanf(fp, "%s %d", option[0], &replica);
-  if (num != 2) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (strcmp(option[0], "replica") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-  if (replica == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-  pVnode->syncCfg.replica = (int8_t)replica;
-
-  for (int32_t i = 0; i < replica; ++i) {
-    int32_t  dnodeId = -1;
-    uint32_t dnodeIp = -1;
-    num = fscanf(fp, "%s %s %d %s %u %s %s", option[0], option[1], &dnodeId, option[2], &dnodeIp, option[3], pVnode->syncCfg.nodeInfo[i].name);
-    if (num != 7) return TSDB_CODE_INVALID_FILE_FORMAT;
-    if (strcmp(option[1], "nodeId") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-    if (strcmp(option[2], "nodeIp") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-    if (strcmp(option[3], "name") != 0) return TSDB_CODE_INVALID_FILE_FORMAT;
-    if (dnodeId == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-    if (dnodeIp == -1) return TSDB_CODE_INVALID_FILE_FORMAT;
-    pVnode->syncCfg.nodeInfo[i].nodeId = dnodeId;
-    pVnode->syncCfg.nodeInfo[i].nodeIp = dnodeIp;
+  cJSON *arbitratorIp = cJSON_GetObjectItem(root, "arbitratorIp");
+  if (!arbitratorIp || arbitratorIp->type != cJSON_String || arbitratorIp->valuestring == NULL) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, arbitratorIp not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->syncCfg.arbitratorIp = inet_addr(arbitratorIp->valuestring);
+
+  cJSON *quorum = cJSON_GetObjectItem(root, "quorum");
+  if (!quorum || quorum->type != cJSON_Number) {
+    dError("failed to read vnode cfg, quorum not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->syncCfg.quorum = (int8_t)quorum->valueint;
+
+  cJSON *replica = cJSON_GetObjectItem(root, "replica");
+  if (!replica || replica->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, replica not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->syncCfg.replica = (int8_t)replica->valueint;
+
+  cJSON *nodeInfos = cJSON_GetObjectItem(root, "nodeInfos");
+  if (!nodeInfos || nodeInfos->type != cJSON_Array) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, nodeInfos not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+
+  int size = cJSON_GetArraySize(nodeInfos);
+  if (size != pVnode->syncCfg.replica) {
+    dError("pVnode:%p vgId:%d, failed to read vnode cfg, nodeInfos size not matched", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+
+  for (int i = 0; i < size; ++i) {
+    cJSON *nodeInfo = cJSON_GetArrayItem(nodeInfos, i);
+    if (nodeInfo == NULL) continue;
+
+    cJSON *nodeId = cJSON_GetObjectItem(nodeInfo, "nodeId");
+    if (!nodeId || nodeId->type != cJSON_Number) {
+      dError("pVnode:%p vgId:%d, failed to read vnode cfg, nodeId not found", pVnode, pVnode->vgId);
+      goto PARSE_OVER;
+    }
+    pVnode->syncCfg.nodeInfo[i].nodeId = nodeId->valueint;
+
+    cJSON *nodeIp = cJSON_GetObjectItem(nodeInfo, "nodeIp");
+    if (!nodeIp || nodeIp->type != cJSON_String || nodeIp->valuestring == NULL) {
+      dError("pVnode:%p vgId:%d, failed to read vnode cfg, nodeIp not found", pVnode, pVnode->vgId);
+      goto PARSE_OVER;
+    }
+    pVnode->syncCfg.nodeInfo[i].nodeIp = inet_addr(nodeIp->valuestring);
+
+    cJSON *nodeName = cJSON_GetObjectItem(nodeInfo, "nodeName");
+    if (!nodeName || nodeName->type != cJSON_String || nodeName->valuestring == NULL) {
+      dError("pVnode:%p vgId:%d, failed to read vnode cfg, nodeName not found", pVnode, pVnode->vgId);
+      goto PARSE_OVER;
+    }
+    strncpy(pVnode->syncCfg.nodeInfo[i].name, nodeName->valuestring, TSDB_NODE_NAME_LEN);
   }
 
+  ret = 0;
+
+  dPrint("pVnode:%p vgId:%d, read vnode cfg successed, replcia:%d", pVnode, pVnode->vgId, pVnode->syncCfg.replica);
+  for (int32_t i = 0; i < pVnode->syncCfg.replica; i++) {
+    dPrint("pVnode:%p vgId:%d, dnode:%d, ip:%s name:%s", pVnode, pVnode->vgId, pVnode->syncCfg.nodeInfo[i].nodeId,
+           taosIpStr(pVnode->syncCfg.nodeInfo[i].nodeIp), pVnode->syncCfg.nodeInfo[i].name);
+  }
+
+PARSE_OVER:
+  free(content);
+  cJSON_Delete(root);
   fclose(fp);
-  dTrace("pVnode:%p vgId:%d, read vnode cfg successed", pVnode, pVnode->vgId);
+  return ret;
+}
 
-  return TSDB_CODE_SUCCESS;
+
+static int32_t vnodeSaveVersion(SVnodeObj *pVnode) {
+  char versionFile[TSDB_FILENAME_LEN + 30] = {0};
+  sprintf(versionFile, "%s/vnode%d/version.json", tsVnodeDir, pVnode->vgId);
+  FILE *fp = fopen(versionFile, "w");
+  if (!fp) {
+    dError("pVnode:%p vgId:%d, failed to open vnode version file for write, error:%s", pVnode, pVnode->vgId);
+    return errno;
+  }
+
+  int32_t len = 0;
+  int32_t maxLen = 30;
+  char *  content = calloc(1, maxLen + 1);
+
+  len += snprintf(content + len, maxLen - len, "{\n");
+  len += snprintf(content + len, maxLen - len, "  \"version\": %" PRId64 "\n", pVnode->version);
+  len += snprintf(content + len, maxLen - len, "}\n");
+
+  fwrite(content, 1, len, fp);
+  fclose(fp);
+  free(content);
+
+  dPrint("pVnode:%p vgId:%d, save vnode version successed", pVnode, pVnode->vgId);
+
+  return 0;
+}
+
+static int32_t vnodeReadVersion(SVnodeObj *pVnode) {
+  char versionFile[TSDB_FILENAME_LEN + 30] = {0};
+  sprintf(versionFile, "%s/vnode%d/version.json", tsVnodeDir, pVnode->vgId);
+  FILE *fp = fopen(versionFile, "w");
+  if (!fp) {
+    dError("pVnode:%p vgId:%d, failed to open vnode version file for write, error:%s", pVnode, pVnode->vgId);
+    return errno;
+  }
+
+  int   ret = TSDB_CODE_OTHERS;
+  int   maxLen = 100;
+  char *content = calloc(1, maxLen + 1);
+  int   len = fread(content, 1, maxLen, fp);
+  if (len <= 0) {
+    free(content);
+    fclose(fp);
+    dError("pVnode:%p vgId:%d, failed to read vnode version, content is null", pVnode, pVnode->vgId);
+    return false;
+  }
+
+  cJSON *root = cJSON_Parse(content);
+  if (root == NULL) {
+    dError("pVnode:%p vgId:%d, failed to read vnode version, invalid json format", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+
+  cJSON *version = cJSON_GetObjectItem(root, "version");
+  if (!version || version->type != cJSON_Number) {
+    dError("pVnode:%p vgId:%d, failed to read vnode version, version not found", pVnode, pVnode->vgId);
+    goto PARSE_OVER;
+  }
+  pVnode->version = version->valueint;
+
+  ret = 0;
+
+  dPrint("pVnode:%p vgId:%d, read vnode version successed, version:%%" PRId64, pVnode, pVnode->vgId, pVnode->version);
+
+PARSE_OVER:
+  free(content);
+  cJSON_Delete(root);
+  fclose(fp);
+  return ret;
 }
\ No newline at end of file

tests/pytest/insert/basic.py

@@ -17,6 +17,7 @@ from util.log import *
 from util.cases import *
 from util.sql import *
 
+
 class TDTestCase:
     def init(self, conn):
         tdLog.debug("start to execute %s" % __file__)
@@ -24,25 +25,24 @@ class TDTestCase:
 
     def run(self):
         tdSql.prepare()
-        tdSql.execute('show databases')
-        tdSql.execute('drop database if exists db')
-        tdSql.execute('create database db')
-        tdSql.execute('use db')
-        tdSql.execute('create table tb (ts timestamp, speed int)')
+
+        ret = tdSql.execute('create table tb (ts timestamp, speed int)')
 
         insertRows = 10
         tdLog.info("insert %d rows" % (insertRows))
         for i in range(0, insertRows):
-            tdSql.execute('insert into tb values (now + %dm, %d)' % (i, i))
+            ret = tdSql.execute(
+                'insert into tb values (now + %dm, %d)' %
+                (i, i))
 
-#        tdLog.info("insert earlier data")
-#        tdSql.execute('insert into tb values (now - 5m , 10)')
-#        tdSql.execute('insert into tb values (now - 6m , 10)')
-#        tdSql.execute('insert into tb values (now - 7m , 10)')
-#        tdSql.execute('insert into tb values (now - 8m , 10)')
+        tdLog.info("insert earlier data")
+        tdSql.execute('insert into tb values (now - 5m , 10)')
+        tdSql.execute('insert into tb values (now - 6m , 10)')
+        tdSql.execute('insert into tb values (now - 7m , 10)')
+        tdSql.execute('insert into tb values (now - 8m , 10)')
 
         tdSql.query("select * from tb")
-        tdSql.checkRows(insertRows)
+        tdSql.checkRows(insertRows + 4)
 
     def stop(self):
         tdSql.close()

tests/pytest/simpletest.sh

-sudo python ./test.py -f insert/basic.py
+#!/bin/bash
+python3 ./test.py -f insert/basic.py $1
+python3 ./test.py -s $1

tests/pytest/test.py

@@ -15,22 +15,25 @@
 # -*- coding: utf-8 -*-
 import sys
 import getopt
+import subprocess
+from distutils.log import warn as printf
+
 from util.log import *
 from util.dnodes import *
 from util.cases import *
 
 import taos
 
-# add testcase here:
-from insert.basic import *
 
 if __name__ == "__main__":
     fileName = "all"
     deployPath = ""
     masterIp = ""
     testCluster = False
-    opts, args = getopt.getopt(sys.argv[1:], 'f:p:m:sch', [
-                               'file=', 'path=', 'master', 'stop', 'cluster', 'help'])
+    valgrind = 0
+    stop = 0
+    opts, args = getopt.gnu_getopt(sys.argv[1:], 'f:p:m:scgh', [
+        'file=', 'path=', 'master', 'stop', 'cluster', 'valgrind', 'help'])
     for key, value in opts:
         if key in ['-h', '--help']:
             tdLog.printNoPrefix(
@@ -41,21 +44,50 @@ if __name__ == "__main__":
             tdLog.printNoPrefix('-c Test Cluster Flag')
             tdLog.printNoPrefix('-s stop All dnodes')
             sys.exit(0)
+
         if key in ['-f', '--file']:
             fileName = value
+
         if key in ['-p', '--path']:
             deployPath = value
+
         if key in ['-m', '--master']:
             masterIp = value
+
         if key in ['-c', '--cluster']:
             testCluster = True
+
+        if key in ['-g', '--valgrind']:
+            valgrind = 1
+
         if key in ['-s', '--stop']:
-            cmd = "ps -ef|grep -w taosd | grep 'taosd' | grep -v grep | awk '{print $2}' && pkill -9 taosd"
-            os.system(cmd)
+            stop = 1
+
+    if (stop != 0):
+        if (valgrind == 0):
+            toBeKilled = "taosd"
+        else:
+            toBeKilled = "valgrind.bin"
+
+        killCmd = "ps -ef|grep -w %s| grep -v grep | awk '{print $2}' | xargs kill -HUP " % toBeKilled
+#        os.system(killCmd)
+#        time.sleep(1)
+
+        psCmd = "ps -ef|grep -w %s| grep -v grep | awk '{print $2}'" % toBeKilled
+        processID = subprocess.check_output(psCmd, shell=True)
+
+        while(processID):
+            os.system(killCmd)
+            time.sleep(1)
+            processID = subprocess.check_output(psCmd, shell=True)
+
         tdLog.exit('stop All dnodes')
 
     if masterIp == "":
         tdDnodes.init(deployPath)
+        tdDnodes.setTestCluster(testCluster)
+        tdDnodes.setValgrind(valgrind)
+
         if testCluster:
             tdLog.notice("Procedures for testing cluster")
             if fileName == "all":

tests/pytest/util/cases.py

@@ -15,6 +15,8 @@ import sys
 import os
 import time
 import datetime
+import inspect
+import importlib
 from util.log import *
 
 
@@ -30,6 +32,10 @@ class TDCases:
         self.windowsCases = []
         self.clusterCases = []
 
+    def __dynamicLoadModule(self, fileName):
+        moduleName = fileName.replace(".py", "").replace("/", ".")
+        return importlib.import_module(moduleName, package='..')
+
     def addWindows(self, name, case):
         self.windowsCases.append(TDCase(name, case))
 
@@ -40,64 +46,93 @@ class TDCases:
         self.clusterCases.append(TDCase(name, case))
 
     def runAllLinux(self, conn):
-        tdLog.notice("run total %d cases" % (len(self.linuxCases)))
-        for case in self.linuxCases:
-            case.case.init(conn)
-            case.case.run()
-            case.case.stop()
-        tdLog.notice("total %d cases executed" % (len(self.linuxCases)))
+        # TODO: load all Linux cases here
+        runNum = 0
+        for tmp in self.linuxCases:
+            if tmp.name.find(fileName) != -1:
+                case = testModule.TDTestCase()
+                case.init(conn)
+                case.run()
+                case.stop()
+                runNum += 1
+                continue
+
+        tdLog.notice("total %d Linux test case(s) executed" % (runNum))
 
     def runOneLinux(self, conn, fileName):
-        tdLog.notice("run cases like %s" % (fileName))
+        testModule = self.__dynamicLoadModule(fileName)
+
         runNum = 0
-        for case in self.linuxCases:
-            if case.name.find(fileName) != -1:
-                case.case.init(conn)
-                case.case.run()
-                case.case.stop()
-                time.sleep(5)
+        for tmp in self.linuxCases:
+            if tmp.name.find(fileName) != -1:
+                case = testModule.TDTestCase()
+                case.init(conn)
+                case.run()
+                case.stop()
                 runNum += 1
-        tdLog.notice("total %d cases executed" % (runNum))
+                continue
+
+        tdLog.notice("total %d Linux test case(s) executed" % (runNum))
 
     def runAllWindows(self, conn):
-        tdLog.notice("run total %d cases" % (len(self.windowsCases)))
-        for case in self.windowsCases:
-            case.case.init(conn)
-            case.case.run()
-            case.case.stop()
-        tdLog.notice("total %d cases executed" % (len(self.windowsCases)))
+        # TODO: load all Windows cases here
+        runNum = 0
+        for tmp in self.windowsCases:
+            if tmp.name.find(fileName) != -1:
+                case = testModule.TDTestCase()
+                case.init(conn)
+                case.run()
+                case.stop()
+                runNum += 1
+                continue
+
+        tdLog.notice("total %d Windows test case(s) executed" % (runNum))
 
     def runOneWindows(self, conn, fileName):
-        tdLog.notice("run cases like %s" % (fileName))
+        testModule = self.__dynamicLoadModule(fileName)
+
         runNum = 0
-        for case in self.windowsCases:
-            if case.name.find(fileName) != -1:
-                case.case.init(conn)
-                case.case.run()
-                case.case.stop()
-                time.sleep(2)
+        for tmp in self.windowsCases:
+            if tmp.name.find(fileName) != -1:
+                case = testModule.TDTestCase()
+                case.init(conn)
+                case.run()
+                case.stop()
                 runNum += 1
-        tdLog.notice("total %d cases executed" % (runNum))
+                continue
+        tdLog.notice("total %d Windows case(s) executed" % (runNum))
 
     def runAllCluster(self):
-        tdLog.notice("run total %d cases" % (len(self.clusterCases)))
-        for case in self.clusterCases:
-            case.case.init()
-            case.case.run()
-            case.case.stop()
-        tdLog.notice("total %d cases executed" % (len(self.clusterCases)))
+        # TODO: load all cluster case module here
 
-    def runOneCluster(self, fileName):
+        runNum = 0
+        for tmp in self.clusterCases:
+            if tmp.name.find(fileName) != -1:
                 tdLog.notice("run cases like %s" % (fileName))
+                case = testModule.TDTestCase()
+                case.init()
+                case.run()
+                case.stop()
+                runNum += 1
+                continue
+
+        tdLog.notice("total %d Cluster test case(s) executed" % (runNum))
+
+    def runOneCluster(self, fileName):
+        testModule = self.__dynamicLoadModule(fileName)
+
         runNum = 0
-        for case in self.clusterCases:
-            if case.name.find(fileName) != -1:
-                case.case.init()
-                case.case.run()
-                case.case.stop()
-                time.sleep(2)
+        for tmp in self.clusterCases:
+            if tmp.name.find(fileName) != -1:
+                tdLog.notice("run cases like %s" % (fileName))
+                case = testModule.TDTestCase()
+                case.init()
+                case.run()
+                case.stop()
                 runNum += 1
-        tdLog.notice("total %d cases executed" % (runNum))
+                continue
+
+        tdLog.notice("total %d Cluster test case(s) executed" % (runNum))
 
 
 tdCases = TDCases()

tests/pytest/util/dnodes.py

@@ -14,6 +14,7 @@
 import sys
 import os
 import os.path
+import subprocess
 from util.log import *
 
 
@@ -78,10 +79,18 @@ class TDDnode:
         self.index = index
         self.running = 0
         self.deployed = 0
+        self.testCluster = False
+        self.valgrind = 0
 
     def init(self, path):
         self.path = path
 
+    def setTestCluster(self, value):
+        self.testCluster = value
+
+    def setValgrind(self, value):
+        self.valgrind = value
+
     def deploy(self):
         self.logDir = "%s/sim/dnode%d/log" % (self.path, self.index)
         self.dataDir = "%s/sim/dnode%d/data" % (self.path, self.index)
@@ -116,7 +125,9 @@ class TDDnode:
         if os.system(cmd) != 0:
             tdLog.exit(cmd)
 
+        if self.testCluster:
             self.startIP()
+
         self.cfg("masterIp", "192.168.0.1")
         self.cfg("secondIp", "192.168.0.2")
         self.cfg("publicIp", "192.168.0.%d" % (self.index))
@@ -164,9 +175,18 @@ class TDDnode:
 
         if self.deployed == 0:
             tdLog.exit("dnode:%d is not deployed" % (self.index))
+
+        if self.valgrind == 0:
             cmd = "nohup %staosd -c %s > /dev/null 2>&1 & " % (
                 binPath, self.cfgDir)
+        else:
+            valgrindCmdline = "valgrind --tool=memcheck --leak-check=full --show-reachable=no --track-origins=yes --show-leak-kinds=all -v --workaround-gcc296-bugs=yes"
+
+            cmd = "nohup %s %staosd -c %s 2>&1 & " % (
+                valgrindCmdline, binPath, self.cfgDir)
+
             print(cmd)
+
         if os.system(cmd) != 0:
             tdLog.exit(cmd)
         self.running = 1
@@ -275,8 +295,16 @@ class TDDnodes:
         self.sim.init(self.path)
         self.sim.deploy()
 
+    def setTestCluster(self, value):
+        self.testCluster = value
+
+    def setValgrind(self, value):
+        self.valgrind = value
+
     def deploy(self, index):
         self.check(index)
+        self.dnodes[index - 1].setTestCluster(self.testCluster)
+        self.dnodes[index - 1].setValgrind(self.valgrind)
         self.dnodes[index - 1].deploy()
 
     def cfg(self, index, option, value):
@@ -297,10 +325,14 @@ class TDDnodes:
 
     def startIP(self, index):
         self.check(index)
+
+        if self.testCluster:
             self.dnodes[index - 1].startIP()
 
     def stopIP(self, index):
         self.check(index)
+
+        if self.dnodes[index - 1].testCluster:
             self.dnodes[index - 1].stopIP()
 
     def check(self, index):
@@ -312,11 +344,14 @@ class TDDnodes:
         for i in range(len(self.dnodes)):
             self.dnodes[i].stop()
 
+        psCmd = "ps -ef | grep -w taosd | grep 'root' | grep -v grep | awk '{print $2}'"
+        processID = subprocess.check_output(psCmd, shell=True)
+        if processID:
             cmd = "sudo systemctl stop taosd"
             os.system(cmd)
         # if os.system(cmd) != 0 :
         # tdLog.exit(cmd)
-        cmd = "ps -ef | grep -w taosd | grep 'dnode' | grep -v grep | awk '{print $2}' && sudo pkill -sigkill taosd"
+        cmd = "ps -ef | grep -w taosd | grep 'dnode' | grep -v grep | awk '{print $2}' && pkill -sigkill taosd"
         os.system(cmd)
         # if os.system(cmd) != 0 :
         # tdLog.exit(cmd)

tests/pytest/util/log.py

@@ -42,7 +42,7 @@ class TDLog:
         printf("\033[1;31m%s %s\033[0m" % (datetime.datetime.now(), err))
         sys.exit(1)
 
-    def printfNoPrefix(self, info):
+    def printNoPrefix(self, info):
         printf("\033[1;36m%s\033[0m" % (info))
 
 

tests/script/basicSuite.sim

@@ -3,6 +3,6 @@
 run general/db/testSuite.sim
 run general/insert/testSuite.sim
 run general/table/testSuite.sim
-run general/user/testSuite.sim
+run general/user/basicSuite.sim
 
 ################################## 

tests/script/general/user/basic1.sim

@@ -66,6 +66,4 @@ print $data10 $data11 $data22
 print $data20 $data11 $data22
 print $data30 $data31 $data32
 
-
-  
+system sh/exec.sh -n dnode1 -s stop
\ No newline at end of file
-  

tests/script/general/user/basicSuite.sim

+run general/user/basic1.sim
\ No newline at end of file

tests/script/general/user/testSuite.sim

 run general/user/basic1.sim
+run general/user/pass_alter.sim
+run general/user/pass_len.sim
+run general/user/user_create.sim
+run general/user/user_len.sim
+#run general/user/monitor.sim
\ No newline at end of file

tests/script/general/account/user_create.sim → tests/script/general/user/user_create.sim

@@ -4,8 +4,6 @@ system sh/ip.sh -i 1 -s up
 system sh/deploy.sh -n dnode1 -m 192.168.0.1 -i 192.168.0.1
 system sh/cfg.sh -n dnode1 -c clog -v 0
 system sh/exec.sh -n dnode1 -s start
-
-sleep 3000
 sql connect
 
 print =============== step1

tests/script/sh/deploy.sh

@@ -93,7 +93,7 @@ echo "privateIp           $NODE_IP"       >> $TAOS_CFG
 echo "dDebugFlag          199"            >> $TAOS_CFG
 echo "mDebugFlag          199"            >> $TAOS_CFG
 echo "sdbDebugFlag        199"            >> $TAOS_CFG
-echo "rpcDebugFlag        131"            >> $TAOS_CFG
+echo "rpcDebugFlag        135"            >> $TAOS_CFG
 echo "tmrDebugFlag        131"            >> $TAOS_CFG
 echo "cDebugFlag          135"            >> $TAOS_CFG
 echo "httpDebugFlag       131"            >> $TAOS_CFG
@@ -105,7 +105,7 @@ echo "numOfThreadsPerCore 2.0"            >> $TAOS_CFG
 echo "defaultPass         taosdata"       >> $TAOS_CFG
 echo "numOfLogLines       100000000"      >> $TAOS_CFG
 echo "mgmtEqualVnodeNum   0"              >> $TAOS_CFG
-echo "clog                0"              >> $TAOS_CFG
+echo "clog                2"              >> $TAOS_CFG
 echo "statusInterval      1"              >> $TAOS_CFG
 echo "numOfTotalVnodes    4"              >> $TAOS_CFG
 echo "asyncLog            0"              >> $TAOS_CFG

tests/script/test.sh

@@ -11,7 +11,8 @@ set +e
 FILE_NAME=
 RELEASE=0
 ASYNC=0
-while getopts "f:a" arg
+VALGRIND=0
+while getopts "f:av" arg
 do
   case $arg in
     f)
@@ -20,6 +21,9 @@ do
     a)
       ASYNC=1
       ;;
+    v)
+      VALGRIND=1
+      ;;
     ?)
       echo "unknow argument"
       ;;
@@ -30,6 +34,8 @@ cd .
 sh/ip.sh -i 1 -s up > /dev/null 2>&1 & 
 sh/ip.sh -i 2 -s up > /dev/null 2>&1 & 
 sh/ip.sh -i 3 -s up > /dev/null 2>&1 & 
+sh/ip.sh -i 4 -s up > /dev/null 2>&1 & 
+sh/ip.sh -i 5 -s up > /dev/null 2>&1 & 
 
 # Get responsible directories
 CODE_DIR=`dirname $0`
@@ -96,10 +102,14 @@ ulimit -c unlimited
 #sudo sysctl -w kernel.core_pattern=$TOP_DIR/core.%p.%e
 
 if [ -n "$FILE_NAME" ]; then
-  echo "ExcuteCmd:" $PROGRAM -c $CFG_DIR -f $FILE_NAME
   echo "------------------------------------------------------------------------"
-  #valgrind --tool=memcheck --leak-check=full --show-reachable=no  --track-origins=yes --show-leak-kinds=all  -v  --workaround-gcc296-bugs=yes  --log-file=valgrind.log $PROGRAM -c $CFG_DIR -f $FILE_NAME
+  if [ $VALGRIND -eq 1 ]; then
+    echo valgrind --tool=memcheck --leak-check=full --show-reachable=no  --track-origins=yes --show-leak-kinds=all  -v  --workaround-gcc296-bugs=yes  --log-file=${CODE_DIR}/../script/valgrind.log $PROGRAM -c $CFG_DIR -f $FILE_NAME
+    valgrind --tool=memcheck --leak-check=full --show-reachable=no  --track-origins=yes --show-leak-kinds=all  -v  --workaround-gcc296-bugs=yes  --log-file=${CODE_DIR}/../script/valgrind.log $PROGRAM -c $CFG_DIR -f $FILE_NAME
+  else
+    echo "ExcuteCmd:" $PROGRAM -c $CFG_DIR -f $FILE_NAME
     $PROGRAM -c $CFG_DIR -f $FILE_NAME
+  fi
 else
   echo "ExcuteCmd:" $PROGRAM -c $CFG_DIR -f basicSuite.sim
   echo "------------------------------------------------------------------------"

tests/script/unique/dnode/balance1.sim

@@ -25,18 +25,18 @@ system sh/cfg.sh -n dnode2 -c mgmtEqualVnodeNum -v 4
 system sh/cfg.sh -n dnode3 -c mgmtEqualVnodeNum -v 4
 system sh/cfg.sh -n dnode4 -c mgmtEqualVnodeNum -v 4
 
-system sh/cfg.sh -n dnode1 -c clog -v 1
-system sh/cfg.sh -n dnode2 -c clog -v 1
-system sh/cfg.sh -n dnode3 -c clog -v 1
-system sh/cfg.sh -n dnode4 -c clog -v 1
+system sh/cfg.sh -n dnode1 -c clog -v 2
+system sh/cfg.sh -n dnode2 -c clog -v 2
+system sh/cfg.sh -n dnode3 -c clog -v 2
+system sh/cfg.sh -n dnode4 -c clog -v 2
 
-system sh/cfg.sh -n dnode1 -c clog -v 1
-system sh/cfg.sh -n dnode2 -c clog -v 1
-system sh/cfg.sh -n dnode3 -c clog -v 1
-system sh/cfg.sh -n dnode4 -c clog -v 1
+system sh/cfg.sh -n dnode1 -c clog -v 2
+system sh/cfg.sh -n dnode2 -c clog -v 2
+system sh/cfg.sh -n dnode3 -c clog -v 2
+system sh/cfg.sh -n dnode4 -c clog -v 2
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sleep 3000
 
@@ -49,15 +49,15 @@ sql insert into d1.t1 values(now+4s, 12)
 sql insert into d1.t1 values(now+5s, 11)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 1 then
 	return -1
 endi
 
 print ========== step2
 sleep 2000
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show2: 
@@ -68,12 +68,12 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 0 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 3 then
+if $data3_2 != 1 then
 	goto show2
 endi
 
@@ -87,12 +87,12 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 0 then
 	return -1
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	return -1
 endi
 
@@ -108,23 +108,23 @@ show4:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 2 then
     goto show4
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
     goto show4
 endi
 if $rows != 1 then
     goto show4
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step5
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 
 $x = 0
 show5: 
@@ -135,16 +135,16 @@ show5:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 0 then
   goto show5
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
   goto show5
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
   goto show5
 endi
 
@@ -158,23 +158,23 @@ sql insert into d3.t3 values(now+4s, 32)
 sql insert into d3.t3 values(now+5s, 31)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 0 then
 	return -1
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	return -1
 endi
-if $data3_192.168.0.3 != 1 then
+if $data3_3 != 3 then
 	return -1
 endi
 
 print ========== step7
 sql create dnode 192.168.0.4
-system sh/exec.sh -n dnode4 -s start
+system sh/exec_up.sh -n dnode4 -s start
 
 $x = 0
 show7: 
@@ -185,20 +185,20 @@ show7:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+if $data3_1 != 0 then
   goto show7
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
   goto show7
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
   goto show7
 endi
-if $data3_192.168.0.4 != 3 then
+if $data3_4 != 1 then
   goto show7
 endi
 
@@ -212,21 +212,21 @@ sql insert into d4.t4 values(now+4s, 42)
 sql insert into d4.t4 values(now+5s, 41)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 0 then
 	return -1
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	return -1
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	return -1
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	return -1
 endi
 
@@ -242,25 +242,25 @@ show9:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 0 then
   goto show9
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
   goto show9
 endi
-if $data3_192.168.0.3 != null then
+if $data3_3 != null then
   goto show9
 endi
-if $data3_192.168.0.4 != 0 then
+if $data3_4 != 4 then
   goto show9
 endi
 
-system sh/exec.sh -n dnode3 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode3 -s stop -x SIGINT
 
 print ========== step10
 sql select * from d1.t1 order by t desc

tests/script/unique/dnode/balance2.sim

@@ -37,13 +37,13 @@ system sh/cfg.sh -n dnode4 -c clog -v 1
 system sh/cfg.sh -n dnode5 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 
 sql create dnode 192.168.0.2
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode2 -s start
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode3 -s start
 sleep 3000
 
 sql create database d1 replica 2  tables 4
@@ -63,16 +63,16 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 4 then
 	return -1
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	return -1
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	return -1
 endi
 
@@ -88,24 +88,24 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 2 then
 	goto show2
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show2
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show2
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step3
 sql create dnode 192.168.0.4
-system sh/exec.sh -n dnode4 -s start
+system sh/exec_up.sh -n dnode4 -s start
 
 $x = 0
 show3: 
@@ -116,20 +116,20 @@ show3:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+if $data3_1 != 4 then
 	goto show3
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show3
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show3
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	goto show3
 endi
 
@@ -143,26 +143,26 @@ sql insert into d3.t3 values(now+4s, 32)
 sql insert into d3.t3 values(now+5s, 31)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+if $data3_1 != 4 then
 	return -1
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	return -1
 endi
-if $data3_192.168.0.3 != 1 then
+if $data3_3 != 1 then
 	return -1
 endi
-if $data3_192.168.0.4 != 1 then
+if $data3_4 != 1 then
 	return -1
 endi
 
 print ========== step5
 sql create dnode 192.168.0.5
-system sh/exec.sh -n dnode5 -s start
+system sh/exec_up.sh -n dnode5 -s start
 
 $x = 0
 show5: 
@@ -173,24 +173,24 @@ show5:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
+if $data3_1 != 4 then
 	goto show5
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show5
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show5
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	goto show5
 endi
-if $data3_192.168.0.5 != 2 then
+if $data3_5 != 2 then
 	goto show5
 endi
 
@@ -206,28 +206,28 @@ show6:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
+if $data3_1 != 4 then
 	goto show6
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show6
 endi
-if $data3_192.168.0.3 != null then
+if $data3_3 != null then
 	goto show6
 endi
-if $data3_192.168.0.4 != 1 then
+if $data3_4 != 1 then
 	goto show6
 endi
-if $data3_192.168.0.5 != 1 then
+if $data3_5 != 1 then
 	goto show6
 endi
 
-system sh/exec.sh -n dnode3 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode3 -s stop -x SIGINT
 
 print ========== step7
 sql select * from d1.t1 order by t desc

tests/script/unique/dnode/balance3.sim

@@ -43,15 +43,15 @@ system sh/cfg.sh -n dnode5 -c clog -v 1
 system sh/cfg.sh -n dnode6 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 
 sql create dnode 192.168.0.2
 sql create dnode 192.168.0.3
 sql create dnode 192.168.0.4
-system sh/exec.sh -n dnode2 -s start
-system sh/exec.sh -n dnode3 -s start
-system sh/exec.sh -n dnode4 -s start
+system sh/exec_up.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode4 -s start
 sleep 3000
 
 sql create database d1 replica 3  tables 4
@@ -71,21 +71,21 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 4 then
 	return -1
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	return -1
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	return -1
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	return -1
 endi
 
@@ -101,29 +101,29 @@ show2:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
 
-if $data3_192.168.0.1 != 2 then
+if $data3_1 != 2 then
 	goto show2
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show2
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show2
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	goto show2
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step
 sql create dnode 192.168.0.5
-system sh/exec.sh -n dnode5 -s start
+system sh/exec_up.sh -n dnode5 -s start
 
 $x = 0
 show3: 
@@ -134,25 +134,25 @@ show3:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 4 then
 	goto show3
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show3
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show3
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	goto show3
 endi
-if $data3_192.168.0.5 != 2 then
+if $data3_5 != 2 then
 	goto show3
 endi
 
@@ -174,31 +174,31 @@ show4:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 4 then
 	goto show4
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show4
 endi
-if $data3_192.168.0.3 != 1 then
+if $data3_3 != 1 then
 	goto show4
 endi
-if $data3_192.168.0.4 != 1 then
+if $data3_4 != 1 then
 	goto show4
 endi
-if $data3_192.168.0.5 != 1 then
+if $data3_5 != 1 then
 	goto show4
 endi
 
 print ========== step5
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show5: 
@@ -209,16 +209,16 @@ show5:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 4 then
 	goto show5
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	goto show5
 endi
 
@@ -236,29 +236,29 @@ show6:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-print 192.168.0.5 freeVnodes $data3_192.168.0.5
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+print 192.168.0.5 openVnodes $data3_5
 
-if $data3_192.168.0.1 != 4 then
+if $data3_1 != 4 then
 	goto show6
 endi
-if $data3_192.168.0.2 != 1 then
+if $data3_2 != 1 then
 	goto show6
 endi
-if $data3_192.168.0.3 != null then
+if $data3_3 != null then
 	goto show6
 endi
-if $data3_192.168.0.4 != 1 then
+if $data3_4 != 1 then
 	goto show6
 endi
-if $data3_192.168.0.5 != 1 then
+if $data3_5 != 1 then
 	goto show6
 endi
 
-system sh/exec.sh -n dnode3 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode3 -s stop -x SIGINT
 
 print ========== step7
 sql select * from d1.t1 order by t desc

tests/script/unique/dnode/balancex.sim

@@ -31,7 +31,7 @@ system sh/cfg.sh -n dnode3 -c clog -v 1
 system sh/cfg.sh -n dnode4 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sleep 3000
 
@@ -52,14 +52,14 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 2 then
 	return -1
 endi
 
 print ========== step2
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show2: 
@@ -70,12 +70,12 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	goto show2
 endi
 
@@ -96,18 +96,18 @@ show3:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 3 then
 	goto show3
 endi
-if $data3_192.168.0.2 != 1 then
+if $data3_2 != 1 then
 	goto show3
 endi
 
 print ========== step3
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 
 $x = 0
 show4: 
@@ -117,16 +117,16 @@ show4:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 4 then
 	goto show4
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	goto show4
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show4
 endi
 
@@ -141,20 +141,20 @@ show5:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 3 then
 	goto show5
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show5
 endi
-if $data3_192.168.0.3 != 1 then
+if $data3_3 != 1 then
 	goto show5
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step6
 sql select * from d1.t1 order by t desc

tests/script/unique/dnode/monitor_bug.sim

@@ -22,19 +22,19 @@ system sh/cfg.sh -n dnode1 -c monitor -v 1
 system sh/cfg.sh -n dnode2 -c monitor -v 0
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sleep 5000
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 3 then
 	return -1
 endi
 
 print ========== step2
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show2: 
@@ -45,12 +45,12 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 3 then
+if $data3_2 != 3 then
 	goto show2
 endi
 

tests/script/unique/dnode/offline1.sim

@@ -29,10 +29,10 @@ system sh/cfg.sh -n dnode2 -c clog -v 1
 system sh/cfg.sh -n dnode3 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 sleep 3000
 
 sql show dnodes
@@ -44,7 +44,7 @@ if $data4_192.168.0.2 != ready then
 endi
 
 print ========== step2
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 sleep 6000
 
 sql show dnodes

tests/script/unique/dnode/offline2.sim

@@ -29,10 +29,10 @@ system sh/cfg.sh -n dnode2 -c clog -v 1
 system sh/cfg.sh -n dnode3 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 sleep 3000
 
 sql create database d1 replica 2 tables 4
@@ -48,7 +48,7 @@ if $data4_192.168.0.2 != ready then
 endi
 
 print ========== step2
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 sleep 5000
 
 sql show dnodes
@@ -72,7 +72,7 @@ endi
 
 print ========== step4
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 sql drop dnode 192.168.0.2
 
 sleep 5000

tests/script/unique/dnode/remove1.sim

@@ -31,7 +31,7 @@ system sh/cfg.sh -n dnode3 -c clog -v 1
 system sh/cfg.sh -n dnode4 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sleep 3000
 
@@ -52,14 +52,14 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 2 then
 	return -1
 endi
 
 print ========== step2
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 sleep 9000
 
 sql create database d3 replica 2 tables 4
@@ -79,12 +79,12 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 3 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 1 then
+if $data3_2 != 1 then
 	goto show2
 endi
 
@@ -101,12 +101,12 @@ show3:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2 $data5_192.168.0.2
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2 $data5_192.168.0.2
 
 print ========== step4
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 
 $x = 0
 show4: 
@@ -117,18 +117,18 @@ show4:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.2 != null then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_2 != null then
 	goto show4
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step5
 sql create dnode 192.168.0.4
-system sh/exec.sh -n dnode4 -s start
+system sh/exec_up.sh -n dnode4 -s start
 
 $x = 0
 show5: 
@@ -138,20 +138,20 @@ show5:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-print 192.168.0.4 freeVnodes $data3_192.168.0.4
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+print 192.168.0.4 openVnodes $data3_4
+if $data3_1 != 4 then
 	goto show5
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
 	goto show5
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
 	goto show5
 endi
-if $data3_192.168.0.4 != 2 then
+if $data3_4 != 2 then
 	goto show5
 endi
 

tests/script/unique/dnode/remove2.sim

@@ -31,7 +31,7 @@ system sh/cfg.sh -n dnode3 -c clog -v 1
 system sh/cfg.sh -n dnode4 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 sleep 3000
 
@@ -52,14 +52,14 @@ sql insert into d2.t2 values(now+4s, 22)
 sql insert into d2.t2 values(now+5s, 21)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 2 then
 	return -1
 endi
 
 print ========== step2
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 sleep 9000
 
 sql create database d3 replica 2 tables 4
@@ -79,17 +79,17 @@ show2:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 3 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 1 then
+if $data3_2 != 1 then
 	goto show2
 endi
 
 print ========== step3
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 sql drop dnode 192.168.0.2
 sleep 7001
 
@@ -102,12 +102,12 @@ show3:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2 $data5_192.168.0.2
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2 $data5_192.168.0.2
 
 print ========== step4
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 
 $x = 0
 show4: 
@@ -118,16 +118,16 @@ show4:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.2 != null then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_2 != null then
 	goto show4
 endi
-if $data3_192.168.0.1 != 3 then
+if $data3_1 != 3 then
 	goto show4
 endi
-if $data3_192.168.0.3 != 1 then
+if $data3_3 != 1 then
 	goto show4
 endi
 

tests/script/unique/dnode/vnode_clean.sim

@@ -31,7 +31,7 @@ system sh/cfg.sh -n dnode3 -c clog -v 1
 system sh/cfg.sh -n dnode4 -c clog -v 1
 
 print ========== step1
-system sh/exec.sh -n dnode1 -s start
+system sh/exec_up.sh -n dnode1 -s start
 sql connect
 
 sql create database d1 tables 4
@@ -43,14 +43,14 @@ sql insert into d1.t1 values(now+4s, 12)
 sql insert into d1.t1 values(now+5s, 11)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-if $data3_192.168.0.1 != 3 then
+print 192.168.0.1 openVnodes $data3_1
+if $data3_1 != 3 then
 	return -1
 endi
 
 print ========== step2
 sql create dnode 192.168.0.2
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show2: 
@@ -60,12 +60,12 @@ show2:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
 	goto show2
 endi
-if $data3_192.168.0.2 != 3 then
+if $data3_2 != 3 then
 	goto show2
 endi
 
@@ -81,12 +81,12 @@ sql insert into d2.t2 values(now+5s, 21)
 
 $x = 0
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
 	return -1
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
 	return -1
 endi
 
@@ -101,19 +101,19 @@ show4:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 2 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 2 then
     goto show4
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
     goto show4
 endi
 if $rows != 1 then
     goto show4
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step5
 sleep 2000
@@ -125,7 +125,7 @@ system sh/cfg.sh -n dnode2 -c balanceMonitorInterval -v 1
 system sh/cfg.sh -n dnode2 -c balanceStartInterval -v 10
 system sh/cfg.sh -n dnode2 -c mgmtEqualVnodeNum -v 4
 system sh/cfg.sh -n dnode2 -c clog -v 1
-system sh/exec.sh -n dnode2 -s start
+system sh/exec_up.sh -n dnode2 -s start
 
 $x = 0
 show5: 
@@ -135,12 +135,12 @@ show5:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
     goto show5
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
     goto show5
 endi
 
@@ -154,18 +154,18 @@ sql insert into d3.t3 values(now+4s, 32)
 sql insert into d3.t3 values(now+5s, 31)
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+if $data3_1 != 4 then
 	return -1
 endi
-if $data3_192.168.0.2 != 1 then
+if $data3_2 != 1 then
 	return -1
 endi
 
 print ========== step7
 sql create dnode 192.168.0.3
-system sh/exec.sh -n dnode3 -s start
+system sh/exec_up.sh -n dnode3 -s start
 
 $x = 0
 show7: 
@@ -176,16 +176,16 @@ show7:
 	endi
 	
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 4 then
   goto show7
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
   goto show7
 endi
-if $data3_192.168.0.3 != 3 then
+if $data3_3 != 3 then
   goto show7
 endi
 
@@ -206,16 +206,16 @@ show8:
 	  return -1
 	endi
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 4 then
   goto show8
 endi
-if $data3_192.168.0.2 != 2 then
+if $data3_2 != 2 then
   goto show8
 endi
-if $data3_192.168.0.3 != 2 then
+if $data3_3 != 2 then
   goto show8
 endi
 
@@ -231,20 +231,20 @@ show9:
 	endi
 
 sql show dnodes
-print 192.168.0.1 freeVnodes $data3_192.168.0.1
-print 192.168.0.2 freeVnodes $data3_192.168.0.2
-print 192.168.0.3 freeVnodes $data3_192.168.0.3
-if $data3_192.168.0.1 != 4 then
+print 192.168.0.1 openVnodes $data3_1
+print 192.168.0.2 openVnodes $data3_2
+print 192.168.0.3 openVnodes $data3_3
+if $data3_1 != 4 then
   goto show9
 endi
-if $data3_192.168.0.2 != null then
+if $data3_2 != null then
   goto show9
 endi
-if $data3_192.168.0.3 != 0 then
+if $data3_3 != 0 then
   goto show9
 endi
 
-system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec_up.sh -n dnode2 -s stop -x SIGINT
 
 print ========== step10
 sql select * from d1.t1  order by t desc

tests/script/unique/mnode/mgmt22.sim

@@ -109,3 +109,7 @@ endi
 if $data3_3 != null then
 	goto show7
 endi
+
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/mgmt24.sim

@@ -78,3 +78,7 @@ endi
 if $data3_2 != slave then
 	goto step5
 endi
+
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/mgmt25.sim

@@ -93,3 +93,7 @@ endi
 if $dnode3Role != slave then
   return -1
 endi
+
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/mgmt26.sim

@@ -123,3 +123,6 @@ if $dnode3Role != slave then
   return -1
 endi
 
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/mgmtr2.sim

@@ -83,3 +83,7 @@ endi
 if $dnode3Role != null then
   return -1
 endi
+
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/secondIp.sim

@@ -38,5 +38,6 @@ if $data4_2 != 4 then
 	return -1
 endi
 
-
-
+system sh/exec_up.sh -n dnode1 -s stop
+system sh/exec_up.sh -n dnode2 -s stop
+system sh/exec_up.sh -n dnode3 -s stop
\ No newline at end of file

tests/script/unique/mnode/testSuite.sim

@@ -7,4 +7,3 @@ run unique/mnode/mgmt33.sim
 run unique/mnode/mgmt34.sim
 run unique/mnode/mgmtr2.sim
 run unique/mnode/secondIp.sim
-

tests/test-all.sh

@@ -8,32 +8,33 @@ GREEN_UNDERLINE='\033[4;32m'
 NC='\033[0m'
 
 cd script
-sudo ./test.sh 2>&1 | grep 'success\|failed' | tee out.txt
+./test.sh -f basicSuite.sim 2>&1 | grep 'success\|failed\|fault' | tee out.txt
 
-total_success=`grep success out.txt | wc -l`
+totalSuccess=`grep success out.txt | wc -l`
+totalBasic=`grep success out.txt | grep Suite | wc -l`
 
-if [ "$total_success" -gt "0" ]; then
-  total_success=`expr $total_success - 1`
-  echo -e "${GREEN} ### Total $total_success TSIM case(s) succeed! ### ${NC}"
+if [ "$totalSuccess" -gt "0" ]; then
+  totalSuccess=`expr $totalSuccess - $totalBasic`
+  echo -e "${GREEN} ### Total $totalSuccess TSIM case(s) succeed! ### ${NC}"
 fi
 
-total_failed=`grep failed out.txt | wc -l`
-if [ "$total_failed" -ne "0" ]; then
-  echo -e "${RED} ### Total $total_failed TSIM case(s) failed! ### ${NC}"
-  exit $total_failed
+totalFailed=`grep 'failed\|fault' out.txt | wc -l`
+if [ "$totalFailed" -ne "0" ]; then
+  echo -e "${RED} ### Total $totalFailed TSIM case(s) failed! ### ${NC}"
+  exit $totalFailed
 fi
 
 cd ../pytest
-sudo ./simpletest.sh 2>&1 | grep 'successfully executed\|failed' | tee pytest-out.txt
-total_py_success=`grep 'successfully executed' pytest-out.txt | wc -l`
+./simpletest.sh 2>&1 | grep 'successfully executed\|failed' | tee pytest-out.txt
+totalPySuccess=`grep 'successfully executed' pytest-out.txt | wc -l`
 
-if [ "$total_py_success" -gt "0" ]; then
-  echo -e "${GREEN} ### Total $total_py_success python case(s) succeed! ### ${NC}"
+if [ "$totalPySuccess" -gt "0" ]; then
+  echo -e "${GREEN} ### Total $totalPySuccess python case(s) succeed! ### ${NC}"
 fi
 
-total_py_failed=`grep 'failed' pytest-out.txt | wc -l`
-if [ "$total_py_failed" -ne "0" ]; then
-  echo -e "${RED} ### Total $total_py_failed python case(s) failed! ### ${NC}"
-  exit $total_py_failed
+totalPyFailed=`grep 'failed' pytest-out.txt | wc -l`
+if [ "$totalPyFailed" -ne "0" ]; then
+  echo -e "${RED} ### Total $totalPyFailed python case(s) failed! ### ${NC}"
+  exit $totalPyFailed
 fi