Merge branch 'v3.0' into fix/tdb-maxLocal

docs/en/14-reference/03-connector/rust.mdx

@@ -250,7 +250,7 @@ The [Taos] structure is the connection manager in [libtaos] and provides two mai
 
     Column information is stored using [ColumnMeta].
 
-    ``rust
+    ```rust
     let cols = &q.column_meta;
     for col in cols {
         println!("name: {}, type: {:?} , bytes: {}", col.name, col.type_, col.bytes);

docs/en/14-reference/05-taosbenchmark.md

@@ -7,7 +7,7 @@ description: "taosBenchmark (once called taosdemo ) is a tool for testing the pe
 
 ## Introduction
 
-taosBenchmark (formerly taosdemo ) is a tool for testing the performance of TDengine products. taosBenchmark can test the performance of TDengine's insert, query, and subscription functions and simulate large amounts of data generated by many devices. taosBenchmark can flexibly control the number and type of databases, supertables, tag columns, number and type of data columns, and sub-tables, and types of databases, super tables, the number and types of data columns, the number of sub-tables, the amount of data per sub-table, the time interval for inserting data, the number of working threads, whether and how to insert disordered data, and so on. The installer provides taosdemo as a soft link to taosBenchmark for compatibility and for the convenience of past users.
+taosBenchmark (formerly taosdemo ) is a tool for testing the performance of TDengine products. taosBenchmark can test the performance of TDengine's insert, query, and subscription functions and simulate large amounts of data generated by many devices. taosBenchmark can be configured to generate user defined databases, supertables, subtables, and the time series data to populate these for performance benchmarking. taosBenchmark is highly configurable and some of the configurations include the time interval for inserting data, the number of working threads and the capability to insert disordered data. The installer provides taosdemo as a soft link to taosBenchmark for compatibility with past users.
 
 ## Installation
 
@@ -21,9 +21,13 @@ There are two ways to install taosBenchmark:
 
 ### Configuration and running methods
 
-TaosBenchmark needs to be executed on the terminal of the operating system, it supports two configuration methods: [Command-line arguments](#Command-line arguments in detailed) and [JSON configuration file](#Configuration file arguments in detailed). These two methods are mutually exclusive. Users can use `-f <json file>` to specify a configuration file. When running taosBenchmark with command-line arguments to control its behavior, users should use other parameters for configuration, but not the `-f` parameter. In addition, taosBenchmark offers a special way of running without parameters.
+TaosBenchmark needs to be executed on the terminal of the operating system, it supports two configuration methods: [Command-line arguments](#command-line-arguments-in-detail) and [JSON configuration file](#configuration-file-parameters-in-detail). These two methods are mutually exclusive. Users can use `-f <json file>` to specify a configuration file. When running taosBenchmark with command-line arguments to control its behavior, users should use other parameters for configuration, but not the `-f` parameter. In addition, taosBenchmark offers a special way of running without parameters.
 
+<<<<<<< HEAD
 taosBenchmark supports complete performance testing of TDengine. taosBenchmark supports the TDengine functions in three categories: write, query, and subscribe. These three functions are mutually exclusive, and users can select only one of them each time taosBenchmark runs. It is important to note that the type of functionality to be tested is not configurable when using the command-line configuration method, which can only test writing performance. To test the query and subscription performance of the TDengine, you must use the configuration file method and specify the function type to test via the parameter `filetype` in the configuration file.
+=======
+taosBenchmark supports the complete performance testing of TDengine by providing functionally to write, query, and subscribe. These three functions are mutually exclusive, users can only select one of them each time taosBenchmark runs. The query and subscribe functionalities are only configurable using a json configuration file by specifying the parameter `filetype`, while write can be performed through both the command-line and a configuration file.
+>>>>>>> 108548b4d6 (docs: typo)
 
 **Make sure that the TDengine cluster is running correctly before running taosBenchmark. **
 
@@ -57,9 +61,8 @@ Use the following command-line to run taosBenchmark and control its behavior via
 taosBenchmark -f <json file>
 ```
 
-**Here are a few examples of configuration files:**
+#### Configuration file examples
-
+##### Example of inserting a scenario JSON configuration file
-#### Example of inserting a scenario JSON configuration file
 
 <details>
 <summary>insert.json</summary>
@@ -70,7 +73,7 @@ taosBenchmark -f <json file>
 
 </details>
 
-#### Query Scenario JSON Profile Example
+##### Query Scenario JSON Profile Example
 
 <details>
 <summary>query.json</summary>
@@ -81,7 +84,7 @@ taosBenchmark -f <json file>
 
 </details>
 
-#### Subscription JSON configuration example
+##### Subscription JSON configuration example
 
 <details>
 <summary>subscribe.json</summary>
@@ -172,7 +175,7 @@ taosBenchmark -A INT,DOUBLE,NCHAR,BINARY\(16\)
   Switch parameter specifying whether to use escape characters in the super table and sub-table names. By default is not used.
 
 - **-C/--chinese** :
-  Switch specifying whether to use Unicode Chinese characters in nchar and binary. By default is not used.
+  specify whether to use Unicode Chinese characters in nchar and binary, the default is no.
 
 - **-N/--normal-table** :
   This parameter indicates that taosBenchmark will create only normal tables instead of super tables. The default value is false. It can be used if the insert mode is taosc, stmt, and rest.
@@ -373,7 +376,7 @@ The configuration parameters for querying the sub-tables or the normal tables ar
 
 - **sqls**.
   - **sql**: the SQL command to be executed.
-  - **result**: the file to save the query result. If it is unspecified, taosBenchark will not save the result.
+  - **result**: the file to save the query result. If it is unspecified, taosBenchmark will not save the result.
 
 #### Configuration parameters of query super table
 

docs/en/14-reference/06-taosdump.md

@@ -36,6 +36,8 @@ There are two ways to install taosdump:
 :::tip
 - taosdump versions after 1.4.1 provide the `-I` argument for parsing Avro file schema and data. If users specify `-s` then only taosdump will parse schema.
 - Backups after taosdump 1.4.2 use the batch count specified by the `-B` parameter. The default value is 16384. If, in some environments, low network speed or disk performance causes "Error actual dump ... batch ...", then try changing the `-B` parameter to a smaller value.
+- The export of taosdump does not support resuming from an interruption. Therefore, if the taosdump process terminates unexpectedly, delete all related files that have been exported or generated.
+- The import of taosdump supports resuming from an interruption, but when the process resumes, you will receive some "table already exists" messages, which could be ignored.
 
 :::
 

docs/en/20-third-party/01-grafana.mdx

@@ -31,38 +31,41 @@ TDengine currently supports Grafana versions 7.5 and above. Users can go to the
 ### Install Grafana Plugin and Configure Data Source
 
 <Tabs defaultValue="script">
-<TabItem value="script" label="Using Script">
+<TabItem value="gui" label="With GUI">
 
-Set the url and authorization environment variables by `export` or a [`.env`(dotenv) file](https://hexdocs.pm/dotenvy/dotenv-file-format.html):
+Under Grafana 8, plugin catalog allows you to [browse and manage plugins within Grafana](https://grafana.com/docs/grafana/next/administration/plugin-management/#plugin-catalog) (but for Grafana 7.x, use **With Script** or **Install & Configure Manually**). Find the page at **Configurations > Plugins**, search **TDengine** and click it to install.
 
-```sh
+![Search tdengine in grafana plugins](./grafana/grafana-plugin-search-tdengine.png)
-export TDENGINE_API=http://tdengine.local:6041
-# user + password
-export TDENGINE_USER=user
-export TDENGINE_PASSWORD=password
 
-# Other useful variables
+Installation may cost some minutes, then you can **Create a TDengine data source**:
-# - If to install TDengine data source, default is true
-export TDENGINE_DS_ENABLED=false
-# - Data source name to be created, default is TDengine
-export TDENGINE_DS_NAME=TDengine
-# - Data source organization id, default is 1
-export GF_ORG_ID=1
-# - Data source is editable in admin ui or not, default is 0 (false)
-export TDENGINE_EDITABLE=1
-```
 
-Run `install.sh`:
+![Install and configure Grafana data source](./grafana/grafana-install-and-config.png)
 
-```sh
+Then you can add a TDengine data source by filling up the configuration options.
-bash -c "$(curl -fsSL https://raw.githubusercontent.com/taosdata/grafanaplugin/master/install.sh)"
-```
 
-With this script, TDengine data source plugin and the Grafana data source will be installed and created automatically with Grafana provisioning configurations. Save the script and type `./install.sh --help` for the full usage of the script.
+![TDengine Database Grafana plugin add data source](./grafana/grafana-data-source.png)
 
-And then, restart Grafana service and open Grafana in web-browser, usually <http://localhost:3000>.
+You can create dashboards with TDengine now.
 
 </TabItem>
+<TabItem value="script" label="With Script">
+
+On a server with Grafana installed, run `install.sh` with TDengine url and username/passwords will install TDengine data source plugin and add a data source named TDengine. This is the recommended way for Grafana 7.x or [Grafana provisioning](https://grafana.com/docs/grafana/latest/administration/provisioning/) users.
+
+```sh
+bash -c "$(curl -fsSL \
+  https://raw.githubusercontent.com/taosdata/grafanaplugin/master/install.sh)" -- \
+  -a http://localhost:6041 \
+  -u root \
+  -p taosdata
+```
+
+Restart Grafana service and open Grafana in web-browser, usually <http://localhost:3000>.
+
+Save the script and type `./install.sh --help` for the full usage of the script.
+
+</TabItem>
+
 <TabItem value="manual" label="Install & Configure Manually">
 
 Follow the installation steps in [Grafana](https://grafana.com/grafana/plugins/tdengine-datasource/?tab=installation) with the [``grafana-cli`` command-line tool](https://grafana.com/docs/grafana/latest/administration/cli/) for plugin installation.
@@ -115,6 +118,73 @@ Click `Save & Test` to test. You should see a success message if the test worked
 
 ![TDengine Database TDinsight plugin add database 4](./grafana/add_datasource4.webp)
 
+</TabItem>
+<TabItem value="container" label="Container">
+
+Please refer to [Install plugins in the Docker container](https://grafana.com/docs/grafana/next/setup-grafana/installation/docker/#install-plugins-in-the-docker-container). This will install `tdengine-datasource` plugin when Grafana container starts:
+
+```bash
+docker run -d \
+  -p 3000:3000 \
+  --name=grafana \
+  -e "GF_INSTALL_PLUGINS=tdengine-datasource" \
+  grafana/grafana
+```
+
+You can setup a zero-configuration stack for TDengine + Grafana by [docker-compose](https://docs.docker.com/compose/) and [Grafana provisioning](https://grafana.com/docs/grafana/latest/administration/provisioning/) file：
+
+1. Save the provisioning configuration file to `tdengine.yml`.
+
+    ```yml
+    apiVersion: 1
+    datasources:
+    - name: TDengine
+      type: tdengine-datasource
+      orgId: 1
+      url: "$TDENGINE_API"
+      isDefault: true
+      secureJsonData:
+        url: "$TDENGINE_URL"
+        basicAuth: "$TDENGINE_BASIC_AUTH"
+        token: "$TDENGINE_CLOUD_TOKEN"
+      version: 1
+      editable: true
+    ```
+
+2. Write `docker-compose.yml` with [TDengine](https://hub.docker.com/r/tdengine/tdengine) and [Grafana](https://hub.docker.com/r/grafana/grafana) image.
+
+    ```yml
+    version: "3.7"
+
+    services:
+      tdengine:
+        image: tdengine/tdengine:2.6.0.2
+        environment:
+          TAOS_FQDN: tdengine
+        volumes:
+          - tdengine-data:/var/lib/taos/
+      grafana:
+        image: grafana/grafana:8.5.6
+        volumes:
+          - ./tdengine.yml/:/etc/grafana/provisioning/tdengine.yml
+          - grafana-data:/var/lib/grafana
+        environment:
+          # install tdengine plugin at start
+          GF_INSTALL_PLUGINS: "tdengine-datasource"
+          TDENGINE_URL: "http://tdengine:6041"
+          #printf "$TDENGINE_USER:$TDENGINE_PASSWORD" | base64
+          TDENGINE_BASIC_AUTH: "cm9vdDp0YmFzZTEyNQ=="
+        ports:
+          - 3000:3000
+    volumes:
+      grafana-data:
+      tdengine-data:
+    ```
+
+3. Start TDengine and Grafana services: `docker-compose up -d`.
+
+Open Grafana <http://localhost:3000>, and you can add dashboard with TDengine now.
+
 </TabItem>
 </Tabs>
 

docs/zh/14-reference/06-taosdump.md

@@ -39,6 +39,8 @@ taosdump 有两种安装方式:
 :::tip
 - taosdump 1.4.1 之后的版本提供 `-I` 参数，用于解析 avro 文件 schema 和数据，如果指定 `-s` 参数将只解析 schema。
 - taosdump 1.4.2 之后的备份使用 `-B` 参数指定的批次数，默认值为 16384，如果在某些环境下由于网络速度或磁盘性能不足导致 "Error actual dump .. batch .." 可以通过 `-B` 参数调整为更小的值进行尝试。
+- taosdump 的导出不支持中断恢复，所以当进程意外终止后，正确的处理方式是删除当前已导出或生成的所有相关文件。
+- taosdump 的导入支持中断恢复，但是当进程重新启动时，会收到一些“表已经存在”的提示，可以忽视。
 
 :::
 

docs/zh/20-third-party/01-grafana.mdx

@@ -29,39 +29,41 @@ TDengine 能够与开源数据可视化系统 [Grafana](https://www.grafana.com/
 ### 安装 Grafana Plugin 并配置数据源
 
 <Tabs defaultValue="script">
+<TabItem value="gui" label="图形化界面安装">
+
+使用 Grafana 最新版本（8.5+），您可以在 Grafana 中[浏览和管理插件](https://grafana.com/docs/grafana/next/administration/plugin-management/#plugin-catalog)（对于 7.x 版本，请使用 **安装脚本** 或 **手动安装并配置** 方式）。在 Grafana 管理界面中的 **Configurations > Plugins** 页面直接搜索并按照提示安装 TDengine。
+
+![Search tdengine in grafana plugins](grafana-plugin-search-tdengine.png)
+
+如图示即安装完毕，按照指示 **Create a TDengine data source** 添加数据源。
+
+![Install and configure Grafana data source](grafana-install-and-config.png)
+
+输入 TDengine 相关配置，完成数据源配置。
+
+![TDengine Database Grafana plugin add data source](./grafana-data-source.png)
+
+配置完毕，现在可以使用 TDengine 创建 Dashboard 了。
+
+</TabItem>
 <TabItem value="script" label="使用安装脚本">
 
-将集群信息设置为环境变量；也可以使用 `.env` 文件，请参考 [dotenv](https://hexdocs.pm/dotenvy/dotenv-file-format.html)：
+对于使用 Grafana 7.x 版本或使用 [Grafana Provisioning](https://grafana.com/docs/grafana/latest/administration/provisioning/) 配置的用户，可以在 Grafana 服务器上使用安装脚本自动安装插件即添加数据源 Provisioning 配置文件。
 
 ```sh
-export TDENGINE_API=http://tdengine.local:6041
+bash -c "$(curl -fsSL \
-# user + password
+  https://raw.githubusercontent.com/taosdata/grafanaplugin/master/install.sh)" -- \
-export TDENGINE_USER=user
+  -a http://localhost:6041 \
-export TDENGINE_PASSWORD=password
+  -u root \
-
+  -p taosdata
-# 其他环境变量：
-# - 是否安装数据源，默认为 true，表示安装
-export TDENGINE_DS_ENABLED=false
-# - 数据源名称，默认为 TDengine
-export TDENGINE_DS_NAME=TDengine
-# - 数据源所属组织 ID，默认为 1
-export GF_ORG_ID=1
-# - 数据源是否可通过管理面板编辑，默认为 0，表示不可编辑
-export TDENGINE_EDITABLE=1
 ```
 
-运行安装脚本：
+安装完毕后，需要重启 Grafana 服务后方可生效。
-
-```sh
-bash -c "$(curl -fsSL https://raw.githubusercontent.com/taosdata/grafanaplugin/master/install.sh)"
-```
-
-该脚本将自动安装 Grafana 插件并配置数据源。安装完毕后，需要重启 Grafana 服务后生效。
 
 保存该脚本并执行 `./install.sh --help` 可查看详细帮助文档。
 
 </TabItem>
-<TabItem value="manual" label="手动安装并配置">
+<TabItem value="manual" label="手动安装">
 
 使用 [`grafana-cli` 命令行工具](https://grafana.com/docs/grafana/latest/administration/cli/) 进行插件[安装](https://grafana.com/grafana/plugins/tdengine-datasource/?tab=installation)。
 
@@ -113,6 +115,73 @@ GF_INSTALL_PLUGINS=tdengine-datasource
 
 ![TDengine Database Grafana plugin add data source](./add_datasource4.webp)
 
+</TabItem>
+<TabItem value="container" label="K8s/Docker 容器">
+
+参考 [Grafana 容器化安装说明](https://grafana.com/docs/grafana/next/setup-grafana/installation/docker/#install-plugins-in-the-docker-container)。使用如下命令启动一个容器，并自动安装 TDengine 插件：
+
+```bash
+docker run -d \
+  -p 3000:3000 \
+  --name=grafana \
+  -e "GF_INSTALL_PLUGINS=tdengine-datasource" \
+  grafana/grafana
+```
+
+使用 docker-compose，配置 Grafana Provisioning 自动化配置，体验 TDengine + Grafana 组合的零配置启动：
+
+1. 保存该文件为 `tdengine.yml`。
+
+    ```yml
+    apiVersion: 1
+    datasources:
+    - name: TDengine
+      type: tdengine-datasource
+      orgId: 1
+      url: "$TDENGINE_API"
+      isDefault: true
+      secureJsonData:
+        url: "$TDENGINE_URL"
+        basicAuth: "$TDENGINE_BASIC_AUTH"
+        token: "$TDENGINE_CLOUD_TOKEN"
+      version: 1
+      editable: true
+    ```
+
+2. 保存该文件为 `docker-compose.yml`。
+
+    ```yml
+    version: "3.7"
+
+    services:
+      tdengine:
+        image: tdengine/tdengine:2.6.0.2
+        environment:
+          TAOS_FQDN: tdengine
+        volumes:
+          - tdengine-data:/var/lib/taos/
+      grafana:
+        image: grafana/grafana:8.5.6
+        volumes:
+          - ./tdengine.yml/:/etc/grafana/provisioning/tdengine.yml
+          - grafana-data:/var/lib/grafana
+        environment:
+          # install tdengine plugin at start
+          GF_INSTALL_PLUGINS: "tdengine-datasource"
+          TDENGINE_URL: "http://tdengine:6041"
+          #printf "$TDENGINE_USER:$TDENGINE_PASSWORD" | base64
+          TDENGINE_BASIC_AUTH: "cm9vdDp0YmFzZTEyNQ=="
+        ports:
+          - 3000:3000
+    volumes:
+      grafana-data:
+      tdengine-data:
+    ```
+
+3. 使用 docker-compose 命令启动 TDengine + Grafana ：`docker-compose up -d`。
+
+打开 Grafana <http://localhost:3000>，现在可以添加 Dashboard 了。
+
 </TabItem>
 </Tabs>
 

examples/c/stream_demo.c

@@ -32,6 +32,13 @@ int32_t init_env() {
   }
   taos_free_result(pRes);
 
+  pRes = taos_query(pConn, "create database if not exists abc2 vgroups 20");
+  if (taos_errno(pRes) != 0) {
+    printf("error in create db, reason:%s\n", taos_errstr(pRes));
+    return -1;
+  }
+  taos_free_result(pRes);
+
   pRes = taos_query(pConn, "use abc1");
   if (taos_errno(pRes) != 0) {
     printf("error in use db, reason:%s\n", taos_errstr(pRes));
@@ -83,7 +90,7 @@ int32_t create_stream() {
   /*pRes = tmq_create_stream(pConn, "stream1", "out1", sql);*/
   pRes = taos_query(
       pConn,
-      "create stream stream1 trigger max_delay 10s into outstb as select _wstartts, sum(k) from st1 interval(10m)");
+      "create stream stream1 trigger at_once into abc1.outstb as select _wstartts, sum(k) from st1 interval(10m) ");
   if (taos_errno(pRes) != 0) {
     printf("failed to create stream stream1, reason:%s\n", taos_errstr(pRes));
     return -1;
@@ -100,11 +107,4 @@ int main(int argc, char* argv[]) {
     code = init_env();
   }
   create_stream();
-#if 0
-  tmq_t*      tmq = build_consumer();
-  tmq_list_t* topic_list = build_topic_list();
-  /*perf_loop(tmq, topic_list);*/
-  /*basic_consume_loop(tmq, topic_list);*/
-  sync_consume_loop(tmq, topic_list);
-#endif
 }

examples/c/tmq.c

@@ -47,7 +47,7 @@ int32_t init_env() {
     return -1;
   }
 
-  TAOS_RES* pRes = taos_query(pConn, "create database if not exists abc1 vgroups 2");
+  TAOS_RES* pRes = taos_query(pConn, "create database if not exists abc1 vgroups 1");
   if (taos_errno(pRes) != 0) {
     printf("error in create db, reason:%s\n", taos_errstr(pRes));
     return -1;
@@ -146,8 +146,8 @@ int32_t create_topic() {
   return 0;
 }
 
-void tmq_commit_cb_print(tmq_t* tmq, tmq_resp_err_t resp, tmq_topic_vgroup_list_t* offsets, void* param) {
+void tmq_commit_cb_print(tmq_t* tmq, int32_t code, void* param) {
-  printf("commit %d tmq %p offsets %p param %p\n", resp, tmq, offsets, param);
+  printf("commit %d tmq %p param %p\n", code, tmq, param);
 }
 
 tmq_t* build_consumer() {
@@ -167,7 +167,7 @@ tmq_t* build_consumer() {
   tmq_conf_set(conf, "td.connect.user", "root");
   tmq_conf_set(conf, "td.connect.pass", "taosdata");
   tmq_conf_set(conf, "msg.with.table.name", "true");
-  tmq_conf_set(conf, "enable.auto.commit", "false");
+  tmq_conf_set(conf, "enable.auto.commit", "true");
   tmq_conf_set_auto_commit_cb(conf, tmq_commit_cb_print, NULL);
   tmq_t* tmq = tmq_consumer_new(conf, NULL, 0);
   assert(tmq);
@@ -183,10 +183,10 @@ tmq_list_t* build_topic_list() {
 }
 
 void basic_consume_loop(tmq_t* tmq, tmq_list_t* topics) {
-  tmq_resp_err_t err;
+  int32_t code;
 
-  if ((err = tmq_subscribe(tmq, topics))) {
+  if ((code = tmq_subscribe(tmq, topics))) {
-    fprintf(stderr, "%% Failed to start consuming topics: %s\n", tmq_err2str(err));
+    fprintf(stderr, "%% Failed to start consuming topics: %s\n", tmq_err2str(code));
     printf("subscribe err\n");
     return;
   }
@@ -201,12 +201,13 @@ void basic_consume_loop(tmq_t* tmq, tmq_list_t* topics) {
       taos_free_result(tmqmessage);
       /*} else {*/
       /*break;*/
+      /*tmq_commit_sync(tmq, NULL);*/
     }
   }
 
-  err = tmq_consumer_close(tmq);
+  code = tmq_consumer_close(tmq);
-  if (err)
+  if (code)
-    fprintf(stderr, "%% Failed to close consumer: %s\n", tmq_err2str(err));
+    fprintf(stderr, "%% Failed to close consumer: %s\n", tmq_err2str(code));
   else
     fprintf(stderr, "%% Consumer closed\n");
 }
@@ -215,10 +216,10 @@ void sync_consume_loop(tmq_t* tmq, tmq_list_t* topics) {
   static const int MIN_COMMIT_COUNT = 1;
 
   int     msg_count = 0;
-  tmq_resp_err_t err;
+  int32_t code;
 
-  if ((err = tmq_subscribe(tmq, topics))) {
+  if ((code = tmq_subscribe(tmq, topics))) {
-    fprintf(stderr, "%% Failed to start consuming topics: %s\n", tmq_err2str(err));
+    fprintf(stderr, "%% Failed to start consuming topics: %s\n", tmq_err2str(code));
     return;
   }
 
@@ -239,14 +240,14 @@ void sync_consume_loop(tmq_t* tmq, tmq_list_t* topics) {
       msg_process(tmqmessage);
       taos_free_result(tmqmessage);
 
-      /*tmq_commit_async(tmq, NULL, tmq_commit_cb_print, NULL);*/
+      /*tmq_commit_sync(tmq, NULL);*/
       /*if ((++msg_count % MIN_COMMIT_COUNT) == 0) tmq_commit(tmq, NULL, 0);*/
     }
   }
 
-  err = tmq_consumer_close(tmq);
+  code = tmq_consumer_close(tmq);
-  if (err)
+  if (code)
-    fprintf(stderr, "%% Failed to close consumer: %s\n", tmq_err2str(err));
+    fprintf(stderr, "%% Failed to close consumer: %s\n", tmq_err2str(code));
   else
     fprintf(stderr, "%% Consumer closed\n");
 }

include/client/taos.h

@@ -192,8 +192,8 @@ DLL_EXPORT int         taos_errno(TAOS_RES *tres);
 
 DLL_EXPORT void        taos_query_a(TAOS *taos, const char *sql, __taos_async_fn_t fp, void *param);
 DLL_EXPORT void        taos_fetch_rows_a(TAOS_RES *res, __taos_async_fn_t fp, void *param);
-DLL_EXPORT void        taos_fetch_raw_block_a(TAOS_RES* res, __taos_async_fn_t fp, void* param);
+DLL_EXPORT void        taos_fetch_raw_block_a(TAOS_RES *res, __taos_async_fn_t fp, void *param);
-DLL_EXPORT const void *taos_get_raw_block(TAOS_RES* res);
+DLL_EXPORT const void *taos_get_raw_block(TAOS_RES *res);
 
 // Shuduo: temporary enable for app build
 #if 1
@@ -209,21 +209,11 @@ DLL_EXPORT TAOS_RES *taos_schemaless_insert(TAOS *taos, char *lines[], int numLi
 
 /* --------------------------TMQ INTERFACE------------------------------- */
 
-enum {
-  TMQ_RESP_ERR__FAIL = -1,
-  TMQ_RESP_ERR__SUCCESS = 0,
-};
-
-typedef int32_t tmq_resp_err_t;
-
 typedef struct tmq_t      tmq_t;
-typedef struct tmq_topic_vgroup_t      tmq_topic_vgroup_t;
-typedef struct tmq_topic_vgroup_list_t tmq_topic_vgroup_list_t;
-
 typedef struct tmq_conf_t tmq_conf_t;
 typedef struct tmq_list_t tmq_list_t;
 
-typedef void(tmq_commit_cb(tmq_t *, tmq_resp_err_t, tmq_topic_vgroup_list_t *, void *param));
+typedef void(tmq_commit_cb(tmq_t *, int32_t code, void *param));
 
 DLL_EXPORT tmq_list_t *tmq_list_new();
 DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
@@ -233,25 +223,17 @@ DLL_EXPORT char      **tmq_list_to_c_array(const tmq_list_t *);
 
 DLL_EXPORT tmq_t *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
 
-DLL_EXPORT const char *tmq_err2str(tmq_resp_err_t);
+DLL_EXPORT const char *tmq_err2str(int32_t code);
 
 /* ------------------------TMQ CONSUMER INTERFACE------------------------ */
 
-DLL_EXPORT tmq_resp_err_t tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
+DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
-DLL_EXPORT tmq_resp_err_t tmq_unsubscribe(tmq_t *tmq);
+DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
-DLL_EXPORT tmq_resp_err_t tmq_subscription(tmq_t *tmq, tmq_list_t **topics);
+DLL_EXPORT int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics);
-
-// timeout: -1 means infinitely waiting
 DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
-
+DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
-DLL_EXPORT tmq_resp_err_t tmq_consumer_close(tmq_t *tmq);
+DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg);
-DLL_EXPORT tmq_resp_err_t tmq_commit_sync(tmq_t *tmq, const tmq_topic_vgroup_list_t *offsets);
+DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
-DLL_EXPORT void tmq_commit_async(tmq_t *tmq, const tmq_topic_vgroup_list_t *offsets, tmq_commit_cb *cb, void *param);
-
-#if 0
-DLL_EXPORT tmq_resp_err_t tmq_commit(tmq_t *tmq, const tmq_topic_vgroup_list_t *offsets, int32_t async);
-DLL_EXPORT tmq_resp_err_t tmq_seek(tmq_t *tmq, const tmq_topic_vgroup_t *offset);
-#endif
 
 /* ----------------------TMQ CONFIGURATION INTERFACE---------------------- */
 
@@ -275,11 +257,6 @@ DLL_EXPORT const char *tmq_get_db_name(TAOS_RES *res);
 DLL_EXPORT int32_t     tmq_get_vgroup_id(TAOS_RES *res);
 DLL_EXPORT const char *tmq_get_table_name(TAOS_RES *res);
 
-#if 0
-DLL_EXPORT int64_t     tmq_get_request_offset(tmq_message_t *message);
-DLL_EXPORT int64_t     tmq_get_response_offset(tmq_message_t *message);
-#endif
-
 /* ------------------------------ TMQ END -------------------------------- */
 
 #if 1  // Shuduo: temporary enable for app build

include/common/tcommon.h

@@ -35,6 +35,7 @@ enum {
   TMQ_MSG_TYPE__DUMMY = 0,
   TMQ_MSG_TYPE__POLL_RSP,
   TMQ_MSG_TYPE__EP_RSP,
+  TMQ_MSG_TYPE__END_RSP,
 };
 
 typedef enum EStreamType {

include/common/tglobal.h

@@ -61,6 +61,7 @@ extern int32_t tsNumOfRpcThreads;
 extern int32_t tsNumOfCommitThreads;
 extern int32_t tsNumOfTaskQueueThreads;
 extern int32_t tsNumOfMnodeQueryThreads;
+extern int32_t tsNumOfMnodeFetchThreads;
 extern int32_t tsNumOfMnodeReadThreads;
 extern int32_t tsNumOfVnodeQueryThreads;
 extern int32_t tsNumOfVnodeFetchThreads;

include/common/tmsg.h

@@ -106,6 +106,7 @@ typedef enum _mgmt_table {
   TSDB_MGMT_TABLE_CONNS,
   TSDB_MGMT_TABLE_QUERIES,
   TSDB_MGMT_TABLE_VNODES,
+  TSDB_MGMT_TABLE_APPS,
   TSDB_MGMT_TABLE_MAX,
 } EShowType;
 
@@ -428,8 +429,10 @@ STSchema* tdGetSTSChemaFromSSChema(SSchema** pSchema, int32_t nCols);
 typedef struct {
   char    name[TSDB_TABLE_FNAME_LEN];
   int8_t  igExists;
-  float   xFilesFactor;
+  int64_t delay1;
-  int32_t delay;
+  int64_t delay2;
+  int64_t watermark1;
+  int64_t watermark2;
   int32_t ttl;
   int32_t numOfColumns;
   int32_t numOfTags;
@@ -1314,8 +1317,6 @@ int32_t tSerializeSSetStandbyReq(void* buf, int32_t bufLen, SSetStandbyReq* pReq
 int32_t tDeserializeSSetStandbyReq(void* buf, int32_t bufLen, SSetStandbyReq* pReq);
 
 typedef struct {
-  int32_t connId;   // todo remove
-  int32_t queryId;  // todo remove
   char    queryStrId[TSDB_QUERY_ID_LEN];
 } SKillQueryReq;
 
@@ -1323,7 +1324,7 @@ int32_t tSerializeSKillQueryReq(void* buf, int32_t bufLen, SKillQueryReq* pReq);
 int32_t tDeserializeSKillQueryReq(void* buf, int32_t bufLen, SKillQueryReq* pReq);
 
 typedef struct {
-  int32_t connId;
+  uint32_t connId;
 } SKillConnReq;
 
 int32_t tSerializeSKillConnReq(void* buf, int32_t bufLen, SKillConnReq* pReq);
@@ -1519,7 +1520,7 @@ typedef struct {
 #define STREAM_TRIGGER_MAX_DELAY    3
 
 typedef struct {
-  char    name[TSDB_TABLE_FNAME_LEN];
+  char    name[TSDB_STREAM_FNAME_LEN];
   char    sourceDB[TSDB_DB_FNAME_LEN];
   char    targetStbFullName[TSDB_TABLE_FNAME_LEN];
   int8_t  igExists;
@@ -1539,7 +1540,7 @@ int32_t tDeserializeSCMCreateStreamReq(void* buf, int32_t bufLen, SCMCreateStrea
 void    tFreeSCMCreateStreamReq(SCMCreateStreamReq* pReq);
 
 typedef struct {
-  char    name[TSDB_TOPIC_FNAME_LEN];
+  char    name[TSDB_STREAM_FNAME_LEN];
   int64_t streamId;
   char*   sql;
   char*   executorMsg;
@@ -1816,6 +1817,8 @@ typedef struct SVCreateTbReq {
   tb_uid_t uid;
   int64_t  ctime;
   int32_t  ttl;
+  int32_t  commentLen;
+  char*    comment;
   int8_t   type;
   union {
     struct {
@@ -1833,6 +1836,7 @@ int tDecodeSVCreateTbReq(SDecoder* pCoder, SVCreateTbReq* pReq);
 
 static FORCE_INLINE void tdDestroySVCreateTbReq(SVCreateTbReq* req) {
   taosMemoryFreeClear(req->name);
+  taosMemoryFreeClear(req->comment);
   if (req->type == TSDB_CHILD_TABLE) {
     taosMemoryFreeClear(req->ctb.pTag);
   } else if (req->type == TSDB_NORMAL_TABLE) {
@@ -1929,7 +1933,7 @@ typedef struct {
   // TSDB_ALTER_TABLE_UPDATE_OPTIONS
   int8_t  updateTTL;
   int32_t newTTL;
-  int8_t  updateComment;
+  int32_t newCommentLen;
   char*   newComment;
 } SVAlterTbReq;
 
@@ -2010,7 +2014,6 @@ typedef struct {
   int64_t  useconds;
   int64_t  stime;  // timestamp precision ms
   int64_t  reqRid;
-  int32_t  pid;
   bool     stableQuery;
   char     fqdn[TSDB_FQDN_LEN];
   int32_t  subPlanNum;
@@ -2019,8 +2022,6 @@ typedef struct {
 
 typedef struct {
   uint32_t connId;
-  int32_t  pid;
-  char     app[TSDB_APP_NAME_LEN];
   SArray*  queryDesc;  // SArray<SQueryDesc>
 } SQueryHbReqBasic;
 
@@ -2035,9 +2036,31 @@ typedef struct {
   SArray*  pQnodeList;
 } SQueryHbRspBasic;
 
+typedef struct SAppClusterSummary {
+  uint64_t numOfInsertsReq;
+  uint64_t numOfInsertRows;
+  uint64_t insertElapsedTime;
+  uint64_t insertBytes;  // submit to tsdb since launched.
+
+  uint64_t fetchBytes;
+  uint64_t queryElapsedTime;
+  uint64_t numOfSlowQueries;
+  uint64_t totalRequests;
+  uint64_t currentRequests;  // the number of SRequestObj
+} SAppClusterSummary;
+
+typedef struct {
+  int64_t            appId;
+  int32_t            pid;
+  char               name[TSDB_APP_NAME_LEN];
+  int64_t            startTime;
+  SAppClusterSummary summary;
+} SAppHbReq;
+
 typedef struct {
   SClientHbKey      connKey;
   int64_t           clusterId;
+  SAppHbReq         app;
   SQueryHbReqBasic* query;
   SHashObj*         info;  // hash<Skv.key, Skv>
 } SClientHbReq;
@@ -2258,12 +2281,12 @@ typedef struct {
 
 typedef struct {
   char   name[TSDB_STREAM_FNAME_LEN];
-  int64_t streamId;
+  int8_t igNotExists;
-} SMDropStreamTaskReq;
+} SMDropStreamReq;
 
 typedef struct {
   int8_t reserved;
-} SMDropStreamTaskRsp;
+} SMDropStreamRsp;
 
 typedef struct {
   SMsgHead head;

include/common/tmsgcb.h

@@ -34,7 +34,6 @@ typedef enum {
   WRITE_QUEUE,
   APPLY_QUEUE,
   SYNC_QUEUE,
-  MERGE_QUEUE,
   QUEUE_MAX,
 } EQueueType;
 

include/common/tmsgdef.h

@@ -198,7 +198,7 @@ enum {
   TD_DEF_MSG_TYPE(TDMT_VND_ALTER_CONFIRM, "alter-confirm", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_VND_ALTER_HASHRANGE, "alter-hashrange", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_VND_COMPACT, "compact", NULL, NULL)
-
+  TD_DEF_MSG_TYPE(TDMT_VND_DROP_TTL_TABLE, "drop-ttl-stb", NULL, NULL)
   TD_NEW_MSG_SEG(TDMT_QND_MSG)
 
   //shared by snode and vnode

include/common/ttokendef.h

@@ -127,92 +127,92 @@
 #define TK_BLOB          109
 #define TK_VARBINARY     110
 #define TK_DECIMAL       111
-#define TK_FILE_FACTOR   112
+#define TK_MAX_DELAY     112
-#define TK_NK_FLOAT      113
+#define TK_WATERMARK     113
 #define TK_ROLLUP        114
 #define TK_TTL           115
 #define TK_SMA           116
-#define TK_SHOW          117
+#define TK_FIRST         117
-#define TK_DATABASES     118
+#define TK_LAST          118
-#define TK_TABLES        119
+#define TK_SHOW          119
-#define TK_STABLES       120
+#define TK_DATABASES     120
-#define TK_MNODES        121
+#define TK_TABLES        121
-#define TK_MODULES       122
+#define TK_STABLES       122
-#define TK_QNODES        123
+#define TK_MNODES        123
-#define TK_FUNCTIONS     124
+#define TK_MODULES       124
-#define TK_INDEXES       125
+#define TK_QNODES        125
-#define TK_ACCOUNTS      126
+#define TK_FUNCTIONS     126
-#define TK_APPS          127
+#define TK_INDEXES       127
-#define TK_CONNECTIONS   128
+#define TK_ACCOUNTS      128
-#define TK_LICENCE       129
+#define TK_APPS          129
-#define TK_GRANTS        130
+#define TK_CONNECTIONS   130
-#define TK_QUERIES       131
+#define TK_LICENCE       131
-#define TK_SCORES        132
+#define TK_GRANTS        132
-#define TK_TOPICS        133
+#define TK_QUERIES       133
-#define TK_VARIABLES     134
+#define TK_SCORES        134
-#define TK_BNODES        135
+#define TK_TOPICS        135
-#define TK_SNODES        136
+#define TK_VARIABLES     136
-#define TK_CLUSTER       137
+#define TK_BNODES        137
-#define TK_TRANSACTIONS  138
+#define TK_SNODES        138
-#define TK_LIKE          139
+#define TK_CLUSTER       139
-#define TK_INDEX         140
+#define TK_TRANSACTIONS  140
-#define TK_FULLTEXT      141
+#define TK_DISTRIBUTED   141
-#define TK_FUNCTION      142
+#define TK_LIKE          142
-#define TK_INTERVAL      143
+#define TK_INDEX         143
-#define TK_TOPIC         144
+#define TK_FULLTEXT      144
-#define TK_AS            145
+#define TK_FUNCTION      145
-#define TK_CONSUMER      146
+#define TK_INTERVAL      146
-#define TK_GROUP         147
+#define TK_TOPIC         147
-#define TK_DESC          148
+#define TK_AS            148
-#define TK_DESCRIBE      149
+#define TK_CONSUMER      149
-#define TK_RESET         150
+#define TK_GROUP         150
-#define TK_QUERY         151
+#define TK_DESC          151
-#define TK_CACHE         152
+#define TK_DESCRIBE      152
-#define TK_EXPLAIN       153
+#define TK_RESET         153
-#define TK_ANALYZE       154
+#define TK_QUERY         154
-#define TK_VERBOSE       155
+#define TK_CACHE         155
-#define TK_NK_BOOL       156
+#define TK_EXPLAIN       156
-#define TK_RATIO         157
+#define TK_ANALYZE       157
-#define TK_COMPACT       158
+#define TK_VERBOSE       158
-#define TK_VNODES        159
+#define TK_NK_BOOL       159
-#define TK_IN            160
+#define TK_RATIO         160
-#define TK_OUTPUTTYPE    161
+#define TK_NK_FLOAT      161
-#define TK_AGGREGATE     162
+#define TK_COMPACT       162
-#define TK_BUFSIZE       163
+#define TK_VNODES        163
-#define TK_STREAM        164
+#define TK_IN            164
-#define TK_INTO          165
+#define TK_OUTPUTTYPE    165
-#define TK_TRIGGER       166
+#define TK_AGGREGATE     166
-#define TK_AT_ONCE       167
+#define TK_BUFSIZE       167
-#define TK_WINDOW_CLOSE  168
+#define TK_STREAM        168
-#define TK_MAX_DELAY     169
+#define TK_INTO          169
-#define TK_WATERMARK     170
+#define TK_TRIGGER       170
-#define TK_KILL          171
+#define TK_AT_ONCE       171
-#define TK_CONNECTION    172
+#define TK_WINDOW_CLOSE  172
-#define TK_TRANSACTION   173
+#define TK_KILL          173
-#define TK_BALANCE       174
+#define TK_CONNECTION    174
-#define TK_VGROUP        175
+#define TK_TRANSACTION   175
-#define TK_MERGE         176
+#define TK_BALANCE       176
-#define TK_REDISTRIBUTE  177
+#define TK_VGROUP        177
-#define TK_SPLIT         178
+#define TK_MERGE         178
-#define TK_SYNCDB        179
+#define TK_REDISTRIBUTE  179
-#define TK_DELETE        180
+#define TK_SPLIT         180
-#define TK_NULL          181
+#define TK_SYNCDB        181
-#define TK_NK_QUESTION   182
+#define TK_DELETE        182
-#define TK_NK_ARROW      183
+#define TK_NULL          183
-#define TK_ROWTS         184
+#define TK_NK_QUESTION   184
-#define TK_TBNAME        185
+#define TK_NK_ARROW      185
-#define TK_QSTARTTS      186
+#define TK_ROWTS         186
-#define TK_QENDTS        187
+#define TK_TBNAME        187
-#define TK_WSTARTTS      188
+#define TK_QSTARTTS      188
-#define TK_WENDTS        189
+#define TK_QENDTS        189
-#define TK_WDURATION     190
+#define TK_WSTARTTS      190
-#define TK_CAST          191
+#define TK_WENDTS        191
-#define TK_NOW           192
+#define TK_WDURATION     192
-#define TK_TODAY         193
+#define TK_CAST          193
-#define TK_TIMEZONE      194
+#define TK_NOW           194
-#define TK_COUNT         195
+#define TK_TODAY         195
-#define TK_FIRST         196
+#define TK_TIMEZONE      196
-#define TK_LAST          197
+#define TK_COUNT         197
 #define TK_LAST_ROW      198
 #define TK_BETWEEN       199
 #define TK_IS            200
@@ -241,20 +241,22 @@
 #define TK_LINEAR        223
 #define TK_NEXT          224
 #define TK_HAVING        225
-#define TK_ORDER         226
+#define TK_RANGE         226
-#define TK_SLIMIT        227
+#define TK_EVERY         227
-#define TK_SOFFSET       228
+#define TK_ORDER         228
-#define TK_LIMIT         229
+#define TK_SLIMIT        229
-#define TK_OFFSET        230
+#define TK_SOFFSET       230
-#define TK_ASC           231
+#define TK_LIMIT         231
-#define TK_NULLS         232
+#define TK_OFFSET        232
-#define TK_ID            233
+#define TK_ASC           233
-#define TK_NK_BITNOT     234
+#define TK_NULLS         234
-#define TK_INSERT        235
+#define TK_ID            235
-#define TK_VALUES        236
+#define TK_NK_BITNOT     236
-#define TK_IMPORT        237
+#define TK_INSERT        237
-#define TK_NK_SEMI       238
+#define TK_VALUES        238
-#define TK_FILE          239
+#define TK_IMPORT        239
+#define TK_NK_SEMI       240
+#define TK_FILE          241
 
 #define TK_NK_SPACE   300
 #define TK_NK_COMMENT 301

include/dnode/mnode/mnode.h

@@ -96,6 +96,7 @@ int32_t mndGetLoad(SMnode *pMnode, SMnodeLoad *pLoad);
 int32_t mndProcessRpcMsg(SRpcMsg *pMsg);
 int32_t mndProcessSyncMsg(SRpcMsg *pMsg);
 int32_t mndPreProcessMsg(SRpcMsg *pMsg);
+void mndAbortPreprocessMsg(SRpcMsg *pMsg);
 
 /**
  * @brief Generate machine code

include/dnode/snode/snode.h

@@ -16,8 +16,8 @@
 #ifndef _TD_SNODE_H_
 #define _TD_SNODE_H_
 
-#include "tmsgcb.h"
 #include "tmsg.h"
+#include "tmsgcb.h"
 #include "trpc.h"
 
 #ifdef __cplusplus
@@ -68,8 +68,8 @@ int32_t sndGetLoad(SSnode *pSnode, SSnodeLoad *pLoad);
  * @param pMsg The request message
  * @param pRsp The response message
  */
-void sndProcessUMsg(SSnode *pSnode, SRpcMsg *pMsg);
+int32_t sndProcessUMsg(SSnode *pSnode, SRpcMsg *pMsg);
-void sndProcessSMsg(SSnode *pSnode, SRpcMsg *pMsg);
+int32_t sndProcessSMsg(SSnode *pSnode, SRpcMsg *pMsg);
 
 #ifdef __cplusplus
 }

include/libs/catalog/catalog.h

@@ -87,7 +87,7 @@ typedef struct SMetaData {
   SArray* pUdfList;     // pRes = SFuncInfo*
   SArray* pIndex;       // pRes = SIndexInfo*
   SArray* pUser;        // pRes = bool*
-  SArray* pQnodeList;   // pRes = SQueryNodeAddr*
+  SArray* pQnodeList;   // pRes = SArray<SQueryNodeLoad>*
 } SMetaData;
 
 typedef struct SCatalogCfg {
@@ -292,6 +292,8 @@ int32_t catalogUpdateVgEpSet(SCatalog* pCtg, const char* dbFName, int32_t vgId,
 
 int32_t ctgdLaunchAsyncCall(SCatalog* pCtg, SRequestConnInfo* pConn, uint64_t reqId, bool forceUpdate);
 
+int32_t catalogClearCache(void);
+
 /**
  * Destroy catalog and relase all resources
  */

include/libs/executor/executor.h

@@ -95,7 +95,7 @@ int32_t qUpdateQualifiedTableId(qTaskInfo_t tinfo, const SArray* tableIdList, bo
  * @return
  */
 int32_t qCreateExecTask(SReadHandle* readHandle, int32_t vgId, uint64_t taskId, struct SSubplan* pPlan,
-                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, EOPTR_EXEC_MODEL model);
+                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, const char* sql, EOPTR_EXEC_MODEL model);
 
 /**
  *
@@ -159,18 +159,14 @@ int64_t qGetQueriedTableUid(qTaskInfo_t tinfo);
  */
 int32_t qGetQualifiedTableIdList(void* pTableList, const char* tagCond, int32_t tagCondLen, SArray* pTableIdList);
 
-/**
- * Update the table id list of a given query.
- * @param uid   child table uid
- * @param type  operation type: ADD|DROP
- * @return
- */
-int32_t qUpdateQueriedTableIdList(qTaskInfo_t tinfo, int64_t uid, int32_t type);
-
 void qProcessFetchRsp(void* parent, struct SRpcMsg* pMsg, struct SEpSet* pEpSet);
 
 int32_t qGetExplainExecInfo(qTaskInfo_t tinfo, int32_t* resNum, SExplainExecInfo** pRes);
 
+int32_t qSerializeTaskStatus(qTaskInfo_t tinfo, char** pOutput, int32_t* len);
+
+int32_t qDeserializeTaskStatus(qTaskInfo_t tinfo, const char* pInput, int32_t len);
+
 #ifdef __cplusplus
 }
 #endif

include/libs/function/function.h

@@ -150,12 +150,8 @@ typedef struct SqlFunctionCtx {
 } SqlFunctionCtx;
 
 enum {
-  TEXPR_NODE_DUMMY     = 0x0,
   TEXPR_BINARYEXPR_NODE= 0x1,
   TEXPR_UNARYEXPR_NODE = 0x2,
-  TEXPR_FUNCTION_NODE  = 0x3,
-  TEXPR_COL_NODE       = 0x4,
-  TEXPR_VALUE_NODE     = 0x8,
 };
 
 typedef struct tExprNode {

include/libs/function/functionMgt.h

@@ -122,6 +122,7 @@ typedef enum EFunctionType {
   // internal function
   FUNCTION_TYPE_SELECT_VALUE,
   FUNCTION_TYPE_BLOCK_DIST,       // block distribution aggregate function
+  FUNCTION_TYPE_BLOCK_DIST_INFO,  // block distribution pseudo column function
 
   // distributed splitting functions
   FUNCTION_TYPE_APERCENTILE_PARTIAL = 4000,
@@ -185,7 +186,10 @@ bool fmIsUserDefinedFunc(int32_t funcId);
 bool fmIsDistExecFunc(int32_t funcId);
 bool fmIsForbidFillFunc(int32_t funcId);
 bool fmIsForbidStreamFunc(int32_t funcId);
+bool fmIsForbidWindowFunc(int32_t funcId);
+bool fmIsForbidGroupByFunc(int32_t funcId);
 bool fmIsIntervalInterpoFunc(int32_t funcId);
+bool fmIsInterpFunc(int32_t funcId);
 
 int32_t fmGetDistMethod(const SFunctionNode* pFunc, SFunctionNode** pPartialFunc, SFunctionNode** pMergeFunc);
 

include/libs/nodes/cmdnodes.h

@@ -87,9 +87,14 @@ typedef struct SAlterDatabaseStmt {
 
 typedef struct STableOptions {
   ENodeType  type;
+  bool       commentNull;
   char       comment[TSDB_TB_COMMENT_LEN];
-  double     filesFactor;
+  SNodeList* pMaxDelay;
-  int32_t    delay;
+  int64_t    maxDelay1;
+  int64_t    maxDelay2;
+  SNodeList* pWatermark;
+  int64_t    watermark1;
+  int64_t    watermark2;
   SNodeList* pRollupFuncs;
   int32_t    ttl;
   SNodeList* pSma;
@@ -122,6 +127,7 @@ typedef struct SCreateSubTableClause {
   bool       ignoreExists;
   SNodeList* pSpecificTags;
   SNodeList* pValsOfTags;
+  STableOptions* pOptions;
 } SCreateSubTableClause;
 
 typedef struct SCreateMultiTableStmt {
@@ -201,14 +207,28 @@ typedef struct SAlterDnodeStmt {
 typedef struct SShowStmt {
   ENodeType     type;
   SNode*        pDbName;  // SValueNode
-  SNode*    pTbNamePattern;  // SValueNode
+  SNode*        pTbName;  // SValueNode
+  EOperatorType tableCondType;
 } SShowStmt;
 
-typedef struct SShowCreatStmt {
+typedef struct SShowCreateDatabaseStmt {
+  ENodeType type;
+  char      dbName[TSDB_DB_NAME_LEN];
+  void*     pCfg;  // SDbCfgInfo
+} SShowCreateDatabaseStmt;
+
+typedef struct SShowCreateTableStmt {
   ENodeType   type;
   char        dbName[TSDB_DB_NAME_LEN];
   char        tableName[TSDB_TABLE_NAME_LEN];
-} SShowCreatStmt;
+  STableMeta* pMeta;
+} SShowCreateTableStmt;
+
+typedef struct SShowTableDistributedStmt {
+  ENodeType type;
+  char      dbName[TSDB_DB_NAME_LEN];
+  char      tableName[TSDB_TABLE_NAME_LEN];
+} SShowTableDistributedStmt;
 
 typedef enum EIndexType { INDEX_TYPE_SMA = 1, INDEX_TYPE_FULLTEXT } EIndexType;
 

include/libs/nodes/nodes.h

@@ -102,6 +102,7 @@ typedef enum ENodeType {
   QUERY_NODE_EXPLAIN_OPTIONS,
   QUERY_NODE_STREAM_OPTIONS,
   QUERY_NODE_LEFT_VALUE,
+  QUERY_NODE_COLUMN_REF,
 
   // Statement nodes are used in parser and planner module.
   QUERY_NODE_SET_OPERATOR,
@@ -184,6 +185,7 @@ typedef enum ENodeType {
   QUERY_NODE_SHOW_CREATE_TABLE_STMT,
   QUERY_NODE_SHOW_CREATE_STABLE_STMT,
   QUERY_NODE_SHOW_TRANSACTIONS_STMT,
+  QUERY_NODE_SHOW_TABLE_DISTRIBUTED_STMT,
   QUERY_NODE_KILL_CONNECTION_STMT,
   QUERY_NODE_KILL_QUERY_STMT,
   QUERY_NODE_KILL_TRANSACTION_STMT,
@@ -203,6 +205,7 @@ typedef enum ENodeType {
   QUERY_NODE_LOGIC_PLAN_SORT,
   QUERY_NODE_LOGIC_PLAN_PARTITION,
   QUERY_NODE_LOGIC_PLAN_INDEF_ROWS_FUNC,
+  QUERY_NODE_LOGIC_PLAN_INTERP_FUNC,
   QUERY_NODE_LOGIC_SUBPLAN,
   QUERY_NODE_LOGIC_PLAN,
 
@@ -213,6 +216,7 @@ typedef enum ENodeType {
   QUERY_NODE_PHYSICAL_PLAN_TABLE_MERGE_SCAN,
   QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN,
   QUERY_NODE_PHYSICAL_PLAN_SYSTABLE_SCAN,
+  QUERY_NODE_PHYSICAL_PLAN_BLOCK_DIST_SCAN,
   QUERY_NODE_PHYSICAL_PLAN_PROJECT,
   QUERY_NODE_PHYSICAL_PLAN_MERGE_JOIN,
   QUERY_NODE_PHYSICAL_PLAN_HASH_AGG,
@@ -227,11 +231,13 @@ typedef enum ENodeType {
   QUERY_NODE_PHYSICAL_PLAN_FILL,
   QUERY_NODE_PHYSICAL_PLAN_MERGE_SESSION,
   QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION,
+  QUERY_NODE_PHYSICAL_PLAN_STREAM_SEMI_SESSION,
   QUERY_NODE_PHYSICAL_PLAN_STREAM_FINAL_SESSION,
   QUERY_NODE_PHYSICAL_PLAN_MERGE_STATE,
   QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE,
   QUERY_NODE_PHYSICAL_PLAN_PARTITION,
   QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC,
+  QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC,
   QUERY_NODE_PHYSICAL_PLAN_DISPATCH,
   QUERY_NODE_PHYSICAL_PLAN_INSERT,
   QUERY_NODE_PHYSICAL_PLAN_DELETE,

include/libs/nodes/plannodes.h

@@ -39,7 +39,8 @@ typedef enum EScanType {
   SCAN_TYPE_TABLE,
   SCAN_TYPE_SYSTEM_TABLE,
   SCAN_TYPE_STREAM,
-  SCAN_TYPE_TABLE_MERGE
+  SCAN_TYPE_TABLE_MERGE,
+  SCAN_TYPE_BLOCK_INFO
 } EScanType;
 
 typedef struct SScanLogicNode {
@@ -97,9 +98,19 @@ typedef struct SProjectLogicNode {
 
 typedef struct SIndefRowsFuncLogicNode {
   SLogicNode node;
-  SNodeList* pVectorFuncs;
+  SNodeList* pFuncs;
 } SIndefRowsFuncLogicNode;
 
+typedef struct SInterpFuncLogicNode {
+  SLogicNode  node;
+  SNodeList*  pFuncs;
+  STimeWindow timeRange;
+  int64_t     interval;
+  EFillMode   fillMode;
+  SNode*      pFillValues;  // SNodeListNode
+  SNode*      pTimeSeries;  // SColumnNode
+} SInterpFuncLogicNode;
+
 typedef enum EModifyTableType { MODIFY_TABLE_TYPE_INSERT = 1, MODIFY_TABLE_TYPE_DELETE } EModifyTableType;
 
 typedef struct SVnodeModifyLogicNode {
@@ -130,13 +141,17 @@ typedef struct SMergeLogicNode {
 
 typedef enum EWindowType { WINDOW_TYPE_INTERVAL = 1, WINDOW_TYPE_SESSION, WINDOW_TYPE_STATE } EWindowType;
 
-typedef enum EIntervalAlgorithm {
+typedef enum EWindowAlgorithm {
   INTERVAL_ALGO_HASH = 1,
   INTERVAL_ALGO_MERGE,
   INTERVAL_ALGO_STREAM_FINAL,
   INTERVAL_ALGO_STREAM_SEMI,
   INTERVAL_ALGO_STREAM_SINGLE,
-} EIntervalAlgorithm;
+  SESSION_ALGO_STREAM_SEMI,
+  SESSION_ALGO_STREAM_FINAL,
+  SESSION_ALGO_STREAM_SINGLE,
+  SESSION_ALGO_MERGE,
+} EWindowAlgorithm;
 
 typedef struct SWindowLogicNode {
   SLogicNode       node;
@@ -149,11 +164,12 @@ typedef struct SWindowLogicNode {
   int8_t           slidingUnit;
   int64_t          sessionGap;
   SNode*           pTspk;
+  SNode*           pTsEnd;
   SNode*           pStateExpr;
   int8_t           triggerType;
   int64_t          watermark;
   double           filesFactor;
-  EIntervalAlgorithm intervalAlgo;
+  EWindowAlgorithm windowAlgo;
 } SWindowLogicNode;
 
 typedef struct SFillLogicNode {
@@ -243,6 +259,7 @@ typedef struct SScanPhysiNode {
 } SScanPhysiNode;
 
 typedef SScanPhysiNode STagScanPhysiNode;
+typedef SScanPhysiNode SBlockDistScanPhysiNode;
 
 typedef struct SSystemTableScanPhysiNode {
   SScanPhysiNode scan;
@@ -286,9 +303,20 @@ typedef struct SProjectPhysiNode {
 typedef struct SIndefRowsFuncPhysiNode {
   SPhysiNode node;
   SNodeList* pExprs;
-  SNodeList* pVectorFuncs;
+  SNodeList* pFuncs;
 } SIndefRowsFuncPhysiNode;
 
+typedef struct SInterpFuncPhysiNode {
+  SPhysiNode  node;
+  SNodeList*  pExprs;
+  SNodeList*  pFuncs;
+  STimeWindow timeRange;
+  int64_t     interval;
+  EFillMode   fillMode;
+  SNode*      pFillValues;  // SNodeListNode
+  SNode*      pTimeSeries;  // SColumnNode
+} SInterpFuncPhysiNode;
+
 typedef struct SJoinPhysiNode {
   SPhysiNode node;
   EJoinType  joinType;
@@ -332,6 +360,7 @@ typedef struct SWinodwPhysiNode {
   SNodeList* pExprs;  // these are expression list of parameter expression of function
   SNodeList* pFuncs;
   SNode*     pTspk;   // timestamp primary key
+  SNode*     pTsEnd;  // window end timestamp
   int8_t     triggerType;
   int64_t    watermark;
   double     filesFactor;
@@ -371,6 +400,8 @@ typedef struct SSessionWinodwPhysiNode {
 } SSessionWinodwPhysiNode;
 
 typedef SSessionWinodwPhysiNode SStreamSessionWinodwPhysiNode;
+typedef SSessionWinodwPhysiNode SStreamSemiSessionWinodwPhysiNode;
+typedef SSessionWinodwPhysiNode SStreamFinalSessionWinodwPhysiNode;
 
 typedef struct SStateWinodwPhysiNode {
   SWinodwPhysiNode window;

include/libs/nodes/querynodes.h

@@ -65,11 +65,16 @@ typedef struct SColumnNode {
   char        tableName[TSDB_TABLE_NAME_LEN];
   char        tableAlias[TSDB_TABLE_NAME_LEN];
   char        colName[TSDB_COL_NAME_LEN];
-  SNode*      pProjectRef;
+  // SNode*      pProjectRef;
   int16_t dataBlockId;
   int16_t slotId;
 } SColumnNode;
 
+typedef struct SColumnRefNode {
+  ENodeType type;
+  char      colName[TSDB_COL_NAME_LEN];
+} SColumnRefNode;
+
 typedef struct STargetNode {
   ENodeType type;
   int16_t   dataBlockId;
@@ -235,6 +240,9 @@ typedef struct SSelectStmt {
   SNode*      pWindow;
   SNodeList*  pGroupByList;  // SGroupingSetNode
   SNode*      pHaving;
+  SNode*      pRange;
+  SNode*      pEvery;
+  SNode*      pFill;
   SNodeList*  pOrderByList;  // SOrderByExprNode
   SLimitNode* pLimit;
   SLimitNode* pSlimit;
@@ -247,6 +255,9 @@ typedef struct SSelectStmt {
   bool        hasIndefiniteRowsFunc;
   bool        hasSelectFunc;
   bool        hasSelectValFunc;
+  bool        hasUniqueFunc;
+  bool        hasTailFunc;
+  bool        hasInterpFunc;
 } SSelectStmt;
 
 typedef enum ESetOperatorType { SET_OP_TYPE_UNION_ALL = 1, SET_OP_TYPE_UNION } ESetOperatorType;
@@ -374,6 +385,9 @@ bool nodesIsComparisonOp(const SOperatorNode* pOp);
 bool nodesIsJsonOp(const SOperatorNode* pOp);
 bool nodesIsRegularOp(const SOperatorNode* pOp);
 
+bool nodesExprHasColumn(SNode* pNode);
+bool nodesExprsHasColumn(SNodeList* pList);
+
 void*   nodesGetValueFromNode(SValueNode* pNode);
 int32_t nodesSetValueNodeValue(SValueNode* pNode, void* value);
 char*   nodesGetStrValueFromNode(SValueNode* pNode);

include/libs/planner/planner.h

@@ -36,7 +36,6 @@ typedef struct SPlanContext {
   int64_t  watermark;
   char*    pMsg;
   int32_t  msgLen;
-  // double   filesFactor;
 } SPlanContext;
 
 // Create the physical plan for the query, according to the AST.

include/libs/qcom/query.h

@@ -33,8 +33,6 @@ typedef enum {
   JOB_TASK_STATUS_PARTIAL_SUCCEED,
   JOB_TASK_STATUS_SUCCEED,
   JOB_TASK_STATUS_FAILED,
-  JOB_TASK_STATUS_CANCELLING,
-  JOB_TASK_STATUS_CANCELLED,
   JOB_TASK_STATUS_DROPPING,
 } EJobTaskType;
 
@@ -49,6 +47,10 @@ typedef enum {
   TARGET_TYPE_OTHER,
 } ETargetType;
 
+#define QUERY_POLICY_VNODE  1
+#define QUERY_POLICY_HYBRID 2
+#define QUERY_POLICY_QNODE  3
+
 typedef struct STableComInfo {
   uint8_t  numOfTags;     // the number of tags in schema
   uint8_t  precision;     // the number of precision

include/libs/qworker/qworker.h

@@ -64,6 +64,8 @@ typedef struct {
 
 int32_t qWorkerInit(int8_t nodeType, int32_t nodeId, SQWorkerCfg *cfg, void **qWorkerMgmt, const SMsgCb *pMsgCb);
 
+int32_t qWorkerAbortPreprocessQueryMsg(void *qWorkerMgmt, SRpcMsg *pMsg);
+
 int32_t qWorkerPreprocessQueryMsg(void *qWorkerMgmt, SRpcMsg *pMsg);
 
 int32_t qWorkerProcessQueryMsg(void *node, void *qWorkerMgmt, SRpcMsg *pMsg, int64_t ts);

include/libs/scheduler/scheduler.h

@@ -73,6 +73,7 @@ typedef void (*schedulerExecCallback)(SQueryResult* pResult, void* param, int32_
 typedef void (*schedulerFetchCallback)(void* pResult, void* param, int32_t code);
 
 typedef struct SSchedulerReq {
+  bool                 *reqKilled;
   SRequestConnInfo     *pConn;
   SArray               *pNodeList;
   SQueryPlan           *pDag;
@@ -127,7 +128,7 @@ void schedulerStopQueryHb(void *pTrans);
  * Free the query job
  * @param pJob
  */
-void schedulerFreeJob(int64_t job);
+void schedulerFreeJob(int64_t job, int32_t errCode);
 
 void schedulerDestroy(void);
 

include/libs/stream/tstream.h

@@ -152,7 +152,7 @@ void*   streamDataBlockDecode(const void* buf, SStreamDataBlock* pInput);
 typedef struct {
   char* qmsg;
   // followings are not applicable to encoder and decoder
-  void* inputHandle;
+  // void* inputHandle;
   void* executor;
 } STaskExec;
 
@@ -240,12 +240,13 @@ struct SStreamTask {
   int8_t  inputType;
   int8_t  status;
 
-  int8_t  sourceType;
   int8_t  execType;
   int8_t  sinkType;
   int8_t  dispatchType;
   int16_t dispatchMsgType;
 
+  int8_t dataScan;
+
   // node info
   int32_t childId;
   int32_t nodeId;

include/libs/sync/sync.h

@@ -62,6 +62,7 @@ typedef struct SSyncCfg {
 
 typedef struct SFsmCbMeta {
   SyncIndex  index;
+  SyncIndex  lastConfigIndex;
   bool       isWeak;
   int32_t    code;
   ESyncState state;
@@ -75,6 +76,7 @@ typedef struct SReConfigCbMeta {
   int32_t   code;
   SyncIndex index;
   SyncTerm  term;
+  SyncIndex lastConfigIndex;
   SyncTerm  currentTerm;
   SSyncCfg  oldCfg;
   SSyncCfg  newCfg;
@@ -105,7 +107,9 @@ typedef struct SSyncFSM {
   void (*FpReConfigCb)(struct SSyncFSM* pFsm, const SRpcMsg* pMsg, SReConfigCbMeta cbMeta);
   void (*FpLeaderTransferCb)(struct SSyncFSM* pFsm, const SRpcMsg* pMsg, SFsmCbMeta cbMeta);
 
-  int32_t (*FpGetSnapshot)(struct SSyncFSM* pFsm, SSnapshot* pSnapshot);
+
+  int32_t (*FpGetSnapshot)(struct SSyncFSM* pFsm, SSnapshot* pSnapshot, void *pReaderParam, void** ppReader);
+  int32_t (*FpGetSnapshotInfo)(struct SSyncFSM* pFsm, SSnapshot* pSnapshot);
 
   int32_t (*FpSnapshotStartRead)(struct SSyncFSM* pFsm, void** ppReader);
   int32_t (*FpSnapshotStopRead)(struct SSyncFSM* pFsm, void* pReader);
@@ -182,6 +186,7 @@ void        syncStart(int64_t rid);
 void        syncStop(int64_t rid);
 int32_t     syncSetStandby(int64_t rid);
 ESyncState  syncGetMyRole(int64_t rid);
+bool        syncIsReady(int64_t rid);
 const char* syncGetMyRoleStr(int64_t rid);
 SyncTerm    syncGetMyTerm(int64_t rid);
 void        syncGetEpSet(int64_t rid, SEpSet* pEpSet);
@@ -190,7 +195,7 @@ int32_t     syncPropose(int64_t rid, const SRpcMsg* pMsg, bool isWeak);
 bool        syncEnvIsStart();
 const char* syncStr(ESyncState state);
 bool        syncIsRestoreFinish(int64_t rid);
-int32_t     syncGetSnapshotMeta(int64_t rid, struct SSnapshotMeta* sMeta);
+
 
 int32_t syncReconfig(int64_t rid, const SSyncCfg* pNewCfg);
 

include/libs/transport/trpc.h

@@ -23,6 +23,7 @@ extern "C" {
 #include <stdint.h>
 #include "taosdef.h"
 #include "tmsg.h"
+#include "ttrace.h"
 
 #define TAOS_CONN_SERVER 0
 #define TAOS_CONN_CLIENT 1
@@ -45,8 +46,9 @@ typedef struct SRpcHandleInfo {
   int64_t  refId;          // refid, used by server
   int32_t  noResp;         // has response or not(default 0, 0: resp, 1: no resp);
   int32_t  persistHandle;  // persist handle or not
+  STraceId traceId;
+  // int64_t traceId;
 
-  SRpcConnInfo connInfo;
   // app info
   void *ahandle;  // app handle set by client
   void *wrapper;  // wrapper handle
@@ -55,6 +57,9 @@ typedef struct SRpcHandleInfo {
   // resp info
   void *  rsp;
   int32_t rspLen;
+
+  // conn info
+  SRpcConnInfo conn;
 } SRpcHandleInfo;
 
 typedef struct SRpcMsg {
@@ -63,7 +68,6 @@ typedef struct SRpcMsg {
   int32_t        contLen;
   int32_t        code;
   SRpcHandleInfo info;
-  SRpcConnInfo   conn;
 } SRpcMsg;
 
 typedef void (*RpcCfp)(void *parent, SRpcMsg *, SEpSet *rf);

include/os/osSysinfo.h

@@ -68,6 +68,7 @@ typedef struct {
 } SysNameInfo;
 
 SysNameInfo taosGetSysNameInfo();
+bool taosCheckCurrentInDll();
 
 #ifdef __cplusplus
 }

include/os/osTime.h

@@ -23,22 +23,22 @@ extern "C" {
 // If the error is in a third-party library, place this header file under the third-party library header file.
 // When you want to use this feature, you should find or add the same function in the following section.
 #ifndef ALLOW_FORBID_FUNC
-    #define strptime STRPTIME_FUNC_TAOS_FORBID
+#define strptime     STRPTIME_FUNC_TAOS_FORBID
-    #define gettimeofday GETTIMEOFDAY_FUNC_TAOS_FORBID
+#define gettimeofday GETTIMEOFDAY_FUNC_TAOS_FORBID
-    #define localtime   LOCALTIME_FUNC_TAOS_FORBID
+#define localtime    LOCALTIME_FUNC_TAOS_FORBID
-    #define localtime_s LOCALTIMES_FUNC_TAOS_FORBID
+#define localtime_s  LOCALTIMES_FUNC_TAOS_FORBID
-    #define localtime_r LOCALTIMER_FUNC_TAOS_FORBID
+#define localtime_r  LOCALTIMER_FUNC_TAOS_FORBID
-    #define time TIME_FUNC_TAOS_FORBID
+#define time         TIME_FUNC_TAOS_FORBID
-    #define mktime MKTIME_FUNC_TAOS_FORBID
+#define mktime       MKTIME_FUNC_TAOS_FORBID
 #endif
 
 #ifdef WINDOWS
 
-  #define CLOCK_REALTIME 	0
+#define CLOCK_REALTIME 0
 
-  #define MILLISECOND_PER_SECOND (1000i64)
+#define MILLISECOND_PER_SECOND (1000i64)
 #else
-  #define MILLISECOND_PER_SECOND ((int64_t)1000L)
+#define MILLISECOND_PER_SECOND ((int64_t)1000L)
 #endif
 
 #define MILLISECOND_PER_MINUTE (MILLISECOND_PER_SECOND * 60)
@@ -82,7 +82,7 @@ static FORCE_INLINE int64_t taosGetTimestampNs() {
   return (int64_t)systemTime.tv_sec * 1000000000L + (int64_t)systemTime.tv_nsec;
 }
 
-char *taosStrpTime(const char *buf, const char *fmt, struct tm *tm);
+char *     taosStrpTime(const char *buf, const char *fmt, struct tm *tm);
 struct tm *taosLocalTime(const time_t *timep, struct tm *result);
 time_t     taosTime(time_t *t);
 time_t     taosMktime(struct tm *timep);

include/util/taoserror.h

@@ -127,6 +127,8 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_TSC_STMT_API_ERROR            TAOS_DEF_ERROR_CODE(0, 0X0225)
 #define TSDB_CODE_TSC_STMT_TBNAME_ERROR         TAOS_DEF_ERROR_CODE(0, 0X0226)
 #define TSDB_CODE_TSC_STMT_CLAUSE_ERROR         TAOS_DEF_ERROR_CODE(0, 0X0227)
+#define TSDB_CODE_TSC_QUERY_KILLED              TAOS_DEF_ERROR_CODE(0, 0X0228)
+#define TSDB_CODE_TSC_NO_EXEC_NODE              TAOS_DEF_ERROR_CODE(0, 0X0229)
 
 // mnode-common
 #define TSDB_CODE_MND_APP_ERROR                 TAOS_DEF_ERROR_CODE(0, 0x0300)
@@ -410,10 +412,10 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_SYN_INVALID_CHECKSUM          TAOS_DEF_ERROR_CODE(0, 0x0908)
 #define TSDB_CODE_SYN_INVALID_MSGLEN            TAOS_DEF_ERROR_CODE(0, 0x0909)
 #define TSDB_CODE_SYN_INVALID_MSGTYPE           TAOS_DEF_ERROR_CODE(0, 0x090A)
-
+#define TSDB_CODE_SYN_IS_LEADER                 TAOS_DEF_ERROR_CODE(0, 0x090B)
-#define TSDB_CODE_SYN_NOT_LEADER                TAOS_DEF_ERROR_CODE(0, 0x0910)
+#define TSDB_CODE_SYN_NOT_LEADER                TAOS_DEF_ERROR_CODE(0, 0x090C)
-#define TSDB_CODE_SYN_ONE_REPLICA               TAOS_DEF_ERROR_CODE(0, 0x0911)
+#define TSDB_CODE_SYN_ONE_REPLICA               TAOS_DEF_ERROR_CODE(0, 0x090D)
-#define TSDB_CODE_SYN_NOT_IN_NEW_CONFIG         TAOS_DEF_ERROR_CODE(0, 0x0912)
+#define TSDB_CODE_SYN_NOT_IN_NEW_CONFIG         TAOS_DEF_ERROR_CODE(0, 0x090E)
 #define TSDB_CODE_SYN_INTERNAL_ERROR            TAOS_DEF_ERROR_CODE(0, 0x09FF)
 
 // tq
@@ -429,7 +431,8 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_TQ_META_KEY_NOT_IN_TXN        TAOS_DEF_ERROR_CODE(0, 0x0A09)
 #define TSDB_CODE_TQ_META_KEY_DUP_IN_TXN        TAOS_DEF_ERROR_CODE(0, 0x0A0A)
 #define TSDB_CODE_TQ_GROUP_NOT_SET              TAOS_DEF_ERROR_CODE(0, 0x0A0B)
-#define TSDB_CODE_TQ_TABLE_SCHEMA_NOT_FOUND     TAOS_DEF_ERROR_CODE(0, 0x0A0B)
+#define TSDB_CODE_TQ_TABLE_SCHEMA_NOT_FOUND     TAOS_DEF_ERROR_CODE(0, 0x0A0C)
+#define TSDB_CODE_TQ_NO_COMMITTED_OFFSET        TAOS_DEF_ERROR_CODE(0, 0x0A0D)
 
 // wal
 #define TSDB_CODE_WAL_APP_ERROR                 TAOS_DEF_ERROR_CODE(0, 0x1000)
@@ -439,108 +442,6 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_WAL_OUT_OF_MEMORY             TAOS_DEF_ERROR_CODE(0, 0x1004)
 #define TSDB_CODE_WAL_LOG_NOT_EXIST             TAOS_DEF_ERROR_CODE(0, 0x1005)
 
-// http
-#define TSDB_CODE_HTTP_SERVER_OFFLINE           TAOS_DEF_ERROR_CODE(0, 0x1100)  //"http server is not online"
-#define TSDB_CODE_HTTP_UNSUPPORT_URL            TAOS_DEF_ERROR_CODE(0, 0x1101)  //"url is not support"
-#define TSDB_CODE_HTTP_INVALID_URL              TAOS_DEF_ERROR_CODE(0, 0x1102)  //invalid url format"
-#define TSDB_CODE_HTTP_NO_ENOUGH_MEMORY         TAOS_DEF_ERROR_CODE(0, 0x1103)  //"no enough memory"
-#define TSDB_CODE_HTTP_REQUSET_TOO_BIG          TAOS_DEF_ERROR_CODE(0, 0x1104)  //"request size is too big"
-#define TSDB_CODE_HTTP_NO_AUTH_INFO             TAOS_DEF_ERROR_CODE(0, 0x1105)  //"no auth info input"
-#define TSDB_CODE_HTTP_NO_MSG_INPUT             TAOS_DEF_ERROR_CODE(0, 0x1106)  //"request is empty"
-#define TSDB_CODE_HTTP_NO_SQL_INPUT             TAOS_DEF_ERROR_CODE(0, 0x1107)  //"no sql input"
-#define TSDB_CODE_HTTP_NO_EXEC_USEDB            TAOS_DEF_ERROR_CODE(0, 0x1108)  //"no need to execute use db cmd"
-#define TSDB_CODE_HTTP_SESSION_FULL             TAOS_DEF_ERROR_CODE(0, 0x1109)  //"session list was full"
-#define TSDB_CODE_HTTP_GEN_TAOSD_TOKEN_ERR      TAOS_DEF_ERROR_CODE(0, 0x110A)  //"generate taosd token error"
-#define TSDB_CODE_HTTP_INVALID_MULTI_REQUEST    TAOS_DEF_ERROR_CODE(0, 0x110B)  //"size of multi request is 0"
-#define TSDB_CODE_HTTP_CREATE_GZIP_FAILED       TAOS_DEF_ERROR_CODE(0, 0x110C)  //"failed to create gzip"
-#define TSDB_CODE_HTTP_FINISH_GZIP_FAILED       TAOS_DEF_ERROR_CODE(0, 0x110D)  //"failed to finish gzip"
-#define TSDB_CODE_HTTP_LOGIN_FAILED             TAOS_DEF_ERROR_CODE(0, 0x110E)  //"failed to login"
-
-#define TSDB_CODE_HTTP_INVALID_VERSION          TAOS_DEF_ERROR_CODE(0, 0x1120)  //"invalid http version"
-#define TSDB_CODE_HTTP_INVALID_CONTENT_LENGTH   TAOS_DEF_ERROR_CODE(0, 0x1121)  //"invalid content length"
-#define TSDB_CODE_HTTP_INVALID_AUTH_TYPE        TAOS_DEF_ERROR_CODE(0, 0x1122)  //"invalid type of Authorization"
-#define TSDB_CODE_HTTP_INVALID_AUTH_FORMAT      TAOS_DEF_ERROR_CODE(0, 0x1123)  //"invalid format of Authorization"
-#define TSDB_CODE_HTTP_INVALID_BASIC_AUTH       TAOS_DEF_ERROR_CODE(0, 0x1124)  //"invalid basic Authorization"
-#define TSDB_CODE_HTTP_INVALID_TAOSD_AUTH       TAOS_DEF_ERROR_CODE(0, 0x1125)  //"invalid taosd Authorization"
-#define TSDB_CODE_HTTP_PARSE_METHOD_FAILED      TAOS_DEF_ERROR_CODE(0, 0x1126)  //"failed to parse method"
-#define TSDB_CODE_HTTP_PARSE_TARGET_FAILED      TAOS_DEF_ERROR_CODE(0, 0x1127)  //"failed to parse target"
-#define TSDB_CODE_HTTP_PARSE_VERSION_FAILED     TAOS_DEF_ERROR_CODE(0, 0x1128)  //"failed to parse http version"
-#define TSDB_CODE_HTTP_PARSE_SP_FAILED          TAOS_DEF_ERROR_CODE(0, 0x1129)  //"failed to parse sp"
-#define TSDB_CODE_HTTP_PARSE_STATUS_FAILED      TAOS_DEF_ERROR_CODE(0, 0x112A)  //"failed to parse status"
-#define TSDB_CODE_HTTP_PARSE_PHRASE_FAILED      TAOS_DEF_ERROR_CODE(0, 0x112B)  //"failed to parse phrase"
-#define TSDB_CODE_HTTP_PARSE_CRLF_FAILED        TAOS_DEF_ERROR_CODE(0, 0x112C)  //"failed to parse crlf"
-#define TSDB_CODE_HTTP_PARSE_HEADER_FAILED      TAOS_DEF_ERROR_CODE(0, 0x112D)  //"failed to parse header"
-#define TSDB_CODE_HTTP_PARSE_HEADER_KEY_FAILED  TAOS_DEF_ERROR_CODE(0, 0x112E)  //"failed to parse header key"
-#define TSDB_CODE_HTTP_PARSE_HEADER_VAL_FAILED  TAOS_DEF_ERROR_CODE(0, 0x112F)  //"failed to parse header val"
-#define TSDB_CODE_HTTP_PARSE_CHUNK_SIZE_FAILED  TAOS_DEF_ERROR_CODE(0, 0x1130)  //"failed to parse chunk size"
-#define TSDB_CODE_HTTP_PARSE_CHUNK_FAILED       TAOS_DEF_ERROR_CODE(0, 0x1131)  //"failed to parse chunk"
-#define TSDB_CODE_HTTP_PARSE_END_FAILED         TAOS_DEF_ERROR_CODE(0, 0x1132)  //"failed to parse end section"
-#define TSDB_CODE_HTTP_PARSE_INVALID_STATE      TAOS_DEF_ERROR_CODE(0, 0x1134)  //"invalid parse state"
-#define TSDB_CODE_HTTP_PARSE_ERROR_STATE        TAOS_DEF_ERROR_CODE(0, 0x1135)  //"failed to parse error section"
-
-#define TSDB_CODE_HTTP_GC_QUERY_NULL            TAOS_DEF_ERROR_CODE(0, 0x1150)  //"query size is 0"
-#define TSDB_CODE_HTTP_GC_QUERY_SIZE            TAOS_DEF_ERROR_CODE(0, 0x1151)  //"query size can not more than 100"
-#define TSDB_CODE_HTTP_GC_REQ_PARSE_ERROR       TAOS_DEF_ERROR_CODE(0, 0x1152)  //"parse grafana json error"
-
-#define TSDB_CODE_HTTP_TG_DB_NOT_INPUT          TAOS_DEF_ERROR_CODE(0, 0x1160)  //"database name can not be null"
-#define TSDB_CODE_HTTP_TG_DB_TOO_LONG           TAOS_DEF_ERROR_CODE(0, 0x1161)  //"database name too long"
-#define TSDB_CODE_HTTP_TG_INVALID_JSON          TAOS_DEF_ERROR_CODE(0, 0x1162)  //"invalid telegraf json fromat"
-#define TSDB_CODE_HTTP_TG_METRICS_NULL          TAOS_DEF_ERROR_CODE(0, 0x1163)  //"metrics size is 0"
-#define TSDB_CODE_HTTP_TG_METRICS_SIZE          TAOS_DEF_ERROR_CODE(0, 0x1164)  //"metrics size can not more than 1K"
-#define TSDB_CODE_HTTP_TG_METRIC_NULL           TAOS_DEF_ERROR_CODE(0, 0x1165)  //"metric name not find"
-#define TSDB_CODE_HTTP_TG_METRIC_TYPE           TAOS_DEF_ERROR_CODE(0, 0x1166)  //"metric name type should be string"
-#define TSDB_CODE_HTTP_TG_METRIC_NAME_NULL      TAOS_DEF_ERROR_CODE(0, 0x1167)  //"metric name length is 0"
-#define TSDB_CODE_HTTP_TG_METRIC_NAME_LONG      TAOS_DEF_ERROR_CODE(0, 0x1168)  //"metric name length too long"
-#define TSDB_CODE_HTTP_TG_TIMESTAMP_NULL        TAOS_DEF_ERROR_CODE(0, 0x1169)  //"timestamp not find"
-#define TSDB_CODE_HTTP_TG_TIMESTAMP_TYPE        TAOS_DEF_ERROR_CODE(0, 0x116A)  //"timestamp type should be integer"
-#define TSDB_CODE_HTTP_TG_TIMESTAMP_VAL_NULL    TAOS_DEF_ERROR_CODE(0, 0x116B)  //"timestamp value smaller than 0"
-#define TSDB_CODE_HTTP_TG_TAGS_NULL             TAOS_DEF_ERROR_CODE(0, 0x116C)  //"tags not find"
-#define TSDB_CODE_HTTP_TG_TAGS_SIZE_0           TAOS_DEF_ERROR_CODE(0, 0x116D)  //"tags size is 0"
-#define TSDB_CODE_HTTP_TG_TAGS_SIZE_LONG        TAOS_DEF_ERROR_CODE(0, 0x116E)  //"tags size too long"
-#define TSDB_CODE_HTTP_TG_TAG_NULL              TAOS_DEF_ERROR_CODE(0, 0x116F)  //"tag is null"
-#define TSDB_CODE_HTTP_TG_TAG_NAME_NULL         TAOS_DEF_ERROR_CODE(0, 0x1170)  //"tag name is null"
-#define TSDB_CODE_HTTP_TG_TAG_NAME_SIZE         TAOS_DEF_ERROR_CODE(0, 0x1171)  //"tag name length too long"
-#define TSDB_CODE_HTTP_TG_TAG_VALUE_TYPE        TAOS_DEF_ERROR_CODE(0, 0x1172)  //"tag value type should be number or string"
-#define TSDB_CODE_HTTP_TG_TAG_VALUE_NULL        TAOS_DEF_ERROR_CODE(0, 0x1173)  //"tag value is null"
-#define TSDB_CODE_HTTP_TG_TABLE_NULL            TAOS_DEF_ERROR_CODE(0, 0x1174)  //"table is null"
-#define TSDB_CODE_HTTP_TG_TABLE_SIZE            TAOS_DEF_ERROR_CODE(0, 0x1175)  //"table name length too long"
-#define TSDB_CODE_HTTP_TG_FIELDS_NULL           TAOS_DEF_ERROR_CODE(0, 0x1176)  //"fields not find"
-#define TSDB_CODE_HTTP_TG_FIELDS_SIZE_0         TAOS_DEF_ERROR_CODE(0, 0x1177)  //"fields size is 0"
-#define TSDB_CODE_HTTP_TG_FIELDS_SIZE_LONG      TAOS_DEF_ERROR_CODE(0, 0x1178)  //"fields size too long"
-#define TSDB_CODE_HTTP_TG_FIELD_NULL            TAOS_DEF_ERROR_CODE(0, 0x1179)  //"field is null"
-#define TSDB_CODE_HTTP_TG_FIELD_NAME_NULL       TAOS_DEF_ERROR_CODE(0, 0x117A)  //"field name is null"
-#define TSDB_CODE_HTTP_TG_FIELD_NAME_SIZE       TAOS_DEF_ERROR_CODE(0, 0x117B)  //"field name length too long"
-#define TSDB_CODE_HTTP_TG_FIELD_VALUE_TYPE      TAOS_DEF_ERROR_CODE(0, 0x117C)  //"field value type should be number or string"
-#define TSDB_CODE_HTTP_TG_FIELD_VALUE_NULL      TAOS_DEF_ERROR_CODE(0, 0x117D)  //"field value is null"
-#define TSDB_CODE_HTTP_TG_HOST_NOT_STRING       TAOS_DEF_ERROR_CODE(0, 0x117E)  //"host type should be string"
-#define TSDB_CODE_HTTP_TG_STABLE_NOT_EXIST      TAOS_DEF_ERROR_CODE(0, 0x117F)  //"stable not exist"
-
-#define TSDB_CODE_HTTP_OP_DB_NOT_INPUT          TAOS_DEF_ERROR_CODE(0, 0x1190)  //"database name can not be null"
-#define TSDB_CODE_HTTP_OP_DB_TOO_LONG           TAOS_DEF_ERROR_CODE(0, 0x1191)  //"database name too long"
-#define TSDB_CODE_HTTP_OP_INVALID_JSON          TAOS_DEF_ERROR_CODE(0, 0x1192)  //"invalid opentsdb json fromat"
-#define TSDB_CODE_HTTP_OP_METRICS_NULL          TAOS_DEF_ERROR_CODE(0, 0x1193)  //"metrics size is 0"
-#define TSDB_CODE_HTTP_OP_METRICS_SIZE          TAOS_DEF_ERROR_CODE(0, 0x1194)  //"metrics size can not more than 10K"
-#define TSDB_CODE_HTTP_OP_METRIC_NULL           TAOS_DEF_ERROR_CODE(0, 0x1195)  //"metric name not find"
-#define TSDB_CODE_HTTP_OP_METRIC_TYPE           TAOS_DEF_ERROR_CODE(0, 0x1196)  //"metric name type should be string"
-#define TSDB_CODE_HTTP_OP_METRIC_NAME_NULL      TAOS_DEF_ERROR_CODE(0, 0x1197)  //"metric name length is 0"
-#define TSDB_CODE_HTTP_OP_METRIC_NAME_LONG      TAOS_DEF_ERROR_CODE(0, 0x1198)  //"metric name length can not more than 22"
-#define TSDB_CODE_HTTP_OP_TIMESTAMP_NULL        TAOS_DEF_ERROR_CODE(0, 0x1199)  //"timestamp not find"
-#define TSDB_CODE_HTTP_OP_TIMESTAMP_TYPE        TAOS_DEF_ERROR_CODE(0, 0x119A)  //"timestamp type should be integer"
-#define TSDB_CODE_HTTP_OP_TIMESTAMP_VAL_NULL    TAOS_DEF_ERROR_CODE(0, 0x119B)  //"timestamp value smaller than 0"
-#define TSDB_CODE_HTTP_OP_TAGS_NULL             TAOS_DEF_ERROR_CODE(0, 0x119C)  //"tags not find"
-#define TSDB_CODE_HTTP_OP_TAGS_SIZE_0           TAOS_DEF_ERROR_CODE(0, 0x119D)  //"tags size is 0"
-#define TSDB_CODE_HTTP_OP_TAGS_SIZE_LONG        TAOS_DEF_ERROR_CODE(0, 0x119E)  //"tags size too long"
-#define TSDB_CODE_HTTP_OP_TAG_NULL              TAOS_DEF_ERROR_CODE(0, 0x119F)  //"tag is null"
-#define TSDB_CODE_HTTP_OP_TAG_NAME_NULL         TAOS_DEF_ERROR_CODE(0, 0x11A0)  //"tag name is null"
-#define TSDB_CODE_HTTP_OP_TAG_NAME_SIZE         TAOS_DEF_ERROR_CODE(0, 0x11A1)  //"tag name length too long"
-#define TSDB_CODE_HTTP_OP_TAG_VALUE_TYPE        TAOS_DEF_ERROR_CODE(0, 0x11A2)  //"tag value type should be boolean number or string"
-#define TSDB_CODE_HTTP_OP_TAG_VALUE_NULL        TAOS_DEF_ERROR_CODE(0, 0x11A3)  //"tag value is null"
-#define TSDB_CODE_HTTP_OP_TAG_VALUE_TOO_LONG    TAOS_DEF_ERROR_CODE(0, 0x11A4)  //"tag value can not more than 64"
-#define TSDB_CODE_HTTP_OP_VALUE_NULL            TAOS_DEF_ERROR_CODE(0, 0x11A5)  //"value not find"
-#define TSDB_CODE_HTTP_OP_VALUE_TYPE            TAOS_DEF_ERROR_CODE(0, 0x11A6)  //"value type should be boolean number or string"
-
-#define TSDB_CODE_HTTP_REQUEST_JSON_ERROR       TAOS_DEF_ERROR_CODE(0, 0x1F00)  //"http request json error"
-
 // tfs
 #define TSDB_CODE_FS_APP_ERROR                  TAOS_DEF_ERROR_CODE(0, 0x2200)
 #define TSDB_CODE_FS_INVLD_CFG                  TAOS_DEF_ERROR_CODE(0, 0x2201)
@@ -570,6 +471,7 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_SCH_INTERNAL_ERROR            TAOS_DEF_ERROR_CODE(0, 0x2502)
 #define TSDB_CODE_SCH_IGNORE_ERROR              TAOS_DEF_ERROR_CODE(0, 0x2503)
 #define TSDB_CODE_SCH_TIMEOUT_ERROR             TAOS_DEF_ERROR_CODE(0, 0x2504)
+#define TSDB_CODE_SCH_JOB_IS_DROPPING           TAOS_DEF_ERROR_CODE(0, 0x2505)
 #define TSDB_CODE_QW_MSG_ERROR                  TAOS_DEF_ERROR_CODE(0, 0x2550)
 
 //parser
@@ -658,6 +560,11 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_PAR_INVALID_REDISTRIBUTE_VG   TAOS_DEF_ERROR_CODE(0, 0x2656)
 #define TSDB_CODE_PAR_FILL_NOT_ALLOWED_FUNC     TAOS_DEF_ERROR_CODE(0, 0x2657)
 #define TSDB_CODE_PAR_INVALID_WINDOW_PC         TAOS_DEF_ERROR_CODE(0, 0x2658)
+#define TSDB_CODE_PAR_WINDOW_NOT_ALLOWED_FUNC   TAOS_DEF_ERROR_CODE(0, 0x2659)
+#define TSDB_CODE_PAR_STREAM_NOT_ALLOWED_FUNC   TAOS_DEF_ERROR_CODE(0, 0x265A)
+#define TSDB_CODE_PAR_GROUP_BY_NOT_ALLOWED_FUNC TAOS_DEF_ERROR_CODE(0, 0x265B)
+#define TSDB_CODE_PAR_INVALID_TABLE_OPTION      TAOS_DEF_ERROR_CODE(0, 0x265C)
+#define TSDB_CODE_PAR_INVALID_INTERP_CLAUSE     TAOS_DEF_ERROR_CODE(0, 0x265D)
 
 //planner
 #define TSDB_CODE_PLAN_INTERNAL_ERROR           TAOS_DEF_ERROR_CODE(0, 0x2700)
@@ -706,6 +613,8 @@ int32_t* taosGetErrno();
 //index
 #define TSDB_CODE_INDEX_REBUILDING               TAOS_DEF_ERROR_CODE(0, 0x3200)
 
+//tmq
+#define TSDB_CODE_TMQ_INVALID_MSG                TAOS_DEF_ERROR_CODE(0, 0x4000)
 
 #ifdef __cplusplus
 }

include/util/tdef.h

@@ -210,7 +210,7 @@ typedef enum ELogicConditionType {
 #define TSDB_TYPE_STR_MAX_LEN    32
 #define TSDB_TABLE_FNAME_LEN     (TSDB_DB_FNAME_LEN + TSDB_TABLE_NAME_LEN + TSDB_NAME_DELIMITER_LEN)
 #define TSDB_TOPIC_FNAME_LEN     (TSDB_ACCT_ID_LEN + TSDB_TABLE_NAME_LEN + TSDB_NAME_DELIMITER_LEN)
-#define TSDB_STREAM_FNAME_LEN    TSDB_TABLE_FNAME_LEN
+#define TSDB_STREAM_FNAME_LEN    (TSDB_ACCT_ID_LEN + TSDB_TABLE_NAME_LEN + TSDB_NAME_DELIMITER_LEN)
 #define TSDB_SUBSCRIBE_KEY_LEN   (TSDB_CGROUP_LEN + TSDB_TOPIC_FNAME_LEN + 2)
 #define TSDB_PARTITION_KEY_LEN   (TSDB_SUBSCRIBE_KEY_LEN + 20)
 #define TSDB_COL_NAME_LEN        65
@@ -344,12 +344,11 @@ typedef enum ELogicConditionType {
 #define TSDB_DB_SCHEMALESS_OFF          0
 #define TSDB_DEFAULT_DB_SCHEMALESS      TSDB_DB_SCHEMALESS_OFF
 
-// #define TSDB_MIN_ROLLUP_DELAY     1
+#define TSDB_MIN_ROLLUP_MAX_DELAY     1  // unit millisecond
-// #define TSDB_MAX_ROLLUP_DELAY     10
+#define TSDB_MAX_ROLLUP_MAX_DELAY     (15 * 60 * 1000)
-// #define TSDB_DEFAULT_ROLLUP_DELAY 1
+#define TSDB_MIN_ROLLUP_WATERMARK     0  // unit millisecond
-#define TSDB_MIN_ROLLUP_FILE_FACTOR     0
+#define TSDB_MAX_ROLLUP_WATERMARK     (15 * 60 * 1000)
-#define TSDB_MAX_ROLLUP_FILE_FACTOR     10
+#define TSDB_DEFAULT_ROLLUP_WATERMARK 5000
-#define TSDB_DEFAULT_ROLLUP_FILE_FACTOR 0.1
 #define TSDB_MIN_TABLE_TTL            0
 #define TSDB_DEFAULT_TABLE_TTL        0
 

include/util/ttrace.h

@@ -0,0 +1,57 @@
+/*
+ * Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
+ *
+ * This program is free software: you can use, redistribute, and/or modify
+ * it under the terms of the GNU Affero General Public License, version 3
+ * or later ("AGPL"), as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+#ifndef _TD_TRACE_H_
+#define _TD_TRACE_H_
+
+#include <stdint.h>
+#include <stdlib.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#pragma(push, 1)
+
+typedef struct STraceId {
+  int64_t rootId;
+  int64_t msgId;
+} STraceId;
+
+#pragma(pop)
+
+#define TRACE_SET_ROOTID(traceId, root) \
+  do {                                  \
+    (traceId)->rootId = root;           \
+  } while (0);
+
+#define TRACE_GET_ROOTID(traceId) (traceId)->rootId
+
+#define TRACE_SET_MSGID(traceId, mId) \
+  do {                                \
+    (traceId)->msgId = mId;           \
+  } while (0)
+
+#define TRACE_GET_MSGID(traceId) (traceId)->msgId
+
+#define TRACE_TO_STR(traceId, buf)                                                \
+  do {                                                                            \
+    sprintf(buf, "0x%" PRIx64 ":0x%" PRIx64 "", traceId->rootId, traceId->msgId); \
+  } while (0)
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

source/client/inc/clientInt.h

@@ -22,6 +22,7 @@ extern "C" {
 
 #include "parser.h"
 #include "planner.h"
+#include "catalog.h"
 #include "query.h"
 #include "taos.h"
 #include "tcommon.h"
@@ -76,10 +77,12 @@ typedef int32_t (*FHbReqHandle)(SClientHbKey* connKey, void* param, SClientHbReq
 
 typedef struct {
   int8_t inited;
+  int64_t       appId;
   // ctl
   int8_t        threadStop;
   TdThread      thread;
   TdThreadMutex lock;       // used when app init and cleanup
+  SHashObj     *appSummary;
   SArray*       appHbMgrs;  // SArray<SAppHbMgr*> one for each cluster
   FHbReqHandle  reqHandle[CONN_TYPE__MAX];
   FHbRspHandle  rspHandle[CONN_TYPE__MAX];
@@ -92,19 +95,6 @@ typedef struct SQueryExecMetric {
   int64_t rsp;     // receive response from server, us
 } SQueryExecMetric;
 
-typedef struct SInstanceSummary {
-  uint64_t numOfInsertsReq;
-  uint64_t numOfInsertRows;
-  uint64_t insertElapsedTime;
-  uint64_t insertBytes;  // submit to tsdb since launched.
-
-  uint64_t fetchBytes;
-  uint64_t queryElapsedTime;
-  uint64_t numOfSlowQueries;
-  uint64_t totalRequests;
-  uint64_t currentRequests;  // the number of SRequestObj
-} SInstanceSummary;
-
 typedef struct SHeartBeatInfo {
   void* pTimer;  // timer, used to send request msg to mnode
 } SHeartBeatInfo;
@@ -114,7 +104,7 @@ struct SAppInstInfo {
   SCorEpSet          mgmtEp;
   TdThreadMutex      qnodeMutex;
   SArray*            pQnodeList;
-  SInstanceSummary summary;
+  SAppClusterSummary summary;
   SList*             pConnList;  // STscObj linked list
   uint64_t           clusterId;
   void*              pTransporter;
@@ -139,7 +129,7 @@ typedef struct STscObj {
   int8_t        connType;
   int32_t       acctId;
   uint32_t      connId;
-  uint64_t      id;         // ref ID returned by taosAddRef
+  TAOS         *id;         // ref ID returned by taosAddRef
   TdThreadMutex mutex;      // used to protect the operation on db
   int32_t       numOfReqs;  // number of sqlObj bound to this connection
   SAppInstInfo* pAppInfo;
@@ -173,6 +163,7 @@ typedef struct SReqResultInfo {
   int32_t        precision;
   bool           convertUcs4;
   int32_t        payloadLen;
+  char*          convertJson;
 } SReqResultInfo;
 
 typedef struct SRequestSendRecvBody {
@@ -215,6 +206,7 @@ typedef struct SRequestObj {
   SRequestSendRecvBody body;
   bool                 stableQuery;
 
+  bool                 killed;
   uint32_t             prevCode; //previous error code: todo refactor, add update flag for catalog
   uint32_t             retry;
 } SRequestObj;
@@ -252,6 +244,7 @@ static FORCE_INLINE SReqResultInfo* tmqGetNextResInfo(TAOS_RES* res, bool conver
       taosMemoryFreeClear(msg->resInfo.pCol);
       taosMemoryFreeClear(msg->resInfo.length);
       taosMemoryFreeClear(msg->resInfo.convertBuf);
+      taosMemoryFreeClear(msg->resInfo.convertJson);
     }
     setQueryResultFromRsp(&msg->resInfo, pRetrieve, convertUcs4, false);
     return &msg->resInfo;
@@ -302,7 +295,7 @@ SRequestObj* launchQuery(STscObj* pTscObj, const char* sql, int sqlLen);
 
 int32_t parseSql(SRequestObj* pRequest, bool topicQuery, SQuery** pQuery, SStmtCallback* pStmtCb);
 
-int32_t getPlan(SRequestObj* pRequest, SQuery* pQuery, SQueryPlan** pPlan, SArray** pNodeList);
+int32_t getPlan(SRequestObj* pRequest, SQuery* pQuery, SQueryPlan** pPlan, SArray* pNodeList);
 
 int32_t buildRequest(STscObj* pTscObj, const char* sql, int sqlLen, SRequestObj** pRequest);
 
@@ -327,12 +320,13 @@ void hbMgrInitMqHbRspHandle();
 
 SRequestObj* launchQueryImpl(SRequestObj* pRequest, SQuery* pQuery, bool keepQuery, void** res);
 int32_t      scheduleQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNodeList);
-void         launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery);
+void         launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery, SMetaData *pResultMeta);
 int32_t      refreshMeta(STscObj* pTscObj, SRequestObj* pRequest);
 int32_t      updateQnodeList(SAppInstInfo* pInfo, SArray* pNodeList);
 void         doAsyncQuery(SRequestObj* pRequest, bool forceUpdateMeta);
 int32_t      removeMeta(STscObj* pTscObj, SArray* tbList);// todo move to clientImpl.c and become a static function
 int32_t      handleAlterTbExecRes(void* res, struct SCatalog* pCatalog);// todo move to xxx
+bool         qnodeRequired(SRequestObj* pRequest);
 
 #ifdef __cplusplus
 }

source/client/src/clientEnv.c

@@ -38,7 +38,7 @@ static TdThreadOnce tscinit = PTHREAD_ONCE_INIT;
 volatile int32_t    tscInitRes = 0;
 
 static void registerRequest(SRequestObj *pRequest) {
-  STscObj *pTscObj = acquireTscObj(pRequest->pTscObj->id);
+  STscObj *pTscObj = acquireTscObj(*(int64_t*)pRequest->pTscObj->id);
 
   assert(pTscObj != NULL);
 
@@ -48,13 +48,13 @@ static void registerRequest(SRequestObj *pRequest) {
   int32_t num = atomic_add_fetch_32(&pTscObj->numOfReqs, 1);
 
   if (pTscObj->pAppInfo) {
-    SInstanceSummary *pSummary = &pTscObj->pAppInfo->summary;
+    SAppClusterSummary *pSummary = &pTscObj->pAppInfo->summary;
 
     int32_t total = atomic_add_fetch_64((int64_t *)&pSummary->totalRequests, 1);
     int32_t currentInst = atomic_add_fetch_64((int64_t *)&pSummary->currentRequests, 1);
     tscDebug("0x%" PRIx64 " new Request from connObj:0x%" PRIx64
              ", current:%d, app current:%d, total:%d, reqId:0x%" PRIx64,
-             pRequest->self, pRequest->pTscObj->id, num, currentInst, total, pRequest->requestId);
+             pRequest->self, *(int64_t*)pRequest->pTscObj->id, num, currentInst, total, pRequest->requestId);
   }
 }
 
@@ -62,7 +62,7 @@ static void deregisterRequest(SRequestObj *pRequest) {
   assert(pRequest != NULL);
 
   STscObj          *pTscObj = pRequest->pTscObj;
-  SInstanceSummary *pActivity = &pTscObj->pAppInfo->summary;
+  SAppClusterSummary *pActivity = &pTscObj->pAppInfo->summary;
 
   int32_t currentInst = atomic_sub_fetch_64((int64_t *)&pActivity->currentRequests, 1);
   int32_t num = atomic_sub_fetch_32(&pTscObj->numOfReqs, 1);
@@ -70,8 +70,8 @@ static void deregisterRequest(SRequestObj *pRequest) {
   int64_t duration = taosGetTimestampUs() - pRequest->metric.start;
   tscDebug("0x%" PRIx64 " free Request from connObj: 0x%" PRIx64 ", reqId:0x%" PRIx64 " elapsed:%" PRIu64
            " ms, current:%d, app current:%d",
-           pRequest->self, pTscObj->id, pRequest->requestId, duration / 1000, num, currentInst);
+           pRequest->self, *(int64_t*)pTscObj->id, pRequest->requestId, duration / 1000, num, currentInst);
-  releaseTscObj(pTscObj->id);
+  releaseTscObj(*(int64_t*)pTscObj->id);
 }
 
 // todo close the transporter properly
@@ -80,7 +80,7 @@ void closeTransporter(STscObj *pTscObj) {
     return;
   }
 
-  tscDebug("free transporter:%p in connObj: 0x%" PRIx64, pTscObj->pAppInfo->pTransporter, pTscObj->id);
+  tscDebug("free transporter:%p in connObj: 0x%" PRIx64, pTscObj->pAppInfo->pTransporter, *(int64_t*)pTscObj->id);
   rpcClose(pTscObj->pAppInfo->pTransporter);
 }
 
@@ -128,7 +128,7 @@ void closeAllRequests(SHashObj *pRequests) {
 void destroyTscObj(void *pObj) {
   STscObj *pTscObj = pObj;
 
-  SClientHbKey connKey = {.tscRid = pTscObj->id, .connType = pTscObj->connType};
+  SClientHbKey connKey = {.tscRid = *(int64_t*)pTscObj->id, .connType = pTscObj->connType};
   hbDeregisterConn(pTscObj->pAppInfo->pAppHbMgr, connKey);
   int64_t connNum = atomic_sub_fetch_64(&pTscObj->pAppInfo->numOfConns, 1);
   closeAllRequests(pTscObj->pRequests);
@@ -137,7 +137,7 @@ void destroyTscObj(void *pObj) {
     // TODO 
     //closeTransporter(pTscObj);
   }
-  tscDebug("connObj 0x%" PRIx64 " destroyed, totalConn:%" PRId64, pTscObj->id, pTscObj->pAppInfo->numOfConns);
+  tscDebug("connObj 0x%" PRIx64 " destroyed, totalConn:%" PRId64, *(int64_t*)pTscObj->id, pTscObj->pAppInfo->numOfConns);
   taosThreadMutexDestroy(&pTscObj->mutex);
   taosMemoryFreeClear(pTscObj);
 }
@@ -166,10 +166,11 @@ void *createTscObj(const char *user, const char *auth, const char *db, int32_t c
   }
 
   taosThreadMutexInit(&pObj->mutex, NULL);
-  pObj->id = taosAddRef(clientConnRefPool, pObj);
+  pObj->id = taosMemoryMalloc(sizeof(int64_t));
+  *(int64_t*)pObj->id = taosAddRef(clientConnRefPool, pObj);
   pObj->schemalessType = 1;
 
-  tscDebug("connObj created, 0x%" PRIx64, pObj->id);
+  tscDebug("connObj created, 0x%" PRIx64, *(int64_t*)pObj->id);
   return pObj;
 }
 
@@ -211,6 +212,7 @@ void doFreeReqResultInfo(SReqResultInfo *pResInfo) {
   taosMemoryFreeClear(pResInfo->pCol);
   taosMemoryFreeClear(pResInfo->fields);
   taosMemoryFreeClear(pResInfo->userFields);
+  taosMemoryFreeClear(pResInfo->convertJson);
 
   if (pResInfo->convertBuf != NULL) {
     for (int32_t i = 0; i < pResInfo->numOfCols; ++i) {
@@ -229,7 +231,7 @@ static void doDestroyRequest(void *p) {
   taosHashRemove(pRequest->pTscObj->pRequests, &pRequest->self, sizeof(pRequest->self));
 
   if (pRequest->body.queryJob != 0) {
-    schedulerFreeJob(pRequest->body.queryJob);
+    schedulerFreeJob(pRequest->body.queryJob, 0);
   }
 
   taosMemoryFreeClear(pRequest->msgBuf);

source/client/src/clientHb.c

@@ -164,6 +164,7 @@ static int32_t hbQueryHbRspHandle(SAppHbMgr *pAppHbMgr, SClientHbRsp *pRsp) {
       pTscObj->connId = pRsp->query->connId;
 
       if (pRsp->query->killRid) {
+        tscDebug("request rid %" PRIx64 " need to be killed now", pRsp->query->killRid);
         SRequestObj *pRequest = acquireRequest(pRsp->query->killRid);
         if (NULL == pRequest) {
           tscDebug("request 0x%" PRIx64 " not exist to kill", pRsp->query->killRid);
@@ -304,7 +305,7 @@ int32_t hbBuildQueryDesc(SQueryHbReqBasic *hbBasic, STscObj *pObj) {
   while (pIter != NULL) {
     int64_t     *rid = pIter;
     SRequestObj *pRequest = acquireRequest(*rid);
-    if (NULL == pRequest) {
+    if (NULL == pRequest || pRequest->killed) {
       pIter = taosHashIterate(pObj->pRequests, pIter);
       continue;
     }
@@ -314,7 +315,6 @@ int32_t hbBuildQueryDesc(SQueryHbReqBasic *hbBasic, STscObj *pObj) {
     desc.queryId = pRequest->requestId;
     desc.useconds = now - pRequest->metric.start;
     desc.reqRid = pRequest->self;
-    desc.pid = hbBasic->pid;
     desc.stableQuery = pRequest->stableQuery;
     taosGetFqdn(desc.fqdn);
     desc.subPlanNum = pRequest->body.pDag ? pRequest->body.pDag->numOfSubplans : 0;
@@ -360,8 +360,6 @@ int32_t hbGetQueryBasicInfo(SClientHbKey *connKey, SClientHbReq *req) {
   }
   
   hbBasic->connId = pTscObj->connId;
-  hbBasic->pid = taosGetPId();
-  taosGetAppName(hbBasic->app, NULL);
 
   int32_t numOfQueries = pTscObj->pRequests ? taosHashGetSize(pTscObj->pRequests) : 0;
   if (numOfQueries <= 0) {
@@ -507,6 +505,21 @@ int32_t hbGetExpiredStbInfo(SClientHbKey *connKey, struct SCatalog *pCatalog, SC
   return TSDB_CODE_SUCCESS;
 }
 
+int32_t hbGetAppInfo(int64_t clusterId, SClientHbReq *req) {
+  SAppHbReq* pApp = taosHashGet(clientHbMgr.appSummary, &clusterId, sizeof(clusterId));
+  if (NULL != pApp) {
+    memcpy(&req->app, pApp, sizeof(*pApp));
+  } else {
+    memset(&req->app.summary, 0, sizeof(req->app.summary));
+    req->app.pid = taosGetPId();
+    req->app.appId = clientHbMgr.appId;
+    taosGetAppName(req->app.name, NULL);    
+  }
+
+  return TSDB_CODE_SUCCESS;
+}
+
+
 int32_t hbQueryHbReqHandle(SClientHbKey *connKey, void *param, SClientHbReq *req) {
   int64_t         *clusterId = (int64_t *)param;
   struct SCatalog *pCatalog = NULL;
@@ -517,6 +530,8 @@ int32_t hbQueryHbReqHandle(SClientHbKey *connKey, void *param, SClientHbReq *req
     return code;
   }
 
+  hbGetAppInfo(*clusterId, req);
+
   hbGetQueryBasicInfo(connKey, req);
 
   code = hbGetExpiredUserInfo(connKey, pCatalog, req);
@@ -589,10 +604,56 @@ void hbThreadFuncUnexpectedStopped(void) {
   atomic_store_8(&clientHbMgr.threadStop, 2);
 }
 
+void hbMergeSummary(SAppClusterSummary* dst, SAppClusterSummary* src) {
+  dst->numOfInsertsReq += src->numOfInsertsReq;
+  dst->numOfInsertRows += src->numOfInsertRows;
+  dst->insertElapsedTime += src->insertElapsedTime;
+  dst->insertBytes += src->insertBytes;
+  dst->fetchBytes += src->fetchBytes;
+  dst->queryElapsedTime += src->queryElapsedTime;
+  dst->numOfSlowQueries += src->numOfSlowQueries;
+  dst->totalRequests += src->totalRequests;
+  dst->currentRequests += src->currentRequests;
+}
+
+int32_t hbGatherAppInfo(void) {
+  SAppHbReq req = {0};
+  int sz = taosArrayGetSize(clientHbMgr.appHbMgrs);
+  if (sz > 0) {
+    req.pid = taosGetPId();
+    req.appId = clientHbMgr.appId;
+    taosGetAppName(req.name, NULL);
+  }
+
+  taosHashClear(clientHbMgr.appSummary);
+  
+  for (int32_t i = 0; i < sz; ++i) {
+    SAppHbMgr *pAppHbMgr = taosArrayGetP(clientHbMgr.appHbMgrs, i);
+    uint64_t clusterId = pAppHbMgr->pAppInstInfo->clusterId;
+    SAppHbReq* pApp = taosHashGet(clientHbMgr.appSummary, &clusterId, sizeof(clusterId));
+    if (NULL == pApp) {
+      memcpy(&req.summary, &pAppHbMgr->pAppInstInfo->summary, sizeof(req.summary));
+      req.startTime = pAppHbMgr->startTime;
+      taosHashPut(clientHbMgr.appSummary, &clusterId, sizeof(clusterId), &req, sizeof(req));
+    } else {
+      if (pAppHbMgr->startTime < pApp->startTime) {
+        pApp->startTime = pAppHbMgr->startTime;
+      }
+      
+      hbMergeSummary(&pApp->summary, &pAppHbMgr->pAppInstInfo->summary);
+    }
+  }
+
+  return TSDB_CODE_SUCCESS;
+}
+
+
 static void *hbThreadFunc(void *param) {
   setThreadName("hb");
 #ifdef WINDOWS
+  if (taosCheckCurrentInDll()) {
     atexit(hbThreadFuncUnexpectedStopped);
+  }
 #endif
   while (1) {
     int8_t threadStop = atomic_val_compare_exchange_8(&clientHbMgr.threadStop, 1, 2);
@@ -603,6 +664,10 @@ static void *hbThreadFunc(void *param) {
     taosThreadMutexLock(&clientHbMgr.lock);
 
     int sz = taosArrayGetSize(clientHbMgr.appHbMgrs);
+    if (sz > 0) {
+      hbGatherAppInfo();
+    }
+    
     for (int i = 0; i < sz; i++) {
       SAppHbMgr *pAppHbMgr = taosArrayGetP(clientHbMgr.appHbMgrs, i);
 
@@ -746,6 +811,10 @@ int hbMgrInit() {
   int8_t old = atomic_val_compare_exchange_8(&clientHbMgr.inited, 0, 1);
   if (old == 1) return 0;
 
+  clientHbMgr.appId = tGenIdPI64();
+  tscDebug("app %" PRIx64 " initialized", clientHbMgr.appId);
+  
+  clientHbMgr.appSummary = taosHashInit(10, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BIGINT), false, HASH_NO_LOCK);
   clientHbMgr.appHbMgrs = taosArrayInit(0, sizeof(void *));
   taosThreadMutexInit(&clientHbMgr.lock, NULL);
 

source/client/src/clientImpl.c

@@ -263,6 +263,7 @@ void asyncExecLocalCmd(SRequestObj* pRequest, SQuery* pQuery) {
 int32_t asyncExecDdlQuery(SRequestObj* pRequest, SQuery* pQuery) {
   // drop table if exists not_exists_table
   if (NULL == pQuery->pCmdMsg) {
+    pRequest->body.queryFp(pRequest->body.param, pRequest, 0);
     return TSDB_CODE_SUCCESS;
   }
 
@@ -306,6 +307,21 @@ int32_t updateQnodeList(SAppInstInfo* pInfo, SArray* pNodeList) {
   return TSDB_CODE_SUCCESS;
 }
 
+bool qnodeRequired(SRequestObj* pRequest) {
+  if (QUERY_POLICY_VNODE == tsQueryPolicy) {
+    return false;
+  }
+
+  SAppInstInfo* pInfo = pRequest->pTscObj->pAppInfo;
+  bool required = false;
+  
+  taosThreadMutexLock(&pInfo->qnodeMutex);
+  required = (NULL == pInfo->pQnodeList);
+  taosThreadMutexUnlock(&pInfo->qnodeMutex);
+
+  return required;
+}
+
 int32_t getQnodeList(SRequestObj* pRequest, SArray** pNodeList) {
   SAppInstInfo* pInfo = pRequest->pTscObj->pAppInfo;
   int32_t       code = 0;
@@ -336,7 +352,7 @@ int32_t getQnodeList(SRequestObj* pRequest, SArray** pNodeList) {
   return code;
 }
 
-int32_t getPlan(SRequestObj* pRequest, SQuery* pQuery, SQueryPlan** pPlan, SArray** pNodeList) {
+int32_t getPlan(SRequestObj* pRequest, SQuery* pQuery, SQueryPlan** pPlan, SArray* pNodeList) {
   pRequest->type = pQuery->msgType;
   SAppInstInfo* pAppInfo = getAppInfo(pRequest);
 
@@ -348,12 +364,7 @@ int32_t getPlan(SRequestObj* pRequest, SQuery* pQuery, SQueryPlan** pPlan, SArra
                       .pMsg = pRequest->msgBuf,
                       .msgLen = ERROR_MSG_BUF_DEFAULT_SIZE};
 
-  int32_t code = getQnodeList(pRequest, pNodeList);
+  return qCreateQueryPlan(&cxt, pPlan, pNodeList);
-  if (TSDB_CODE_SUCCESS == code) {
-    code = qCreateQueryPlan(&cxt, pPlan, *pNodeList);
-  }
-
-  return code;
 }
 
 void setResSchemaInfo(SReqResultInfo* pResInfo, const SSchema* pSchema, int32_t numOfCols) {
@@ -396,6 +407,195 @@ void setResPrecision(SReqResultInfo* pResInfo, int32_t precision) {
   pResInfo->precision = precision;
 }
 
+int32_t buildVnodePolicyNodeList(SRequestObj* pRequest, SArray** pNodeList, SArray* pMnodeList, SArray* pDbVgList) {
+  SArray* nodeList = taosArrayInit(4, sizeof(SQueryNodeLoad));
+
+  int32_t dbNum = taosArrayGetSize(pDbVgList);
+  for (int32_t i = 0; i < dbNum; ++i) {
+    SArray* pVg = taosArrayGetP(pDbVgList, i);
+    int32_t vgNum = taosArrayGetSize(pVg);
+    if (vgNum <= 0) {
+      continue;
+    }
+
+    for (int32_t j = 0; j < vgNum; ++j) {
+      SVgroupInfo* pInfo = taosArrayGet(pVg, j);
+      SQueryNodeLoad load = {0};
+      load.addr.nodeId = pInfo->vgId;
+      load.addr.epSet = pInfo->epSet;
+      
+      taosArrayPush(nodeList, &load);
+    }
+  }
+
+  int32_t vnodeNum = taosArrayGetSize(nodeList);
+  if (vnodeNum > 0) {
+    tscDebug("0x%" PRIx64 " vnode policy, use vnode list, num:%d", pRequest->requestId, vnodeNum);
+    goto _return;
+  }
+
+  int32_t mnodeNum = taosArrayGetSize(pMnodeList);
+  if (mnodeNum <= 0) {
+    tscDebug("0x%" PRIx64 " vnode policy, empty node list", pRequest->requestId);
+    goto _return;
+  }
+
+  void* pData = taosArrayGet(pMnodeList, 0);
+  taosArrayAddBatch(nodeList, pData, mnodeNum);
+
+  tscDebug("0x%" PRIx64 " vnode policy, use mnode list, num:%d", pRequest->requestId, mnodeNum);
+
+_return:
+
+  *pNodeList = nodeList;
+
+  return TSDB_CODE_SUCCESS;
+}
+
+int32_t buildQnodePolicyNodeList(SRequestObj* pRequest, SArray** pNodeList, SArray* pMnodeList, SArray* pQnodeList) {
+  SArray* nodeList = taosArrayInit(4, sizeof(SQueryNodeLoad));
+
+  int32_t qNodeNum = taosArrayGetSize(pQnodeList);
+  if (qNodeNum > 0) {
+    void* pData = taosArrayGet(pQnodeList, 0);
+    taosArrayAddBatch(nodeList, pData, qNodeNum);
+    tscDebug("0x%" PRIx64 " qnode policy, use qnode list, num:%d", pRequest->requestId, qNodeNum);
+    goto _return;
+  }
+
+  int32_t mnodeNum = taosArrayGetSize(pMnodeList);
+  if (mnodeNum <= 0) {
+    tscDebug("0x%" PRIx64 " qnode policy, empty node list", pRequest->requestId);
+    goto _return;
+  }
+
+  void* pData = taosArrayGet(pMnodeList, 0);
+  taosArrayAddBatch(nodeList, pData, mnodeNum);
+
+  tscDebug("0x%" PRIx64 " qnode policy, use mnode list, num:%d", pRequest->requestId, mnodeNum);
+
+_return:
+
+  *pNodeList = nodeList;
+
+  return TSDB_CODE_SUCCESS;
+}
+
+
+int32_t buildAsyncExecNodeList(SRequestObj* pRequest, SArray** pNodeList, SArray* pMnodeList, SMetaData *pResultMeta) {
+  SArray* pDbVgList = NULL;
+  SArray* pQnodeList = NULL;
+  int32_t code = 0;
+  
+  switch (tsQueryPolicy) {
+    case QUERY_POLICY_VNODE: {
+      if (pResultMeta) {
+        pDbVgList = taosArrayInit(4, POINTER_BYTES);
+        
+        int32_t dbNum = taosArrayGetSize(pResultMeta->pDbVgroup);
+        for (int32_t i = 0; i < dbNum; ++i) {
+          SMetaRes* pRes = taosArrayGet(pResultMeta->pDbVgroup, i);
+          if (pRes->code || NULL == pRes->pRes) {
+            continue;
+          }
+
+          taosArrayPush(pDbVgList, &pRes->pRes);
+        } 
+      }
+    
+      code = buildVnodePolicyNodeList(pRequest, pNodeList, pMnodeList, pDbVgList);
+      break;
+    }
+    case QUERY_POLICY_HYBRID:
+    case QUERY_POLICY_QNODE: {
+      if (pResultMeta && taosArrayGetSize(pResultMeta->pQnodeList) > 0) {
+        SMetaRes* pRes = taosArrayGet(pResultMeta->pQnodeList, 0);
+        if (pRes->code) {
+          pQnodeList = NULL;
+        } else {
+          pQnodeList = taosArrayDup((SArray*)pRes->pRes);
+        }
+      } else {
+        SAppInstInfo* pInst = pRequest->pTscObj->pAppInfo;
+        taosThreadMutexLock(&pInst->qnodeMutex);
+        if (pInst->pQnodeList) {
+          pQnodeList = taosArrayDup(pInst->pQnodeList);
+        }
+        taosThreadMutexUnlock(&pInst->qnodeMutex);
+      }
+      
+      code = buildQnodePolicyNodeList(pRequest, pNodeList, pMnodeList, pQnodeList);
+      break;
+    }
+    default:
+      tscError("unknown query policy: %d", tsQueryPolicy);
+      return TSDB_CODE_TSC_APP_ERROR;
+  }
+
+  taosArrayDestroy(pDbVgList);
+  taosArrayDestroy(pQnodeList);
+  
+  return code;
+}
+
+int32_t buildSyncExecNodeList(SRequestObj* pRequest, SArray** pNodeList, SArray* pMnodeList) {
+  SArray* pDbVgList = NULL;
+  SArray* pQnodeList = NULL;
+  int32_t code = 0;
+  
+  switch (tsQueryPolicy) {
+    case QUERY_POLICY_VNODE: {
+      int32_t dbNum = taosArrayGetSize(pRequest->dbList);
+      if (dbNum > 0) {
+        SCatalog* pCtg = NULL;
+        SAppInstInfo* pInst = pRequest->pTscObj->pAppInfo;
+        code = catalogGetHandle(pInst->clusterId, &pCtg);
+        if (code != TSDB_CODE_SUCCESS) {
+          goto _return;
+        }
+
+        pDbVgList = taosArrayInit(dbNum, POINTER_BYTES);   
+        SArray* pVgList = NULL;
+        for (int32_t i = 0; i < dbNum; ++i) {
+          char* dbFName = taosArrayGet(pRequest->dbList, i);
+          SRequestConnInfo conn = {.pTrans = pInst->pTransporter,
+                                   .requestId = pRequest->requestId,
+                                   .requestObjRefId = pRequest->self,
+                                   .mgmtEps = getEpSet_s(&pInst->mgmtEp)};    
+                                   
+          code = catalogGetDBVgInfo(pCtg, &conn, dbFName, &pVgList);
+          if (code) {
+            goto _return;
+          }
+        
+          taosArrayPush(pDbVgList, &pVgList);
+        } 
+      }
+    
+      code = buildVnodePolicyNodeList(pRequest, pNodeList, pMnodeList, pDbVgList);
+      break;
+    }
+    case QUERY_POLICY_HYBRID:
+    case QUERY_POLICY_QNODE: {
+      getQnodeList(pRequest, &pQnodeList);
+      
+      code = buildQnodePolicyNodeList(pRequest, pNodeList, pMnodeList, pQnodeList);
+      break;
+    }
+    default:
+      tscError("unknown query policy: %d", tsQueryPolicy);
+      return TSDB_CODE_TSC_APP_ERROR;
+  }
+
+_return:
+
+  taosArrayDestroy(pDbVgList);
+  taosArrayDestroy(pQnodeList);
+  
+  return code;
+}
+
+
 int32_t scheduleAsyncQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNodeList) {
   tsem_init(&schdRspSem, 0, 0);
 
@@ -418,7 +618,7 @@ int32_t scheduleAsyncQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNod
   while (true) {
     if (code != TSDB_CODE_SUCCESS) {
       if (pRequest->body.queryJob != 0) {
-        schedulerFreeJob(pRequest->body.queryJob);
+        schedulerFreeJob(pRequest->body.queryJob, 0);
       }
 
       pRequest->code = code;
@@ -439,7 +639,7 @@ int32_t scheduleAsyncQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNod
     pRequest->body.resInfo.numOfRows = res.numOfRows;
 
     if (pRequest->body.queryJob != 0) {
-      schedulerFreeJob(pRequest->body.queryJob);
+      schedulerFreeJob(pRequest->body.queryJob, 0);
     }
   }
 
@@ -461,14 +661,15 @@ int32_t scheduleQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNodeList
                           .sql = pRequest->sqlstr,
                           .startTs = pRequest->metric.start,
                           .fp = NULL,
-                          .cbParam = NULL};
+                          .cbParam = NULL,
+                          .reqKilled = &pRequest->killed};
 
   int32_t code = schedulerExecJob(&req, &pRequest->body.queryJob, &res);
   pRequest->body.resInfo.execRes = res.res;
 
   if (code != TSDB_CODE_SUCCESS) {
     if (pRequest->body.queryJob != 0) {
-      schedulerFreeJob(pRequest->body.queryJob);
+      schedulerFreeJob(pRequest->body.queryJob, 0);
     }
 
     pRequest->code = code;
@@ -481,7 +682,7 @@ int32_t scheduleQuery(SRequestObj* pRequest, SQueryPlan* pDag, SArray* pNodeList
     pRequest->body.resInfo.numOfRows = res.numOfRows;
 
     if (pRequest->body.queryJob != 0) {
-      schedulerFreeJob(pRequest->body.queryJob);
+      schedulerFreeJob(pRequest->body.queryJob, 0);
     }
   }
 
@@ -608,6 +809,19 @@ void schedulerExecCb(SQueryResult* pResult, void* param, int32_t code) {
   SRequestObj* pRequest = (SRequestObj*)param;
   pRequest->code = code;
 
+  if (TDMT_VND_SUBMIT == pRequest->type || TDMT_VND_DELETE == pRequest->type ||
+      TDMT_VND_CREATE_TABLE == pRequest->type) {
+    pRequest->body.resInfo.numOfRows = pResult->numOfRows;
+
+    if (pRequest->body.queryJob != 0) {
+      schedulerFreeJob(pRequest->body.queryJob, 0);
+      pRequest->body.queryJob = 0;
+    }
+  }
+
+  tscDebug("0x%" PRIx64 " enter scheduler exec cb, code:%d - %s, reqId:0x%" PRIx64,
+             pRequest->self, code, tstrerror(code), pRequest->requestId);
+
   STscObj* pTscObj = pRequest->pTscObj;
   if (code != TSDB_CODE_SUCCESS && NEED_CLIENT_HANDLE_ERROR(code)) {
     tscDebug("0x%" PRIx64 " client retry to handle the error, code:%d - %s, tryCount:%d, reqId:0x%" PRIx64,
@@ -643,12 +857,16 @@ SRequestObj* launchQueryImpl(SRequestObj* pRequest, SQuery* pQuery, bool keepQue
       code = execDdlQuery(pRequest, pQuery);
       break;
     case QUERY_EXEC_MODE_SCHEDULE: {
-      SArray* pNodeList = NULL;
+      SArray* pMnodeList = taosArrayInit(4, sizeof(SQueryNodeLoad));
-      code = getPlan(pRequest, pQuery, &pRequest->body.pDag, &pNodeList);
+      code = getPlan(pRequest, pQuery, &pRequest->body.pDag, pMnodeList);
       if (TSDB_CODE_SUCCESS == code) {
+        SArray* pNodeList = NULL;
+        buildSyncExecNodeList(pRequest, &pNodeList, pMnodeList);
+        
         code = scheduleQuery(pRequest, pRequest->body.pDag, pNodeList);
-      }
         taosArrayDestroy(pNodeList);
+      }
+      taosArrayDestroy(pMnodeList);
       break;
     }
     case QUERY_EXEC_MODE_EMPTY_RESULT:
@@ -697,7 +915,7 @@ SRequestObj* launchQuery(STscObj* pTscObj, const char* sql, int sqlLen) {
   return launchQueryImpl(pRequest, pQuery, false, NULL);
 }
 
-void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery) {
+void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery, SMetaData *pResultMeta) {
   int32_t code = 0;
 
   switch (pQuery->execMode) {
@@ -708,7 +926,7 @@ void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery) {
       code = asyncExecDdlQuery(pRequest, pQuery);
       break;
     case QUERY_EXEC_MODE_SCHEDULE: {
-      SArray* pNodeList = taosArrayInit(4, sizeof(struct SQueryNodeAddr));
+      SArray* pMnodeList = taosArrayInit(4, sizeof(SQueryNodeLoad));
 
       pRequest->type = pQuery->msgType;
 
@@ -721,15 +939,16 @@ void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery) {
                           .msgLen = ERROR_MSG_BUF_DEFAULT_SIZE};
 
       SAppInstInfo* pAppInfo = getAppInfo(pRequest);
-      if (TSDB_CODE_SUCCESS == code) {
+      code = qCreateQueryPlan(&cxt, &pRequest->body.pDag, pMnodeList);
-        code = qCreateQueryPlan(&cxt, &pRequest->body.pDag, pNodeList);
       if (code) {
         tscError("0x%" PRIx64 " failed to create query plan, code:%s 0x%" PRIx64, pRequest->self, tstrerror(code),
                  pRequest->requestId);
       }
-      }
 
       if (TSDB_CODE_SUCCESS == code) {
+        SArray* pNodeList = NULL;
+        buildAsyncExecNodeList(pRequest, &pNodeList, pMnodeList, pResultMeta);
+        
         SRequestConnInfo conn = {
             .pTrans = pAppInfo->pTransporter, .requestId = pRequest->requestId, .requestObjRefId = pRequest->self};
         SSchedulerReq req = {.pConn = &conn,
@@ -738,8 +957,10 @@ void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery) {
                              .sql = pRequest->sqlstr,
                              .startTs = pRequest->metric.start,
                              .fp = schedulerExecCb,
-                             .cbParam = pRequest};
+                             .cbParam = pRequest,
+                             .reqKilled = &pRequest->killed};
         code = schedulerAsyncExecJob(&req, &pRequest->body.queryJob);
+        taosArrayDestroy(pNodeList);
       } else {
         tscError("0x%" PRIx64 " failed to create query plan, code:%s 0x%" PRIx64, pRequest->self, tstrerror(code),
                  pRequest->requestId);
@@ -747,7 +968,7 @@ void launchAsyncQuery(SRequestObj* pRequest, SQuery* pQuery) {
       }
 
       // todo not to be released here
-      taosArrayDestroy(pNodeList);
+      taosArrayDestroy(pMnodeList);
       break;
     }
     case QUERY_EXEC_MODE_EMPTY_RESULT:
@@ -922,7 +1143,7 @@ STscObj* taosConnectImpl(const char* user, const char* auth, const char* db, __t
     taos_close_internal(pTscObj);
     pTscObj = NULL;
   } else {
-    tscDebug("0x%" PRIx64 " connection is opening, connId:%u, dnodeConn:%p, reqId:0x%" PRIx64, pTscObj->id,
+    tscDebug("0x%" PRIx64 " connection is opening, connId:%u, dnodeConn:%p, reqId:0x%" PRIx64, *(int64_t*)pTscObj->id,
              pTscObj->connId, pTscObj->pAppInfo->pTransporter, pRequest->requestId);
     destroyRequest(pRequest);
   }
@@ -1085,10 +1306,10 @@ TAOS* taos_connect_auth(const char* ip, const char* user, const char* auth, cons
 
   STscObj* pObj = taos_connect_internal(ip, user, NULL, auth, db, port, CONN_TYPE__QUERY);
   if (pObj) {
-    return (TAOS*)pObj->id;
+    return pObj->id;
   }
   
-  return (TAOS*)0;
+  return NULL;
 }
 
 TAOS* taos_connect_l(const char* ip, int ipLen, const char* user, int userLen, const char* pass, int passLen,
@@ -1327,27 +1548,126 @@ static int32_t doConvertUCS4(SReqResultInfo* pResultInfo, int32_t numOfRows, int
 
       pResultInfo->pCol[i].pData = pResultInfo->convertBuf[i];
       pResultInfo->row[i] = pResultInfo->pCol[i].pData;
-    } else if (type == TSDB_DATA_TYPE_JSON && colLength[i] > 0) {
+    }
-      char* p = taosMemoryRealloc(pResultInfo->convertBuf[i], colLength[i]);
-      if (p == NULL) {
-        return TSDB_CODE_OUT_OF_MEMORY;
   }
 
-      pResultInfo->convertBuf[i] = p;
+  return TSDB_CODE_SUCCESS;
-      int32_t        len = 0;
+}
-      SResultColumn* pCol = &pResultInfo->pCol[i];
-      for (int32_t j = 0; j < numOfRows; ++j) {
-        if (pCol->offset[j] != -1) {
-          char* pStart = pCol->offset[j] + pCol->pData;
 
-          int32_t jsonInnerType = *pStart;
+static int32_t estimateJsonLen(SReqResultInfo* pResultInfo, int32_t numOfCols, int32_t numOfRows){
-          char*   jsonInnerData = pStart + CHAR_BYTES;
+  char* p = (char*)pResultInfo->pData;
+
+  int32_t len = sizeof(int32_t) + sizeof(uint64_t) + numOfCols * (sizeof(int16_t) + sizeof(int32_t));
+  int32_t* colLength = (int32_t*)(p + len);
+  len += sizeof(int32_t) * numOfCols;
+
+  char* pStart = p + len;
+  for (int32_t i = 0; i < numOfCols; ++i) {
+    int32_t colLen = htonl(colLength[i]);
+
+    if (pResultInfo->fields[i].type == TSDB_DATA_TYPE_JSON) {
+      int32_t* offset = (int32_t*)pStart;
+      int32_t lenTmp = numOfRows * sizeof(int32_t);
+      len += lenTmp;
+      pStart += lenTmp;
+
+      for (int32_t j = 0; j < numOfRows; ++j) {
+        if (offset[j] == -1) {
+          continue;
+        }
+        char* data = offset[j] + pStart;
+
+        int32_t jsonInnerType = *data;
+        char*   jsonInnerData = data + CHAR_BYTES;
+        if (jsonInnerType == TSDB_DATA_TYPE_NULL) {
+          len += (VARSTR_HEADER_SIZE + strlen(TSDB_DATA_NULL_STR_L));
+        } else if (jsonInnerType & TD_TAG_JSON) {
+          len += (VARSTR_HEADER_SIZE + ((const STag*)(data))->len);
+        } else if (jsonInnerType == TSDB_DATA_TYPE_NCHAR) {  // value -> "value"
+          len += varDataTLen(jsonInnerData) + CHAR_BYTES * 2;
+        } else if (jsonInnerType == TSDB_DATA_TYPE_DOUBLE) {
+          len += (VARSTR_HEADER_SIZE + 32);
+        } else if (jsonInnerType == TSDB_DATA_TYPE_BOOL) {
+          len += (VARSTR_HEADER_SIZE + 5);
+        } else {
+          ASSERT(0);
+        }
+
+      }
+    } else if (IS_VAR_DATA_TYPE(pResultInfo->fields[i].type)) {
+      int32_t lenTmp = numOfRows * sizeof(int32_t);
+      len += (lenTmp + colLen);
+      pStart += lenTmp;
+    } else {
+      int32_t lenTmp = BitmapLen(pResultInfo->numOfRows);
+      len += (lenTmp + colLen);
+      pStart += lenTmp;
+    }
+    pStart += colLen;
+  }
+  return len;
+}
+
+static int32_t doConvertJson(SReqResultInfo* pResultInfo, int32_t numOfCols, int32_t numOfRows) {
+  bool needConvert = false;
+  for (int32_t i = 0; i < numOfCols; ++i) {
+    if (pResultInfo->fields[i].type == TSDB_DATA_TYPE_JSON) {
+      needConvert = true;
+      break;
+    }
+  }
+  if(!needConvert) return TSDB_CODE_SUCCESS;
+
+  char* p = (char*)pResultInfo->pData;
+  int32_t dataLen = estimateJsonLen(pResultInfo, numOfCols, numOfRows);
+
+  pResultInfo->convertJson = taosMemoryCalloc(1, dataLen);
+  if(pResultInfo->convertJson == NULL) return TSDB_CODE_OUT_OF_MEMORY;
+  char* p1 = pResultInfo->convertJson;
+
+  int32_t len = sizeof(int32_t) + sizeof(uint64_t) + numOfCols * (sizeof(int16_t) + sizeof(int32_t));
+  memcpy(p1, p, len);
+
+  p += len;
+  p1 += len;
+
+  len = sizeof(int32_t) * numOfCols;
+  int32_t* colLength = (int32_t*)p;
+  int32_t* colLength1 = (int32_t*)p1;
+  memcpy(p1, p, len);
+  p += len;
+  p1 += len;
+
+  char* pStart = p;
+  char* pStart1 = p1;
+  for (int32_t i = 0; i < numOfCols; ++i) {
+    int32_t colLen = htonl(colLength[i]);
+    int32_t colLen1 = htonl(colLength1[i]);
+    ASSERT(colLen < dataLen);
+
+    if (pResultInfo->fields[i].type == TSDB_DATA_TYPE_JSON) {
+      int32_t* offset = (int32_t*)pStart;
+      int32_t* offset1 = (int32_t*)pStart1;
+      len = numOfRows * sizeof(int32_t);
+      memcpy(pStart1, pStart, len);
+      pStart += len;
+      pStart1 += len;
+
+      len = 0;
+      for (int32_t j = 0; j < numOfRows; ++j) {
+        if (offset[j] == -1) {
+          continue;
+        }
+        char* data = offset[j] + pStart;
+
+        int32_t jsonInnerType = *data;
+        char*   jsonInnerData = data + CHAR_BYTES;
         char    dst[TSDB_MAX_JSON_TAG_LEN] = {0};
         if (jsonInnerType == TSDB_DATA_TYPE_NULL) {
           sprintf(varDataVal(dst), "%s", TSDB_DATA_NULL_STR_L);
           varDataSetLen(dst, strlen(varDataVal(dst)));
-          } else if (jsonInnerType == TD_TAG_JSON) {
+        } else if (jsonInnerType & TD_TAG_JSON) {
-            char* jsonString = parseTagDatatoJson(pStart);
+          char* jsonString = parseTagDatatoJson(data);
           STR_TO_VARSTR(dst, jsonString);
           taosMemoryFree(jsonString);
         } else if (jsonInnerType == TSDB_DATA_TYPE_NCHAR) {  // value -> "value"
@@ -1371,26 +1691,31 @@ static int32_t doConvertUCS4(SReqResultInfo* pResultInfo, int32_t numOfRows, int
           ASSERT(0);
         }
 
-          if (len + varDataTLen(dst) > colLength[i]) {
+        offset1[j]= len;
-            p = taosMemoryRealloc(pResultInfo->convertBuf[i], len + varDataTLen(dst));
+        memcpy(pStart1 + len, dst, varDataTLen(dst));
-            if (p == NULL) {
-              return TSDB_CODE_OUT_OF_MEMORY;
-            }
-
-            pResultInfo->convertBuf[i] = p;
-          }
-          p = pResultInfo->convertBuf[i] + len;
-          memcpy(p, dst, varDataTLen(dst));
-          pCol->offset[j] = len;
         len += varDataTLen(dst);
       }
+      colLen1 = len;
+      colLength1[i] = htonl(len);
+    } else if (IS_VAR_DATA_TYPE(pResultInfo->fields[i].type)) {
+      len = numOfRows * sizeof(int32_t);
+      memcpy(pStart1, pStart, len);
+      pStart += len;
+      pStart1 += len;
+      memcpy(pStart1, pStart, colLen);
+    } else {
+      len = BitmapLen(pResultInfo->numOfRows);
+      memcpy(pStart1, pStart, len);
+      pStart += len;
+      pStart1 += len;
+      memcpy(pStart1, pStart, colLen);
+
+    }
+    pStart += colLen;
+    pStart1 += colLen1;
   }
 
-      pResultInfo->pCol[i].pData = pResultInfo->convertBuf[i];
+  pResultInfo->pData = pResultInfo->convertJson;
-      pResultInfo->row[i] = pResultInfo->pCol[i].pData;
-    }
-  }
-
   return TSDB_CODE_SUCCESS;
 }
 
@@ -1405,6 +1730,10 @@ int32_t setResultDataPtr(SReqResultInfo* pResultInfo, TAOS_FIELD* pFields, int32
   if (code != TSDB_CODE_SUCCESS) {
     return code;
   }
+  code = doConvertJson(pResultInfo, numOfCols, numOfRows);
+  if (code != TSDB_CODE_SUCCESS) {
+    return code;
+  }
 
   char* p = (char*)pResultInfo->pData;
 
@@ -1448,8 +1777,7 @@ int32_t setResultDataPtr(SReqResultInfo* pResultInfo, TAOS_FIELD* pFields, int32
     pStart += colLength[i];
   }
 
-  // convert UCS4-LE encoded character to native multi-bytes character in current data block.
+  if(convertUcs4){
-  if (convertUcs4) {
     code = doConvertUCS4(pResultInfo, numOfRows, numOfCols, colLength);
   }
 

source/client/src/clientMain.c

@@ -99,10 +99,10 @@ TAOS *taos_connect(const char *ip, const char *user, const char *pass, const cha
 
   STscObj* pObj = taos_connect_internal(ip, user, pass, NULL, db, port, CONN_TYPE__QUERY);
   if (pObj) {
-    return (TAOS*)pObj->id;
+    return pObj->id;
   }
   
-  return (TAOS*)0;
+  return NULL;
 }
 
 void taos_close_internal(void *taos) {
@@ -111,19 +111,24 @@ void taos_close_internal(void *taos) {
   }
 
   STscObj *pTscObj = (STscObj *)taos;
-  tscDebug("0x%" PRIx64 " try to close connection, numOfReq:%d", pTscObj->id, pTscObj->numOfReqs);
+  tscDebug("0x%" PRIx64 " try to close connection, numOfReq:%d", *(int64_t*)pTscObj->id, pTscObj->numOfReqs);
 
-  taosRemoveRef(clientConnRefPool, pTscObj->id);
+  taosRemoveRef(clientConnRefPool, *(int64_t*)pTscObj->id);
 }
 
 void taos_close(TAOS *taos) {
-  STscObj* pObj = acquireTscObj((int64_t)taos);
+  if (taos == NULL) {
+    return;
+  }
+  
+  STscObj* pObj = acquireTscObj(*(int64_t*)taos);
   if (NULL == pObj) {
     return;
   }
   
   taos_close_internal(pObj);
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
+  taosMemoryFree(taos);
 }
 
 
@@ -206,7 +211,12 @@ static void syncQueryFn(void *param, void *res, int32_t code) {
 }
 
 TAOS_RES *taos_query(TAOS *taos, const char *sql) {
-  STscObj* pTscObj = acquireTscObj((int64_t)taos);
+  if (NULL == taos) {
+    terrno = TSDB_CODE_TSC_DISCONNECTED;
+    return NULL;
+  }
+  
+  STscObj* pTscObj = acquireTscObj(*(int64_t*)taos);
   if (pTscObj == NULL || sql == NULL) {
     terrno = TSDB_CODE_TSC_DISCONNECTED;
     return NULL;
@@ -219,13 +229,13 @@ TAOS_RES *taos_query(TAOS *taos, const char *sql) {
   taos_query_a(taos, sql, syncQueryFn, param);
   tsem_wait(&param->sem);
 
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
 
   return param->pRequest;
 #else
   size_t sqlLen = strlen(sql);
   if (sqlLen > (size_t)TSDB_MAX_ALLOWED_SQL_LEN) {
-    releaseTscObj((int64_t)taos);
+    releaseTscObj(*(int64_t*)taos);
     tscError("sql string exceeds max length:%d", TSDB_MAX_ALLOWED_SQL_LEN);
     terrno = TSDB_CODE_TSC_EXCEED_SQL_LIMIT;
     return NULL;
@@ -233,7 +243,7 @@ TAOS_RES *taos_query(TAOS *taos, const char *sql) {
 
   TAOS_RES* pRes = execQuery(pTscObj, sql, sqlLen);
 
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
 
   return pRes;
 #endif
@@ -246,13 +256,14 @@ TAOS_ROW taos_fetch_row(TAOS_RES *res) {
 
   if (TD_RES_QUERY(res)) {
     SRequestObj *pRequest = (SRequestObj *)res;
+    if (pRequest->type == TSDB_SQL_RETRIEVE_EMPTY_RESULT || pRequest->type == TSDB_SQL_INSERT ||
+        pRequest->code != TSDB_CODE_SUCCESS || taos_num_fields(res) == 0 || pRequest->killed) {
+      return NULL;
+    }
+
 #if SYNC_ON_TOP_OF_ASYNC
     return doAsyncFetchRows(pRequest, true, true);
 #else
-    if (pRequest->type == TSDB_SQL_RETRIEVE_EMPTY_RESULT || pRequest->type == TSDB_SQL_INSERT ||
-        pRequest->code != TSDB_CODE_SUCCESS || taos_num_fields(res) == 0) {
-      return NULL;
-    }
     return doFetchRows(pRequest, true, true);
 #endif
 
@@ -452,15 +463,15 @@ int taos_result_precision(TAOS_RES *res) {
 }
 
 int taos_select_db(TAOS *taos, const char *db) {
-  STscObj* pObj = acquireTscObj((int64_t)taos);
+  STscObj* pObj = acquireTscObj(*(int64_t*)taos);
   if (pObj == NULL) {
-    releaseTscObj((int64_t)taos);
+    releaseTscObj(*(int64_t*)taos);
     terrno = TSDB_CODE_TSC_DISCONNECTED;
     return TSDB_CODE_TSC_DISCONNECTED;
   }
 
   if (db == NULL || strlen(db) == 0) {
-    releaseTscObj((int64_t)taos);
+    releaseTscObj(*(int64_t*)taos);
     terrno = TSDB_CODE_TSC_INVALID_INPUT;
     return terrno;
   }
@@ -472,7 +483,7 @@ int taos_select_db(TAOS *taos, const char *db) {
   int32_t   code = taos_errno(pRequest);
 
   taos_free_result(pRequest);
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
   return code;
 }
 
@@ -482,14 +493,20 @@ void taos_stop_query(TAOS_RES *res) {
   }
 
   SRequestObj *pRequest = (SRequestObj *)res;
-  int32_t      numOfFields = taos_num_fields(pRequest);
+  pRequest->killed = true;
   
+  int32_t      numOfFields = taos_num_fields(pRequest);
   // It is not a query, no need to stop.
   if (numOfFields == 0) {
+    tscDebug("request %" PRIx64 " no need to be killed since not query", pRequest->requestId);
     return;
   }
 
-  schedulerFreeJob(pRequest->body.queryJob);
+  if (pRequest->body.queryJob) {
+    schedulerFreeJob(pRequest->body.queryJob, TSDB_CODE_TSC_QUERY_KILLED);
+  }
+
+  tscDebug("request %" PRIx64 " killed", pRequest->requestId);
 }
 
 bool taos_is_null(TAOS_RES *res, int32_t row, int32_t col) {
@@ -619,7 +636,7 @@ int *taos_get_column_data_offset(TAOS_RES *res, int columnIndex) {
 int taos_validate_sql(TAOS *taos, const char *sql) { return true; }
 
 void taos_reset_current_db(TAOS *taos) {
-  STscObj* pTscObj = acquireTscObj((int64_t)taos);
+  STscObj* pTscObj = acquireTscObj(*(int64_t*)taos);
   if (pTscObj == NULL) {
     terrno = TSDB_CODE_TSC_DISCONNECTED;
     return;
@@ -627,17 +644,17 @@ void taos_reset_current_db(TAOS *taos) {
 
   resetConnectDB(pTscObj);
 
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
 }
 
 const char *taos_get_server_info(TAOS *taos) {
-  STscObj* pTscObj = acquireTscObj((int64_t)taos);
+  STscObj* pTscObj = acquireTscObj(*(int64_t*)taos);
   if (pTscObj == NULL) {
     terrno = TSDB_CODE_TSC_DISCONNECTED;
     return NULL;
   }
 
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
 
   return pTscObj->ver;
 }
@@ -664,6 +681,8 @@ static void destorySqlParseWrapper(SqlParseWrapper *pWrapper) {
 }
 
 void retrieveMetaCallback(SMetaData *pResultMeta, void *param, int32_t code) {
+  tscDebug("enter meta callback, code %s", tstrerror(code));
+  
   SqlParseWrapper *pWrapper = (SqlParseWrapper *)param;
   SQuery          *pQuery = pWrapper->pQuery;
   SRequestObj     *pRequest = pWrapper->pRequest;
@@ -683,7 +702,7 @@ void retrieveMetaCallback(SMetaData *pResultMeta, void *param, int32_t code) {
     TSWAP(pRequest->tableList, (pQuery)->pTableList);
 
     destorySqlParseWrapper(pWrapper);
-    launchAsyncQuery(pRequest, pQuery);
+    launchAsyncQuery(pRequest, pQuery, pResultMeta);
   } else {
     destorySqlParseWrapper(pWrapper);
     tscDebug("error happens, code:%d", code);
@@ -704,11 +723,11 @@ void retrieveMetaCallback(SMetaData *pResultMeta, void *param, int32_t code) {
 }
 
 void taos_query_a(TAOS *taos, const char *sql, __taos_async_fn_t fp, void *param) {
-  STscObj* pTscObj = acquireTscObj((int64_t)taos);
+  STscObj* pTscObj = acquireTscObj(*(int64_t*)taos);
   if (pTscObj == NULL || sql == NULL || NULL == fp) {
     terrno = TSDB_CODE_INVALID_PARA;
     if (pTscObj) {
-      releaseTscObj((int64_t)taos);
+      releaseTscObj(*(int64_t*)taos);
     } else {
       terrno = TSDB_CODE_TSC_DISCONNECTED;
     }
@@ -789,7 +808,7 @@ void doAsyncQuery(SRequestObj *pRequest, bool updateMetaForce) {
 
   SQuery *pQuery = NULL;
 
-  SCatalogReq catalogReq = {.forceUpdate = updateMetaForce};
+  SCatalogReq catalogReq = {.forceUpdate = updateMetaForce, .qNodeRequired = qnodeRequired(pRequest)};
   code = qParseSqlSyntax(pCxt, &pQuery, &catalogReq);
   if (code != TSDB_CODE_SUCCESS) {
     goto _error;
@@ -830,6 +849,9 @@ static void fetchCallback(void *pResult, void *param, int32_t code) {
 
   SReqResultInfo *pResultInfo = &pRequest->body.resInfo;
 
+  tscDebug("0x%" PRIx64 " enter scheduler fetch cb, code:%d - %s, reqId:0x%" PRIx64,
+             pRequest->self, code, tstrerror(code), pRequest->requestId);
+
   pResultInfo->pData = pResult;
   pResultInfo->numOfRows = 0;
 
@@ -926,7 +948,7 @@ int taos_load_table_info(TAOS *taos, const char *tableNameList) {
 }
 
 TAOS_STMT *taos_stmt_init(TAOS *taos) {
-  STscObj* pObj = acquireTscObj((int64_t)taos);
+  STscObj* pObj = acquireTscObj(*(int64_t*)taos);
   if (NULL == pObj) {
     tscError("invalid parameter for %s", __FUNCTION__);
     terrno = TSDB_CODE_TSC_DISCONNECTED;
@@ -935,7 +957,7 @@ TAOS_STMT *taos_stmt_init(TAOS *taos) {
 
   TAOS_STMT* pStmt = stmtInit(pObj);
   
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
 
   return pStmt;
 }

source/client/src/clientMsgHandler.c

@@ -77,7 +77,7 @@ int32_t processConnectRsp(void* param, const SDataBuf* pMsg, int32_t code) {
 
   for (int32_t i = 0; i < connectRsp.epSet.numOfEps; ++i) {
     tscDebug("0x%" PRIx64 " epSet.fqdn[%d]:%s port:%d, connObj:0x%" PRIx64, pRequest->requestId, i,
-             connectRsp.epSet.eps[i].fqdn, connectRsp.epSet.eps[i].port, pTscObj->id);
+             connectRsp.epSet.eps[i].fqdn, connectRsp.epSet.eps[i].port, *(int64_t*)pTscObj->id);
   }
 
   pTscObj->connId = connectRsp.connId;
@@ -90,7 +90,7 @@ int32_t processConnectRsp(void* param, const SDataBuf* pMsg, int32_t code) {
 
   pTscObj->connType = connectRsp.connType;
 
-  hbRegisterConn(pTscObj->pAppInfo->pAppHbMgr, pTscObj->id, connectRsp.clusterId, connectRsp.connType);
+  hbRegisterConn(pTscObj->pAppInfo->pAppHbMgr, *(int64_t*)pTscObj->id, connectRsp.clusterId, connectRsp.connType);
 
   //  pRequest->body.resInfo.pRspMsg = pMsg->pData;
   tscDebug("0x%" PRIx64 " clusterId:%" PRId64 ", totalConn:%" PRId64, pRequest->requestId, connectRsp.clusterId,

source/client/src/clientSml.c

@@ -309,7 +309,7 @@ static int32_t smlApplySchemaAction(SSmlHandle *info, SSchemaAction *action) {
     case SCHEMA_ACTION_ADD_COLUMN: {
       int n = sprintf(result, "alter stable `%s` add column ", action->alterSTable.sTableName);
       smlBuildColumnDescription(action->alterSTable.field, result + n, capacity - n, &outBytes);
-      TAOS_RES *res = taos_query((TAOS*)info->taos->id, result);  // TODO async doAsyncQuery
+      TAOS_RES *res = taos_query(info->taos->id, result);  // TODO async doAsyncQuery
       code = taos_errno(res);
       const char *errStr = taos_errstr(res);
       if (code != TSDB_CODE_SUCCESS) {
@@ -323,7 +323,7 @@ static int32_t smlApplySchemaAction(SSmlHandle *info, SSchemaAction *action) {
     case SCHEMA_ACTION_ADD_TAG: {
       int n = sprintf(result, "alter stable `%s` add tag ", action->alterSTable.sTableName);
       smlBuildColumnDescription(action->alterSTable.field, result + n, capacity - n, &outBytes);
-      TAOS_RES *res = taos_query((TAOS*)info->taos->id, result);  // TODO async doAsyncQuery
+      TAOS_RES *res = taos_query(info->taos->id, result);  // TODO async doAsyncQuery
       code = taos_errno(res);
       const char *errStr = taos_errstr(res);
       if (code != TSDB_CODE_SUCCESS) {
@@ -337,7 +337,7 @@ static int32_t smlApplySchemaAction(SSmlHandle *info, SSchemaAction *action) {
     case SCHEMA_ACTION_CHANGE_COLUMN_SIZE: {
       int n = sprintf(result, "alter stable `%s` modify column ", action->alterSTable.sTableName);
       smlBuildColumnDescription(action->alterSTable.field, result + n, capacity - n, &outBytes);
-      TAOS_RES *res = taos_query((TAOS*)info->taos->id, result);  // TODO async doAsyncQuery
+      TAOS_RES *res = taos_query(info->taos->id, result);  // TODO async doAsyncQuery
       code = taos_errno(res);
       if (code != TSDB_CODE_SUCCESS) {
         uError("SML:0x%" PRIx64 " apply schema action. error : %s", info->id, taos_errstr(res));
@@ -350,7 +350,7 @@ static int32_t smlApplySchemaAction(SSmlHandle *info, SSchemaAction *action) {
     case SCHEMA_ACTION_CHANGE_TAG_SIZE: {
       int n = sprintf(result, "alter stable `%s` modify tag ", action->alterSTable.sTableName);
       smlBuildColumnDescription(action->alterSTable.field, result + n, capacity - n, &outBytes);
-      TAOS_RES *res = taos_query((TAOS*)info->taos->id, result);  // TODO async doAsyncQuery
+      TAOS_RES *res = taos_query(info->taos->id, result);  // TODO async doAsyncQuery
       code = taos_errno(res);
       if (code != TSDB_CODE_SUCCESS) {
         uError("SML:0x%" PRIx64 " apply schema action. error : %s", info->id, taos_errstr(res));
@@ -405,7 +405,7 @@ static int32_t smlApplySchemaAction(SSmlHandle *info, SSchemaAction *action) {
       pos--;
       ++freeBytes;
       outBytes = snprintf(pos, freeBytes, ")");
-      TAOS_RES *res = taos_query((TAOS*)info->taos->id, result);
+      TAOS_RES *res = taos_query(info->taos->id, result);
       code = taos_errno(res);
       if (code != TSDB_CODE_SUCCESS) {
         uError("SML:0x%" PRIx64 " apply schema action. error : %s", info->id, taos_errstr(res));
@@ -2274,7 +2274,7 @@ static int32_t smlInsertData(SSmlHandle *info) {
   //  info->affectedRows = taos_affected_rows(info->pRequest);
   //  return info->pRequest->code;
 
-  launchAsyncQuery(info->pRequest, info->pQuery);
+  launchAsyncQuery(info->pRequest, info->pQuery, NULL);
   return TSDB_CODE_SUCCESS;
 }
 
@@ -2434,7 +2434,7 @@ static void smlInsertCallback(void *param, void *res, int32_t code) {
  */
 
 TAOS_RES* taos_schemaless_insert(TAOS* taos, char* lines[], int numLines, int protocol, int precision) {
-  STscObj* pTscObj = acquireTscObj((int64_t)taos);
+  STscObj* pTscObj = acquireTscObj(*(int64_t*)taos);
   if (NULL == pTscObj) {
     terrno = TSDB_CODE_TSC_DISCONNECTED;
     uError("SML:taos_schemaless_insert invalid taos");
@@ -2443,7 +2443,7 @@ TAOS_RES* taos_schemaless_insert(TAOS* taos, char* lines[], int numLines, int pr
   
   SRequestObj* request = (SRequestObj*)createRequest(pTscObj, TSDB_SQL_INSERT);
   if(!request){
-    releaseTscObj((int64_t)taos);
+    releaseTscObj(*(int64_t*)taos);
     uError("SML:taos_schemaless_insert error request is null");
     return NULL;
   }
@@ -2531,6 +2531,6 @@ end:
 //  ((STscObj *)taos)->schemalessType = 0;
   pTscObj->schemalessType = 1;
   uDebug("resultend:%s", request->msgBuf);
-  releaseTscObj((int64_t)taos);
+  releaseTscObj(*(int64_t*)taos);
   return (TAOS_RES*)request;
 }

source/client/src/tmq.c

@@ -48,14 +48,6 @@ struct tmq_list_t {
   SArray container;
 };
 
-struct tmq_topic_vgroup_t {
-  SMqOffset offset;
-};
-
-struct tmq_topic_vgroup_list_t {
-  SArray container;  // SArray<tmq_topic_vgroup_t*>
-};
-
 struct tmq_conf_t {
   char     clientId[256];
   char     groupId[TSDB_CGROUP_LEN];
@@ -163,7 +155,7 @@ typedef struct {
 typedef struct {
   tmq_t*  tmq;
   tsem_t  rspSem;
-  tmq_resp_err_t rspErr;
+  int32_t rspErr;
 } SMqSubscribeCbParam;
 
 typedef struct {
@@ -189,7 +181,7 @@ typedef struct {
   int8_t         freeOffsets;
   tmq_commit_cb* userCb;
   tsem_t         rspSem;
-  tmq_resp_err_t rspErr;
+  int32_t        rspErr;
   SArray*        offsets;
   void*          userParam;
 } SMqCommitCbParam;
@@ -201,10 +193,10 @@ typedef struct {
   int8_t         freeOffsets;
   int32_t        waitingRspNum;
   int32_t        totalRspNum;
-  tmq_resp_err_t rspErr;
+  int32_t        rspErr;
   tmq_commit_cb* userCb;
-  SArray*        successfulOffsets;
+  /*SArray*        successfulOffsets;*/
-  SArray*        failedOffsets;
+  /*SArray*        failedOffsets;*/
   void*  userParam;
   tsem_t rspSem;
 } SMqCommitCbParamSet;
@@ -347,10 +339,9 @@ int32_t tmqCommitCb(void* param, const SDataBuf* pMsg, int32_t code) {
   pParam->rspErr = code;
   if (pParam->async) {
     if (pParam->automatic && pParam->tmq->commitCb) {
-      pParam->tmq->commitCb(pParam->tmq, pParam->rspErr, (tmq_topic_vgroup_list_t*)pParam->offsets,
+      pParam->tmq->commitCb(pParam->tmq, pParam->rspErr, pParam->tmq->commitCbUserParam);
-                            pParam->tmq->commitCbUserParam);
     } else if (!pParam->automatic && pParam->userCb) {
-      pParam->userCb(pParam->tmq, pParam->rspErr, (tmq_topic_vgroup_list_t*)pParam->offsets, pParam->userParam);
+      pParam->userCb(pParam->tmq, pParam->rspErr, pParam->userParam);
     }
 
     if (pParam->freeOffsets) {
@@ -368,11 +359,16 @@ int32_t tmqCommitCb2(void* param, const SDataBuf* pBuf, int32_t code) {
   SMqCommitCbParam2*   pParam = (SMqCommitCbParam2*)param;
   SMqCommitCbParamSet* pParamSet = (SMqCommitCbParamSet*)pParam->params;
   // push into array
+#if 0
   if (code == 0) {
     taosArrayPush(pParamSet->failedOffsets, &pParam->pOffset);
   } else {
     taosArrayPush(pParamSet->successfulOffsets, &pParam->pOffset);
   }
+#endif
+
+  /*tscDebug("receive offset commit cb of %s on vg %d, offset is %ld", pParam->pOffset->subKey, pParam->->vgId,
+   * pOffset->version);*/
 
   // count down waiting rsp
   int32_t waitingRspNum = atomic_sub_fetch_32(&pParamSet->waitingRspNum, 1);
@@ -383,23 +379,33 @@ int32_t tmqCommitCb2(void* param, const SDataBuf* pBuf, int32_t code) {
     if (pParamSet->async) {
       // call async cb func
       if (pParamSet->automatic && pParamSet->tmq->commitCb) {
-        pParamSet->tmq->commitCb(pParamSet->tmq, pParamSet->rspErr, NULL, pParamSet->tmq->commitCbUserParam);
+        pParamSet->tmq->commitCb(pParamSet->tmq, pParamSet->rspErr, pParamSet->tmq->commitCbUserParam);
       } else if (!pParamSet->automatic && pParamSet->userCb) {
         // sem post
-        pParamSet->userCb(pParamSet->tmq, pParamSet->rspErr, NULL, pParamSet->userParam);
+        pParamSet->userCb(pParamSet->tmq, pParamSet->rspErr, pParamSet->userParam);
       }
+    } else {
+      tsem_post(&pParamSet->rspSem);
     }
 
+#if 0
     taosArrayDestroyP(pParamSet->successfulOffsets, taosMemoryFree);
     taosArrayDestroyP(pParamSet->failedOffsets, taosMemoryFree);
+#endif
   }
   return 0;
 }
 
-int32_t tmqCommitInner2(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8_t automatic, int8_t async,
+int32_t tmqCommitInner2(tmq_t* tmq, const TAOS_RES* msg, int8_t automatic, int8_t async, tmq_commit_cb* userCb,
-                        tmq_commit_cb* userCb, void* userParam) {
+                        void* userParam) {
   int32_t code = -1;
 
+  if (msg != NULL) {
+    SMqRspObj* pRspObj = (SMqRspObj*)msg;
+    if (!TD_RES_TMQ(pRspObj)) {
+      return TSDB_CODE_TMQ_INVALID_MSG;
+    }
+
     SMqCommitCbParamSet* pParamSet = taosMemoryCalloc(1, sizeof(SMqCommitCbParamSet));
     if (pParamSet == NULL) {
       terrno = TSDB_CODE_OUT_OF_MEMORY;
@@ -415,17 +421,30 @@ int32_t tmqCommitInner2(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8
 
     for (int32_t i = 0; i < taosArrayGetSize(tmq->clientTopics); i++) {
       SMqClientTopic* pTopic = taosArrayGet(tmq->clientTopics, i);
+      if (strcmp(pTopic->topicName, pRspObj->topic) == 0) {
         for (int32_t j = 0; j < taosArrayGetSize(pTopic->vgs); j++) {
-      SMqClientVg* pVg = taosArrayGet(pTopic->vgs, i);
+          SMqClientVg* pVg = taosArrayGet(pTopic->vgs, j);
+          if (pVg->vgId == pRspObj->vgId) {
+            if (pVg->currentOffset < 0 || pVg->committedOffset == pVg->currentOffset) {
+              tscDebug("consumer %ld skip commit for topic %s vg %d, current offset is %ld, committed offset is %ld",
+                       tmq->consumerId, pTopic->topicName, pVg->vgId, pVg->currentOffset, pVg->committedOffset);
+
+              return 0;
+            }
+
             STqOffset* pOffset = taosMemoryCalloc(1, sizeof(STqOffset));
             if (pOffset == NULL) {
               terrno = TSDB_CODE_OUT_OF_MEMORY;
               return -1;
             }
-      int32_t tlen = strlen(tmq->groupId);
+            pOffset->type = TMQ_OFFSET__LOG;
-      memcpy(pOffset->subKey, tmq->groupId, tlen);
+            pOffset->version = pVg->currentOffset;
-      pOffset->subKey[tlen] = TMQ_SEPARATOR;
+
-      strcpy(pOffset->subKey + tlen + 1, pTopic->topicName);
+            int32_t groupLen = strlen(tmq->groupId);
+            memcpy(pOffset->subKey, tmq->groupId, groupLen);
+            pOffset->subKey[groupLen] = TMQ_SEPARATOR;
+            strcpy(pOffset->subKey + groupLen + 1, pTopic->topicName);
+
             int32_t len;
             int32_t code;
             tEncodeSize(tEncodeSTqOffset, pOffset, len, code);
@@ -454,24 +473,36 @@ int32_t tmqCommitInner2(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8
             }
             pMsgSendInfo->msgInfo = (SDataBuf){
                 .pData = buf,
-          .len = len,
+                .len = sizeof(SMsgHead) + len,
                 .handle = NULL,
             };
 
+            tscDebug("consumer %ld commit offset of %s on vg %d, offset is %ld", tmq->consumerId, pOffset->subKey,
+                     pVg->vgId, pOffset->version);
+
+            // TODO: put into cb
+            pVg->committedOffset = pVg->currentOffset;
+
             pMsgSendInfo->requestId = generateRequestId();
             pMsgSendInfo->requestObjRefId = 0;
             pMsgSendInfo->param = pParam;
             pMsgSendInfo->fp = tmqCommitCb2;
-      pMsgSendInfo->msgType = TDMT_MND_MQ_COMMIT_OFFSET;
+            pMsgSendInfo->msgType = TDMT_VND_MQ_COMMIT_OFFSET;
             // send msg
 
-      SEpSet  epSet = getEpSet_s(&tmq->pTscObj->pAppInfo->mgmtEp);
             int64_t transporterId = 0;
-      asyncSendMsgToServer(tmq->pTscObj->pAppInfo->pTransporter, &epSet, &transporterId, pMsgSendInfo);
+            asyncSendMsgToServer(tmq->pTscObj->pAppInfo->pTransporter, &pVg->epSet, &transporterId, pMsgSendInfo);
             pParamSet->waitingRspNum++;
             pParamSet->totalRspNum++;
           }
         }
+      }
+    }
+    if (pParamSet->totalRspNum == 0) {
+      tsem_destroy(&pParamSet->rspSem);
+      taosMemoryFree(pParamSet);
+      return 0;
+    }
 
     if (!async) {
       tsem_wait(&pParamSet->rspSem);
@@ -483,21 +514,144 @@ int32_t tmqCommitInner2(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8
 
     if (code != 0 && async) {
       if (automatic) {
-      tmq->commitCb(tmq, code, NULL, tmq->commitCbUserParam);
+        tmq->commitCb(tmq, code, tmq->commitCbUserParam);
       } else {
-      userCb(tmq, code, NULL, userParam);
+        userCb(tmq, code, userParam);
+      }
+    }
+    return 0;
+  }
+
+  SMqCommitCbParamSet* pParamSet = taosMemoryCalloc(1, sizeof(SMqCommitCbParamSet));
+  if (pParamSet == NULL) {
+    terrno = TSDB_CODE_OUT_OF_MEMORY;
+    return -1;
+  }
+  pParamSet->tmq = tmq;
+  pParamSet->automatic = automatic;
+  pParamSet->async = async;
+  pParamSet->freeOffsets = 1;
+  pParamSet->userCb = userCb;
+  pParamSet->userParam = userParam;
+  tsem_init(&pParamSet->rspSem, 0, 0);
+
+  for (int32_t i = 0; i < taosArrayGetSize(tmq->clientTopics); i++) {
+    SMqClientTopic* pTopic = taosArrayGet(tmq->clientTopics, i);
+
+    tscDebug("consumer %ld begin commit for topic %s, vgNum %d", tmq->consumerId, pTopic->topicName,
+             (int32_t)taosArrayGetSize(pTopic->vgs));
+
+    for (int32_t j = 0; j < taosArrayGetSize(pTopic->vgs); j++) {
+      SMqClientVg* pVg = taosArrayGet(pTopic->vgs, j);
+
+      tscDebug("consumer %ld begin commit for topic %s, vgId %d", tmq->consumerId, pTopic->topicName, pVg->vgId);
+
+      /*if (pVg->currentOffset < 0) {*/
+      if (pVg->currentOffset < 0 || pVg->committedOffset == pVg->currentOffset) {
+        tscDebug("consumer %ld skip commit for topic %s vg %d, current offset is %ld, committed offset is %ld",
+                 tmq->consumerId, pTopic->topicName, pVg->vgId, pVg->currentOffset, pVg->committedOffset);
+
+        continue;
+      }
+      STqOffset* pOffset = taosMemoryCalloc(1, sizeof(STqOffset));
+      if (pOffset == NULL) {
+        terrno = TSDB_CODE_OUT_OF_MEMORY;
+        return -1;
+      }
+      pOffset->type = TMQ_OFFSET__LOG;
+      pOffset->version = pVg->currentOffset;
+
+      int32_t groupLen = strlen(tmq->groupId);
+      memcpy(pOffset->subKey, tmq->groupId, groupLen);
+      pOffset->subKey[groupLen] = TMQ_SEPARATOR;
+      strcpy(pOffset->subKey + groupLen + 1, pTopic->topicName);
+
+      int32_t len;
+      int32_t code;
+      tEncodeSize(tEncodeSTqOffset, pOffset, len, code);
+      if (code < 0) {
+        ASSERT(0);
+      }
+      void* buf = taosMemoryCalloc(1, sizeof(SMsgHead) + len);
+      ((SMsgHead*)buf)->vgId = htonl(pVg->vgId);
+
+      void* abuf = POINTER_SHIFT(buf, sizeof(SMsgHead));
+
+      SEncoder encoder;
+      tEncoderInit(&encoder, abuf, len);
+      tEncodeSTqOffset(&encoder, pOffset);
+
+      // build param
+      SMqCommitCbParam2* pParam = taosMemoryCalloc(1, sizeof(SMqCommitCbParam2));
+      pParam->params = pParamSet;
+      pParam->pOffset = pOffset;
+
+      // build send info
+      SMsgSendInfo* pMsgSendInfo = taosMemoryCalloc(1, sizeof(SMsgSendInfo));
+      if (pMsgSendInfo == NULL) {
+        // TODO
+        continue;
+      }
+      pMsgSendInfo->msgInfo = (SDataBuf){
+          .pData = buf,
+          .len = sizeof(SMsgHead) + len,
+          .handle = NULL,
+      };
+
+      tscDebug("consumer %ld commit offset of %s on vg %d, offset is %ld", tmq->consumerId, pOffset->subKey, pVg->vgId,
+               pOffset->version);
+
+      // TODO: put into cb
+      pVg->committedOffset = pVg->currentOffset;
+
+      pMsgSendInfo->requestId = generateRequestId();
+      pMsgSendInfo->requestObjRefId = 0;
+      pMsgSendInfo->param = pParam;
+      pMsgSendInfo->fp = tmqCommitCb2;
+      pMsgSendInfo->msgType = TDMT_VND_MQ_COMMIT_OFFSET;
+      // send msg
+
+      int64_t transporterId = 0;
+      asyncSendMsgToServer(tmq->pTscObj->pAppInfo->pTransporter, &pVg->epSet, &transporterId, pMsgSendInfo);
+      pParamSet->waitingRspNum++;
+      pParamSet->totalRspNum++;
     }
   }
 
+  if (pParamSet->totalRspNum == 0) {
+    tsem_destroy(&pParamSet->rspSem);
+    taosMemoryFree(pParamSet);
+    return 0;
+  }
+
+  if (!async) {
+    tsem_wait(&pParamSet->rspSem);
+    code = pParamSet->rspErr;
+    tsem_destroy(&pParamSet->rspSem);
+  } else {
+    code = 0;
+  }
+
+  if (code != 0 && async) {
+    if (automatic) {
+      tmq->commitCb(tmq, code, tmq->commitCbUserParam);
+    } else {
+      userCb(tmq, code, userParam);
+    }
+  }
+
+#if 0
   if (!async) {
     taosArrayDestroyP(pParamSet->successfulOffsets, taosMemoryFree);
     taosArrayDestroyP(pParamSet->failedOffsets, taosMemoryFree);
   }
+#endif
 
   return 0;
 }
 
-int32_t tmqCommitInner(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8_t automatic, int8_t async,
+#if 0
+int32_t tmqCommitInner(tmq_t* tmq, const TAOS_RES* msg, int8_t automatic, int8_t async,
                        tmq_commit_cb* userCb, void* userParam) {
   SMqCMCommitOffsetReq req;
   SArray*              pOffsets = NULL;
@@ -507,7 +661,7 @@ int32_t tmqCommitInner(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8_
   int8_t               freeOffsets;
   int32_t              code = -1;
 
-  if (offsets == NULL) {
+  if (msg == NULL) {
     freeOffsets = 1;
     pOffsets = taosArrayInit(0, sizeof(SMqOffset));
     for (int32_t i = 0; i < taosArrayGetSize(tmq->clientTopics); i++) {
@@ -524,7 +678,7 @@ int32_t tmqCommitInner(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int8_
     }
   } else {
     freeOffsets = 0;
-    pOffsets = (SArray*)&offsets->container;
+    pOffsets = (SArray*)&msg->container;
   }
 
   req.num = (int32_t)pOffsets->size;
@@ -611,6 +765,7 @@ END:
   }
   return code;
 }
+#endif
 
 void tmqAssignDelayedHbTask(void* param, void* tmrId) {
   tmq_t*  tmq = (tmq_t*)param;
@@ -648,7 +803,7 @@ int32_t tmqHandleAllDelayedTask(tmq_t* tmq) {
       tmqAskEp(tmq, true);
       taosTmrReset(tmqAssignDelayedHbTask, 1000, tmq, tmqMgmt.timer, &tmq->hbTimer);
     } else if (*pTaskType == TMQ_DELAYED_TASK__COMMIT) {
-      tmqCommitInner(tmq, NULL, 1, 1, tmq->commitCb, tmq->commitCbUserParam);
+      tmqCommitInner2(tmq, NULL, 1, 1, tmq->commitCb, tmq->commitCbUserParam);
       taosTmrReset(tmqAssignDelayedCommitTask, tmq->autoCommitInterval, tmq, tmqMgmt.timer, &tmq->commitTimer);
     } else if (*pTaskType == TMQ_DELAYED_TASK__REPORT) {
     } else {
@@ -689,7 +844,7 @@ int32_t tmqSubscribeCb(void* param, const SDataBuf* pMsg, int32_t code) {
   return 0;
 }
 
-tmq_resp_err_t tmq_subscription(tmq_t* tmq, tmq_list_t** topics) {
+int32_t tmq_subscription(tmq_t* tmq, tmq_list_t** topics) {
   if (*topics == NULL) {
     *topics = tmq_list_new();
   }
@@ -697,12 +852,12 @@ tmq_resp_err_t tmq_subscription(tmq_t* tmq, tmq_list_t** topics) {
     SMqClientTopic* topic = taosArrayGet(tmq->clientTopics, i);
     tmq_list_append(*topics, strchr(topic->topicName, '.') + 1);
   }
-  return TMQ_RESP_ERR__SUCCESS;
+  return 0;
 }
 
-tmq_resp_err_t tmq_unsubscribe(tmq_t* tmq) {
+int32_t tmq_unsubscribe(tmq_t* tmq) {
   tmq_list_t* lst = tmq_list_new();
-  tmq_resp_err_t rsp = tmq_subscribe(tmq, lst);
+  int32_t     rsp = tmq_subscribe(tmq, lst);
   tmq_list_destroy(lst);
   return rsp;
 }
@@ -818,11 +973,13 @@ FAIL:
   return NULL;
 }
 
-tmq_resp_err_t tmq_commit(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int32_t async) {
+#if 0
-  return tmqCommitInner(tmq, offsets, 0, async, tmq->commitCb, tmq->commitCbUserParam);
+int32_t tmq_commit(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, int32_t async) {
+  return tmqCommitInner2(tmq, offsets, 0, async, tmq->commitCb, tmq->commitCbUserParam);
 }
+#endif
 
-tmq_resp_err_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) {
+int32_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) {
   const SArray*   container = &topic_list->container;
   int32_t         sz = taosArrayGetSize(container);
   void*           buf = NULL;
@@ -862,7 +1019,7 @@ tmq_resp_err_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) {
   if (sendInfo == NULL) goto FAIL;
 
   SMqSubscribeCbParam param = {
-      .rspErr = TMQ_RESP_ERR__SUCCESS,
+      .rspErr = 0,
       .tmq = tmq,
   };
 
@@ -943,6 +1100,19 @@ int32_t tmqPollCb(void* param, const SDataBuf* pMsg, int32_t code) {
   if (code != 0) {
     tscWarn("msg discard from vg %d, epoch %d, code:%x", vgId, epoch, code);
     if (pMsg->pData) taosMemoryFree(pMsg->pData);
+    if (code == TSDB_CODE_TQ_NO_COMMITTED_OFFSET) {
+      SMqPollRspWrapper* pRspWrapper = taosAllocateQitem(sizeof(SMqPollRspWrapper), DEF_QITEM);
+      if (pRspWrapper == NULL) {
+        taosMemoryFree(pMsg->pData);
+        tscWarn("msg discard from vg %d, epoch %d since out of memory", vgId, epoch);
+        goto CREATE_MSG_FAIL;
+      }
+      pRspWrapper->tmqRspType = TMQ_MSG_TYPE__END_RSP;
+      /*pRspWrapper->vgHandle = pVg;*/
+      /*pRspWrapper->topicHandle = pTopic;*/
+      taosWriteQitem(tmq->mqueue, pRspWrapper);
+      tsem_post(&tmq->rspSem);
+    }
     goto CREATE_MSG_FAIL;
   }
 
@@ -1017,14 +1187,13 @@ bool tmqUpdateEp2(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
     if (pTopicCur->vgs) {
       int32_t vgNumCur = taosArrayGetSize(pTopicCur->vgs);
       tscDebug("consumer %ld new vg num: %d", tmq->consumerId, vgNumCur);
-      if (vgNumCur == 0) break;
       for (int32_t j = 0; j < vgNumCur; j++) {
         SMqClientVg* pVgCur = taosArrayGet(pTopicCur->vgs, j);
         sprintf(vgKey, "%s:%d", pTopicCur->topicName, pVgCur->vgId);
-        tscDebug("consumer %ld epoch %d vg %d build %s", tmq->consumerId, epoch, pVgCur->vgId, vgKey);
+        tscDebug("consumer %ld epoch %d vg %d vgKey is %s, offset is %ld", tmq->consumerId, epoch, pVgCur->vgId, vgKey,
+                 pVgCur->currentOffset);
         taosHashPut(pHash, vgKey, strlen(vgKey), &pVgCur->currentOffset, sizeof(int64_t));
       }
-      break;
     }
   }
 
@@ -1032,7 +1201,6 @@ bool tmqUpdateEp2(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
     SMqClientTopic topic = {0};
     SMqSubTopicEp* pTopicEp = taosArrayGet(pRsp->topics, i);
     topic.schema = pTopicEp->schema;
-    taosHashClear(pHash);
     topic.topicName = strdup(pTopicEp->topic);
     tstrncpy(topic.db, pTopicEp->db, TSDB_DB_FNAME_LEN);
 
@@ -1049,7 +1217,8 @@ bool tmqUpdateEp2(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
         offset = *pOffset;
       }
 
-      tscDebug("consumer %ld(epoch %d) offset of vg %d updated to %ld", tmq->consumerId, epoch, pVgEp->vgId, offset);
+      tscDebug("consumer %ld(epoch %d) offset of vg %d updated to %ld, vgKey is %s", tmq->consumerId, epoch,
+               pVgEp->vgId, offset, vgKey);
       SMqClientVg clientVg = {
           .pollCnt = 0,
           .currentOffset = offset,
@@ -1076,6 +1245,7 @@ bool tmqUpdateEp2(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
   return set;
 }
 
+#if 1
 bool tmqUpdateEp(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
   /*printf("call update ep %d\n", epoch);*/
   bool    set = false;
@@ -1160,6 +1330,7 @@ bool tmqUpdateEp(tmq_t* tmq, int32_t epoch, SMqAskEpRsp* pRsp) {
   atomic_store_32(&tmq->epoch, epoch);
   return set;
 }
+#endif
 
 int32_t tmqAskEpCb(void* param, const SDataBuf* pMsg, int32_t code) {
   SMqAskEpCbParam* pParam = (SMqAskEpCbParam*)param;
@@ -1186,7 +1357,7 @@ int32_t tmqAskEpCb(void* param, const SDataBuf* pMsg, int32_t code) {
     tDecodeSMqAskEpRsp(POINTER_SHIFT(pMsg->pData, sizeof(SMqRspHead)), &rsp);
     /*printf("rsp epoch %ld sz %ld\n", rsp.epoch, rsp.topics->size);*/
     /*printf("tmq epoch %ld sz %ld\n", tmq->epoch, tmq->clientTopics->size);*/
-    tmqUpdateEp(tmq, head->epoch, &rsp);
+    tmqUpdateEp2(tmq, head->epoch, &rsp);
     tDeleteSMqAskEpRsp(&rsp);
   } else {
     SMqAskEpRspWrapper* pWrapper = taosAllocateQitem(sizeof(SMqAskEpRspWrapper), DEF_QITEM);
@@ -1283,7 +1454,8 @@ int32_t tmqAskEp(tmq_t* tmq, bool async) {
   return code;
 }
 
-tmq_resp_err_t tmq_seek(tmq_t* tmq, const tmq_topic_vgroup_t* offset) {
+#if 0
+int32_t tmq_seek(tmq_t* tmq, const tmq_topic_vgroup_t* offset) {
   const SMqOffset* pOffset = &offset->offset;
   if (strcmp(pOffset->cgroup, tmq->groupId) != 0) {
     return TMQ_RESP_ERR__FAIL;
@@ -1305,16 +1477,17 @@ tmq_resp_err_t tmq_seek(tmq_t* tmq, const tmq_topic_vgroup_t* offset) {
   }
   return TMQ_RESP_ERR__FAIL;
 }
+#endif
 
 SMqPollReq* tmqBuildConsumeReqImpl(tmq_t* tmq, int64_t timeout, SMqClientTopic* pTopic, SMqClientVg* pVg) {
   int64_t reqOffset;
   if (pVg->currentOffset >= 0) {
     reqOffset = pVg->currentOffset;
   } else {
-    if (tmq->resetOffsetCfg == TMQ_CONF__RESET_OFFSET__NONE) {
+    /*if (tmq->resetOffsetCfg == TMQ_CONF__RESET_OFFSET__NONE) {*/
-      tscError("unable to poll since no committed offset but reset offset is set to none");
+    /*tscError("unable to poll since no committed offset but reset offset is set to none");*/
-      return NULL;
+    /*return NULL;*/
-    }
+    /*}*/
     reqOffset = tmq->resetOffsetCfg;
   }
 
@@ -1326,10 +1499,10 @@ SMqPollReq* tmqBuildConsumeReqImpl(tmq_t* tmq, int64_t timeout, SMqClientTopic*
   /*strcpy(pReq->topic, pTopic->topicName);*/
   /*strcpy(pReq->cgroup, tmq->groupId);*/
 
-  int32_t tlen = strlen(tmq->groupId);
+  int32_t groupLen = strlen(tmq->groupId);
-  memcpy(pReq->subKey, tmq->groupId, tlen);
+  memcpy(pReq->subKey, tmq->groupId, groupLen);
-  pReq->subKey[tlen] = TMQ_SEPARATOR;
+  pReq->subKey[groupLen] = TMQ_SEPARATOR;
-  strcpy(pReq->subKey + tlen + 1, pTopic->topicName);
+  strcpy(pReq->subKey + groupLen + 1, pTopic->topicName);
 
   pReq->withTbName = tmq->withTbName;
   pReq->timeout = timeout;
@@ -1440,7 +1613,7 @@ int32_t tmqHandleNoPollRsp(tmq_t* tmq, SMqRspWrapper* rspWrapper, bool* pReset)
     if (rspWrapper->epoch > atomic_load_32(&tmq->epoch)) {
       SMqAskEpRspWrapper* pEpRspWrapper = (SMqAskEpRspWrapper*)rspWrapper;
       SMqAskEpRsp*        rspMsg = &pEpRspWrapper->msg;
-      tmqUpdateEp(tmq, rspWrapper->epoch, rspMsg);
+      tmqUpdateEp2(tmq, rspWrapper->epoch, rspMsg);
       /*tmqClearUnhandleMsg(tmq);*/
       *pReset = true;
     } else {
@@ -1462,7 +1635,11 @@ SMqRspObj* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) {
       if (rspWrapper == NULL) return NULL;
     }
 
-    if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__POLL_RSP) {
+    if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__END_RSP) {
+      taosFreeQitem(rspWrapper);
+      terrno = TSDB_CODE_TQ_NO_COMMITTED_OFFSET;
+      return NULL;
+    } else if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__POLL_RSP) {
       SMqPollRspWrapper* pollRspWrapper = (SMqPollRspWrapper*)rspWrapper;
       /*atomic_sub_fetch_32(&tmq->readyRequest, 1);*/
       int32_t consumerEpoch = atomic_load_32(&tmq->epoch);
@@ -1523,6 +1700,8 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) {
     rspObj = tmqHandleAllRsp(tmq, timeout, false);
     if (rspObj) {
       return (TAOS_RES*)rspObj;
+    } else if (terrno == TSDB_CODE_TQ_NO_COMMITTED_OFFSET) {
+      return NULL;
     }
     if (timeout != -1) {
       int64_t endTime = taosGetTimestampMs();
@@ -1539,10 +1718,10 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) {
   }
 }
 
-tmq_resp_err_t tmq_consumer_close(tmq_t* tmq) {
+int32_t tmq_consumer_close(tmq_t* tmq) {
   if (tmq->status == TMQ_CONSUMER_STATUS__READY) {
-    tmq_resp_err_t rsp = tmq_commit_sync(tmq, NULL);
+    int32_t rsp = tmq_commit_sync(tmq, NULL);
-    if (rsp != TMQ_RESP_ERR__SUCCESS) {
+    if (rsp != 0) {
       return rsp;
     }
 
@@ -1550,18 +1729,18 @@ tmq_resp_err_t tmq_consumer_close(tmq_t* tmq) {
     rsp = tmq_subscribe(tmq, lst);
     tmq_list_destroy(lst);
 
-    if (rsp != TMQ_RESP_ERR__SUCCESS) {
+    if (rsp != 0) {
       return rsp;
     }
   }
   // TODO: free resources
-  return TMQ_RESP_ERR__SUCCESS;
+  return 0;
 }
 
-const char* tmq_err2str(tmq_resp_err_t err) {
+const char* tmq_err2str(int32_t err) {
-  if (err == TMQ_RESP_ERR__SUCCESS) {
+  if (err == 0) {
     return "success";
-  } else if (err == TMQ_RESP_ERR__FAIL) {
+  } else if (err == -1) {
     return "fail";
   } else {
     return tstrerror(err);
@@ -1607,10 +1786,8 @@ const char* tmq_get_table_name(TAOS_RES* res) {
   return NULL;
 }
 
-void tmq_commit_async(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets, tmq_commit_cb* cb, void* param) {
+void tmq_commit_async(tmq_t* tmq, const TAOS_RES* msg, tmq_commit_cb* cb, void* param) {
-  tmqCommitInner(tmq, offsets, 0, 1, cb, param);
+  tmqCommitInner2(tmq, msg, 0, 1, cb, param);
 }
 
-tmq_resp_err_t tmq_commit_sync(tmq_t* tmq, const tmq_topic_vgroup_list_t* offsets) {
+int32_t tmq_commit_sync(tmq_t* tmq, const TAOS_RES* msg) { return tmqCommitInner2(tmq, msg, 0, 0, NULL, NULL); }
-  return tmqCommitInner(tmq, offsets, 0, 0, NULL, NULL);
-}

source/client/test/clientTests.cpp

@@ -44,6 +44,7 @@ void showDB(TAOS* pConn) {
 }
 
 void fetchCallback(void* param, void* res, int32_t numOfRow) {
+#if 0
   printf("numOfRow = %d \n", numOfRow);
   int         numFields = taos_num_fields(res);
   TAOS_FIELD *fields = taos_fetch_fields(res);
@@ -63,6 +64,13 @@ void fetchCallback(void* param, void* res, int32_t numOfRow) {
 //    taos_close(_taos);
 //    taos_cleanup();
   }
+#endif
+  if (numOfRow == 0) {
+    printf("completed\n");
+    return;
+  }
+
+  taos_fetch_raw_block_a(res, fetchCallback, param);
 }
 
 void queryCallback(void* param, void* res, int32_t code) {
@@ -70,7 +78,7 @@ void queryCallback(void* param, void* res, int32_t code) {
     printf("failed to execute, reason:%s\n", taos_errstr(res));
   }
   printf("start to fetch data\n");
-  taos_fetch_rows_a(res, fetchCallback, param);
+  taos_fetch_raw_block_a(res, fetchCallback, param);
 }
 
 void queryCallback1(void* param, void* res, int32_t code) {
@@ -778,9 +786,9 @@ TEST(testCase, async_api_test) {
   TAOS* pConn = taos_connect("localhost", "root", "taosdata", NULL, 0);
   ASSERT_NE(pConn, nullptr);
 
-  taos_query(pConn, "use nest");
+  taos_query(pConn, "use table_alltype_hyperloglog");
-
+#if 0
-  TAOS_RES* pRes = taos_query(pConn, "select NOW() from (select * from regular_table_2 where tbname in ('regular_table_2_1') and  q_bigint <= 9223372036854775807 and  q_tinyint <= 127 and    q_bool in ( true , false) ) order by ts;");
+  TAOS_RES* pRes = taos_query(pConn, "insert into tu(ts) values('2022-02-27 12:12:61')");
   if (taos_errno(pRes) != 0) {
     printf("failed, reason:%s\n", taos_errstr(pRes));
   }
@@ -802,8 +810,9 @@ TEST(testCase, async_api_test) {
     printf("%s\n", str);
     memset(str, 0, sizeof(str));
   }
+#endif
 
-  taos_query_a(pConn, "alter table test.m1 comment 'abcde' ", queryCallback, pConn);
+  taos_query_a(pConn, "select HYPERLOGLOG(q_ts) from stable_1_2 where ts between 1630000001000 and 1630100001000 interval(19d) Fill(NONE);", queryCallback, pConn);
   getchar();
   taos_close(pConn);
 }

source/common/src/systable.c

@@ -77,7 +77,7 @@ static const SSysDbTableSchema userDBSchema[] = {
     {.name = "ntables", .bytes = 8, .type = TSDB_DATA_TYPE_BIGINT},
     {.name = "replica", .bytes = 1, .type = TSDB_DATA_TYPE_TINYINT},
     {.name = "strict", .bytes = 9 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
-    {.name = "duration", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
+    {.name = "duration", .bytes = 10 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "keep", .bytes = 32 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "buffer", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "pagesize", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
@@ -124,7 +124,7 @@ static const SSysDbTableSchema userStbsSchema[] = {
     {.name = "columns", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "tags", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "last_update", .bytes = 8, .type = TSDB_DATA_TYPE_TIMESTAMP},
-    {.name = "table_comment", .bytes = 1024 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "table_comment", .bytes = TSDB_TB_COMMENT_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
 };
 
 static const SSysDbTableSchema streamSchema[] = {
@@ -148,7 +148,7 @@ static const SSysDbTableSchema userTblsSchema[] = {
     {.name = "uid", .bytes = 8, .type = TSDB_DATA_TYPE_BIGINT},
     {.name = "vgroup_id", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "ttl", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
-    {.name = "table_comment", .bytes = 512 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "table_comment", .bytes = TSDB_TB_COMMENT_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "type", .bytes = 20 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
 };
 
@@ -220,7 +220,8 @@ static const SSysDbTableSchema transSchema[] = {
     {.name = "id", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "create_time", .bytes = 8, .type = TSDB_DATA_TYPE_TIMESTAMP},
     {.name = "stage", .bytes = TSDB_TRANS_STAGE_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
-    {.name = "db", .bytes = SYSTABLE_SCH_DB_NAME_LEN, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "db1", .bytes = SYSTABLE_SCH_DB_NAME_LEN, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "db2", .bytes = SYSTABLE_SCH_DB_NAME_LEN, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "failed_times", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "last_exec_time", .bytes = 8, .type = TSDB_DATA_TYPE_TIMESTAMP},
     {.name = "last_action_info",
@@ -302,7 +303,7 @@ static const SSysDbTableSchema offsetSchema[] = {
 };
 
 static const SSysDbTableSchema querySchema[] = {
-    {.name = "query_id", .bytes = 26 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "query_id", .bytes = TSDB_QUERY_ID_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "req_id", .bytes = 8, .type = TSDB_DATA_TYPE_UBIGINT},
     {.name = "connId", .bytes = 4, .type = TSDB_DATA_TYPE_UINT},
     {.name = "app", .bytes = TSDB_APP_NAME_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},    
@@ -317,6 +318,24 @@ static const SSysDbTableSchema querySchema[] = {
     {.name = "sql", .bytes = TSDB_SHOW_SQL_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
 };
 
+static const SSysDbTableSchema appSchema[] = {
+    {.name = "app_id", .bytes = 8, .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "ip", .bytes = TSDB_IPv4ADDR_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "pid", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
+    {.name = "name", .bytes = TSDB_APP_NAME_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},    
+    {.name = "start_time", .bytes = 8 , .type = TSDB_DATA_TYPE_TIMESTAMP},
+    {.name = "insert_req", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "insert_row", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "insert_time", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "insert_bytes", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "fetch_bytes", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "query_time", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "show_query", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "total_req", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "current_req", .bytes = 8 , .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "last_access", .bytes = 8, .type = TSDB_DATA_TYPE_TIMESTAMP},
+};
+
 static const SSysTableMeta perfsMeta[] = {
     {TSDB_PERFS_TABLE_CONNECTIONS, connectionsSchema, tListLen(connectionsSchema)},
     {TSDB_PERFS_TABLE_QUERIES, querySchema, tListLen(querySchema)},
@@ -327,6 +346,7 @@ static const SSysTableMeta perfsMeta[] = {
     {TSDB_PERFS_TABLE_TRANS, transSchema, tListLen(transSchema)},
     {TSDB_PERFS_TABLE_SMAS, smaSchema, tListLen(smaSchema)},
     {TSDB_PERFS_TABLE_STREAMS, streamSchema, tListLen(streamSchema)},
+    {TSDB_PERFS_TABLE_APPS, appSchema, tListLen(appSchema)}
 };
 
 void getInfosDbMeta(const SSysTableMeta** pInfosTableMeta, size_t* size) {

source/common/src/tdatablock.c

@@ -1226,6 +1226,7 @@ SSDataBlock* createOneDataBlock(const SSDataBlock* pDataBlock, bool copyData) {
     SColumnInfoData  colInfo = {0};
     SColumnInfoData* p = taosArrayGet(pDataBlock->pDataBlock, i);
     colInfo.info = p->info;
+    colInfo.hasNull = true;
     taosArrayPush(pBlock->pDataBlock, &colInfo);
   }
 
@@ -1583,6 +1584,11 @@ int32_t buildSubmitReqFromDataBlock(SSubmitReq** pReq, const SArray* pDataBlocks
     int32_t      rowSize = pDataBlock->info.rowSize;
     int64_t      groupId = pDataBlock->info.groupId;
 
+    if (colNum <= 1) {
+      // invalid if only with TS col
+      continue;
+    }
+
     if (rb.nCols != colNum) {
       tdSRowSetTpInfo(&rb, colNum, pTSchema->flen);
     }
@@ -1679,6 +1685,7 @@ int32_t buildSubmitReqFromDataBlock(SSubmitReq** pReq, const SArray* pDataBlocks
     msgLen += pSubmitBlk->dataLen;
   }
 
+  if (numOfBlks > 0) {
     (*pReq)->length = msgLen;
 
     (*pReq)->header.vgId = htonl(vgId);
@@ -1697,11 +1704,16 @@ int32_t buildSubmitReqFromDataBlock(SSubmitReq** pReq, const SArray* pDataBlocks
       blk->numOfRows = htons(blk->numOfRows);
       blk = (SSubmitBlk*)(blk->data + dataLen);
     }
+  } else {
+    // no valid rows
+    taosMemoryFreeClear(*pReq);
+  }
 
   return TSDB_CODE_SUCCESS;
 }
 
 char* buildCtbNameByGroupId(const char* stbName, uint64_t groupId) {
+  ASSERT(stbName[0] != 0);
   SArray* tags = taosArrayInit(0, sizeof(void*));
   SSmlKv* pTag = taosMemoryCalloc(1, sizeof(SSmlKv));
   pTag->key = "group_id";

source/common/src/tglobal.c

@@ -51,15 +51,16 @@ int32_t tsNumOfShmThreads = 1;
 int32_t tsNumOfRpcThreads = 1;
 int32_t tsNumOfCommitThreads = 2;
 int32_t tsNumOfTaskQueueThreads = 1;
-int32_t tsNumOfMnodeQueryThreads = 1;
+int32_t tsNumOfMnodeQueryThreads = 2;
+int32_t tsNumOfMnodeFetchThreads = 1;
 int32_t tsNumOfMnodeReadThreads = 1;
 int32_t tsNumOfVnodeQueryThreads = 2;
-int32_t tsNumOfVnodeFetchThreads = 2;
+int32_t tsNumOfVnodeFetchThreads = 1;
 int32_t tsNumOfVnodeWriteThreads = 2;
 int32_t tsNumOfVnodeSyncThreads = 2;
 int32_t tsNumOfVnodeMergeThreads = 2;
 int32_t tsNumOfQnodeQueryThreads = 2;
-int32_t tsNumOfQnodeFetchThreads = 2;
+int32_t tsNumOfQnodeFetchThreads = 1;
 int32_t tsNumOfSnodeSharedThreads = 2;
 int32_t tsNumOfSnodeUniqueThreads = 2;
 
@@ -417,8 +418,7 @@ static int32_t taosAddServerCfg(SConfig *pCfg) {
   tsNumOfVnodeQueryThreads = TMAX(tsNumOfVnodeQueryThreads, 1);
   if (cfgAddInt32(pCfg, "numOfVnodeQueryThreads", tsNumOfVnodeQueryThreads, 1, 1024, 0) != 0) return -1;
 
-  tsNumOfVnodeFetchThreads = tsNumOfCores / 2;
+  tsNumOfVnodeFetchThreads = TRANGE(tsNumOfVnodeFetchThreads, 1, 1);
-  tsNumOfVnodeFetchThreads = TRANGE(tsNumOfVnodeFetchThreads, 2, 4);
   if (cfgAddInt32(pCfg, "numOfVnodeFetchThreads", tsNumOfVnodeFetchThreads, 1, 1024, 0) != 0) return -1;
 
   tsNumOfVnodeWriteThreads = tsNumOfCores;
@@ -437,8 +437,7 @@ static int32_t taosAddServerCfg(SConfig *pCfg) {
   tsNumOfQnodeQueryThreads = TMAX(tsNumOfQnodeQueryThreads, 1);
   if (cfgAddInt32(pCfg, "numOfQnodeQueryThreads", tsNumOfQnodeQueryThreads, 1, 1024, 0) != 0) return -1;
 
-  tsNumOfQnodeFetchThreads = tsNumOfCores / 2;
+  tsNumOfQnodeFetchThreads = TRANGE(tsNumOfQnodeFetchThreads, 1, 1);
-  tsNumOfQnodeFetchThreads = TRANGE(tsNumOfQnodeFetchThreads, 2, 4);
   if (cfgAddInt32(pCfg, "numOfQnodeFetchThreads", tsNumOfQnodeFetchThreads, 1, 1024, 0) != 0) return -1;
 
   tsNumOfSnodeSharedThreads = tsNumOfCores / 4;

source/common/src/tmsg.c

@@ -191,13 +191,25 @@ static int32_t tSerializeSClientHbReq(SEncoder *pEncoder, const SClientHbReq *pR
   if (tEncodeSClientHbKey(pEncoder, &pReq->connKey) < 0) return -1;
 
   if (pReq->connKey.connType == CONN_TYPE__QUERY) {
+    if (tEncodeI64(pEncoder, pReq->app.appId) < 0) return -1;
+    if (tEncodeI32(pEncoder, pReq->app.pid) < 0) return -1;
+    if (tEncodeCStr(pEncoder, pReq->app.name) < 0) return -1;
+    if (tEncodeI64(pEncoder, pReq->app.startTime) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.numOfInsertsReq) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.numOfInsertRows) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.insertElapsedTime) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.insertBytes) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.fetchBytes) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.queryElapsedTime) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.numOfSlowQueries) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.totalRequests) < 0) return -1;
+    if (tEncodeU64(pEncoder, pReq->app.summary.currentRequests) < 0) return -1;
+
     int32_t queryNum = 0;
     if (pReq->query) {
       queryNum = 1;
       if (tEncodeI32(pEncoder, queryNum) < 0) return -1;
       if (tEncodeU32(pEncoder, pReq->query->connId) < 0) return -1;
-      if (tEncodeI32(pEncoder, pReq->query->pid) < 0) return -1;
-      if (tEncodeCStr(pEncoder, pReq->query->app) < 0) return -1;
 
       int32_t num = taosArrayGetSize(pReq->query->queryDesc);
       if (tEncodeI32(pEncoder, num) < 0) return -1;
@@ -209,7 +221,6 @@ static int32_t tSerializeSClientHbReq(SEncoder *pEncoder, const SClientHbReq *pR
         if (tEncodeI64(pEncoder, desc->useconds) < 0) return -1;
         if (tEncodeI64(pEncoder, desc->stime) < 0) return -1;
         if (tEncodeI64(pEncoder, desc->reqRid) < 0) return -1;
-        if (tEncodeI32(pEncoder, desc->pid) < 0) return -1;
         if (tEncodeI8(pEncoder, desc->stableQuery) < 0) return -1;
         if (tEncodeCStr(pEncoder, desc->fqdn) < 0) return -1;
         if (tEncodeI32(pEncoder, desc->subPlanNum) < 0) return -1;
@@ -243,14 +254,26 @@ static int32_t tDeserializeSClientHbReq(SDecoder *pDecoder, SClientHbReq *pReq)
   if (tDecodeSClientHbKey(pDecoder, &pReq->connKey) < 0) return -1;
 
   if (pReq->connKey.connType == CONN_TYPE__QUERY) {
+    if (tDecodeI64(pDecoder, &pReq->app.appId) < 0) return -1;
+    if (tDecodeI32(pDecoder, &pReq->app.pid) < 0) return -1;
+    if (tDecodeCStrTo(pDecoder, pReq->app.name) < 0) return -1;
+    if (tDecodeI64(pDecoder, &pReq->app.startTime) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.numOfInsertsReq) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.numOfInsertRows) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.insertElapsedTime) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.insertBytes) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.fetchBytes) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.queryElapsedTime) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.numOfSlowQueries) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.totalRequests) < 0) return -1;
+    if (tDecodeU64(pDecoder, &pReq->app.summary.currentRequests) < 0) return -1;
+
     int32_t queryNum = 0;
     if (tDecodeI32(pDecoder, &queryNum) < 0) return -1;
     if (queryNum) {
       pReq->query = taosMemoryCalloc(1, sizeof(*pReq->query));
       if (NULL == pReq->query) return -1;
       if (tDecodeU32(pDecoder, &pReq->query->connId) < 0) return -1;
-      if (tDecodeI32(pDecoder, &pReq->query->pid) < 0) return -1;
-      if (tDecodeCStrTo(pDecoder, pReq->query->app) < 0) return -1;
 
       int32_t num = 0;
       if (tDecodeI32(pDecoder, &num) < 0) return -1;
@@ -265,8 +288,7 @@ static int32_t tDeserializeSClientHbReq(SDecoder *pDecoder, SClientHbReq *pReq)
           if (tDecodeI64(pDecoder, &desc.useconds) < 0) return -1;
           if (tDecodeI64(pDecoder, &desc.stime) < 0) return -1;
           if (tDecodeI64(pDecoder, &desc.reqRid) < 0) return -1;
-          if (tDecodeI32(pDecoder, &desc.pid) < 0) return -1;
+          if (tDecodeI8(pDecoder, (int8_t *)&desc.stableQuery) < 0) return -1;
-          if (tDecodeI8(pDecoder, (int8_t*)&desc.stableQuery) < 0) return -1;
           if (tDecodeCStrTo(pDecoder, desc.fqdn) < 0) return -1;
           if (tDecodeI32(pDecoder, &desc.subPlanNum) < 0) return -1;
 
@@ -474,8 +496,10 @@ int32_t tSerializeSMCreateStbReq(void *buf, int32_t bufLen, SMCreateStbReq *pReq
   if (tStartEncode(&encoder) < 0) return -1;
   if (tEncodeCStr(&encoder, pReq->name) < 0) return -1;
   if (tEncodeI8(&encoder, pReq->igExists) < 0) return -1;
-  if (tEncodeFloat(&encoder, pReq->xFilesFactor) < 0) return -1;
+  if (tEncodeI64(&encoder, pReq->delay1) < 0) return -1;
-  if (tEncodeI32(&encoder, pReq->delay) < 0) return -1;
+  if (tEncodeI64(&encoder, pReq->delay2) < 0) return -1;
+  if (tEncodeI64(&encoder, pReq->watermark1) < 0) return -1;
+  if (tEncodeI64(&encoder, pReq->watermark2) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->ttl) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->numOfColumns) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->numOfTags) < 0) return -1;
@@ -500,7 +524,7 @@ int32_t tSerializeSMCreateStbReq(void *buf, int32_t bufLen, SMCreateStbReq *pReq
   }
 
   if (pReq->commentLen > 0) {
-    if (tEncodeBinary(&encoder, pReq->comment, pReq->commentLen) < 0) return -1;
+    if (tEncodeCStr(&encoder, pReq->comment) < 0) return -1;
   }
   if (pReq->ast1Len > 0) {
     if (tEncodeBinary(&encoder, pReq->pAst1, pReq->ast1Len) < 0) return -1;
@@ -522,8 +546,10 @@ int32_t tDeserializeSMCreateStbReq(void *buf, int32_t bufLen, SMCreateStbReq *pR
   if (tStartDecode(&decoder) < 0) return -1;
   if (tDecodeCStrTo(&decoder, pReq->name) < 0) return -1;
   if (tDecodeI8(&decoder, &pReq->igExists) < 0) return -1;
-  if (tDecodeFloat(&decoder, &pReq->xFilesFactor) < 0) return -1;
+  if (tDecodeI64(&decoder, &pReq->delay1) < 0) return -1;
-  if (tDecodeI32(&decoder, &pReq->delay) < 0) return -1;
+  if (tDecodeI64(&decoder, &pReq->delay2) < 0) return -1;
+  if (tDecodeI64(&decoder, &pReq->watermark1) < 0) return -1;
+  if (tDecodeI64(&decoder, &pReq->watermark2) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->ttl) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->numOfColumns) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->numOfTags) < 0) return -1;
@@ -563,7 +589,7 @@ int32_t tDeserializeSMCreateStbReq(void *buf, int32_t bufLen, SMCreateStbReq *pR
   }
 
   if (pReq->commentLen > 0) {
-    pReq->comment = taosMemoryMalloc(pReq->commentLen);
+    pReq->comment = taosMemoryMalloc(pReq->commentLen + 1);
     if (pReq->comment == NULL) return -1;
     if (tDecodeCStrTo(&decoder, pReq->comment) < 0) return -1;
   }
@@ -681,7 +707,7 @@ int32_t tDeserializeSMAlterStbReq(void *buf, int32_t bufLen, SMAlterStbReq *pReq
   if (tDecodeI32(&decoder, &pReq->ttl) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->commentLen) < 0) return -1;
   if (pReq->commentLen > 0) {
-    pReq->comment = taosMemoryMalloc(pReq->commentLen);
+    pReq->comment = taosMemoryMalloc(pReq->commentLen + 1);
     if (pReq->comment == NULL) return -1;
     if (tDecodeCStrTo(&decoder, pReq->comment) < 0) return -1;
   }
@@ -3410,7 +3436,7 @@ int32_t tSerializeSKillConnReq(void *buf, int32_t bufLen, SKillConnReq *pReq) {
   tEncoderInit(&encoder, buf, bufLen);
 
   if (tStartEncode(&encoder) < 0) return -1;
-  if (tEncodeI32(&encoder, pReq->connId) < 0) return -1;
+  if (tEncodeU32(&encoder, pReq->connId) < 0) return -1;
   tEndEncode(&encoder);
 
   int32_t tlen = encoder.pos;
@@ -3423,7 +3449,7 @@ int32_t tDeserializeSKillConnReq(void *buf, int32_t bufLen, SKillConnReq *pReq)
   tDecoderInit(&decoder, buf, bufLen);
 
   if (tStartDecode(&decoder) < 0) return -1;
-  if (tDecodeI32(&decoder, &pReq->connId) < 0) return -1;
+  if (tDecodeU32(&decoder, &pReq->connId) < 0) return -1;
   tEndDecode(&decoder);
 
   tDecoderClear(&decoder);
@@ -4348,6 +4374,10 @@ int tEncodeSVCreateTbReq(SEncoder *pCoder, const SVCreateTbReq *pReq) {
   if (tEncodeI64(pCoder, pReq->ctime) < 0) return -1;
   if (tEncodeI32(pCoder, pReq->ttl) < 0) return -1;
   if (tEncodeI8(pCoder, pReq->type) < 0) return -1;
+  if (tEncodeI32(pCoder, pReq->commentLen) < 0) return -1;
+  if (pReq->commentLen > 0) {
+    if (tEncodeCStr(pCoder, pReq->comment) < 0) return -1;
+  }
 
   if (pReq->type == TSDB_CHILD_TABLE) {
     if (tEncodeI64(pCoder, pReq->ctb.suid) < 0) return -1;
@@ -4371,6 +4401,12 @@ int tDecodeSVCreateTbReq(SDecoder *pCoder, SVCreateTbReq *pReq) {
   if (tDecodeI64(pCoder, &pReq->ctime) < 0) return -1;
   if (tDecodeI32(pCoder, &pReq->ttl) < 0) return -1;
   if (tDecodeI8(pCoder, &pReq->type) < 0) return -1;
+  if (tDecodeI32(pCoder, &pReq->commentLen) < 0) return -1;
+  if (pReq->commentLen > 0) {
+    pReq->comment = taosMemoryMalloc(pReq->commentLen + 1);
+    if (pReq->comment == NULL) return -1;
+    if (tDecodeCStrTo(pCoder, pReq->comment) < 0) return -1;
+  }
 
   if (pReq->type == TSDB_CHILD_TABLE) {
     if (tDecodeI64(pCoder, &pReq->ctb.suid) < 0) return -1;
@@ -4724,8 +4760,8 @@ int32_t tEncodeSVAlterTbReq(SEncoder *pEncoder, const SVAlterTbReq *pReq) {
       if (pReq->updateTTL) {
         if (tEncodeI32v(pEncoder, pReq->newTTL) < 0) return -1;
       }
-      if (tEncodeI8(pEncoder, pReq->updateComment) < 0) return -1;
+      if (tEncodeI32v(pEncoder, pReq->newCommentLen) < 0) return -1;
-      if (pReq->updateComment) {
+      if (pReq->newCommentLen > 0) {
         if (tEncodeCStr(pEncoder, pReq->newComment) < 0) return -1;
       }
       break;
@@ -4772,8 +4808,8 @@ int32_t tDecodeSVAlterTbReq(SDecoder *pDecoder, SVAlterTbReq *pReq) {
       if (pReq->updateTTL) {
         if (tDecodeI32v(pDecoder, &pReq->newTTL) < 0) return -1;
       }
-      if (tDecodeI8(pDecoder, &pReq->updateComment) < 0) return -1;
+      if (tDecodeI32v(pDecoder, &pReq->newCommentLen) < 0) return -1;
-      if (pReq->updateComment) {
+      if (pReq->newCommentLen > 0) {
         if (tDecodeCStr(pDecoder, &pReq->newComment) < 0) return -1;
       }
       break;

source/dnode/mgmt/mgmt_dnode/src/dmWorker.c

@@ -101,7 +101,8 @@ void dmStopMonitorThread(SDnodeMgmt *pMgmt) {
 static void dmProcessMgmtQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
   SDnodeMgmt *pMgmt = pInfo->ahandle;
   int32_t     code = -1;
-  dTrace("msg:%p, will be processed in dnode queue, type:%s", pMsg, TMSG_INFO(pMsg->msgType));
+  STraceId *  trace = &pMsg->info.traceId;
+  dGTrace("msg:%p, will be processed in dnode queue, type:%s", pMsg, TMSG_INFO(pMsg->msgType));
 
   switch (pMsg->msgType) {
     case TDMT_DND_CONFIG_DNODE:

source/dnode/mgmt/mgmt_mnode/inc/mmInt.h

@@ -30,6 +30,7 @@ typedef struct SMnodeMgmt {
   const char    *path;
   const char    *name;
   SSingleWorker  queryWorker;
+  SSingleWorker  fetchWorker;
   SSingleWorker  readWorker;
   SSingleWorker  writeWorker;
   SSingleWorker  syncWorker;
@@ -57,6 +58,7 @@ int32_t mmPutMsgToWriteQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
 int32_t mmPutMsgToSyncQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
 int32_t mmPutMsgToReadQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
 int32_t mmPutMsgToQueryQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
+int32_t mmPutMsgToFetchQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
 int32_t mmPutMsgToMonitorQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg);
 int32_t mmPutMsgToQueue(SMnodeMgmt *pMgmt, EQueueType qtype, SRpcMsg *pRpc);
 

source/dnode/mgmt/mgmt_mnode/src/mmHandle.c

@@ -204,8 +204,8 @@ SArray *mmGetMsgHandles() {
 
   if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY, mmPutMsgToQueryQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY_CONTINUE, mmPutMsgToQueryQueue, 1) == NULL) goto _OVER;
-  if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY_HEARTBEAT, mmPutMsgToQueryQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY_HEARTBEAT, mmPutMsgToFetchQueue, 1) == NULL) goto _OVER;
-  if (dmSetMgmtHandle(pArray, TDMT_VND_FETCH, mmPutMsgToQueryQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_VND_FETCH, mmPutMsgToFetchQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_CREATE_STB_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_ALTER_STB_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_STB_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
@@ -213,7 +213,7 @@ SArray *mmGetMsgHandles() {
   if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_SMA_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_MQ_VG_CHANGE_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_MQ_VG_DELETE_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
-  if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_TASK, mmPutMsgToQueryQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_TASK, mmPutMsgToFetchQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DEPLOY_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_DROP_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_ALTER_CONFIG_RSP, mmPutMsgToWriteQueue, 0) == NULL) goto _OVER;

source/dnode/mgmt/mgmt_mnode/src/mmInt.c

@@ -148,6 +148,10 @@ static int32_t mmStart(SMnodeMgmt *pMgmt) {
 
 static void mmStop(SMnodeMgmt *pMgmt) {
   dDebug("mnode-mgmt start to stop");
+  taosThreadRwlockWrlock(&pMgmt->lock);
+  pMgmt->stopped = 1;
+  taosThreadRwlockUnlock(&pMgmt->lock);
+
   mndStop(pMgmt->pMnode);
 }
 

source/dnode/mgmt/mgmt_mnode/src/mmWorker.c

@@ -48,7 +48,8 @@ static inline void mmSendRsp(SRpcMsg *pMsg, int32_t code) {
 static void mmProcessRpcMsg(SQueueInfo *pInfo, SRpcMsg *pMsg) {
   SMnodeMgmt *pMgmt = pInfo->ahandle;
   int32_t     code = -1;
-  dTrace("msg:%p, get from mnode queue", pMsg);
+  STraceId *  trace = &pMsg->info.traceId;
+  dGTrace("msg:%p, get from mnode queue", pMsg);
 
   switch (pMsg->msgType) {
     case TDMT_MON_MM_INFO:
@@ -122,6 +123,13 @@ int32_t mmPutMsgToQueryQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg) {
   return mmPutMsgToWorker(pMgmt, &pMgmt->queryWorker, pMsg);
 }
 
+int32_t mmPutMsgToFetchQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg) {
+  pMsg->info.node = pMgmt->pMnode;
+
+  return mmPutMsgToWorker(pMgmt, &pMgmt->fetchWorker, pMsg);
+}
+
+
 int32_t mmPutMsgToMonitorQueue(SMnodeMgmt *pMgmt, SRpcMsg *pMsg) {
   return mmPutMsgToWorker(pMgmt, &pMgmt->monitorWorker, pMsg);
 }
@@ -135,6 +143,9 @@ int32_t mmPutMsgToQueue(SMnodeMgmt *pMgmt, EQueueType qtype, SRpcMsg *pRpc) {
     case QUERY_QUEUE:
       pWorker = &pMgmt->queryWorker;
       break;
+    case FETCH_QUEUE:
+      pWorker = &pMgmt->fetchWorker;
+      break;
     case READ_QUEUE:
       pWorker = &pMgmt->readWorker;
       break;
@@ -167,6 +178,18 @@ int32_t mmStartWorker(SMnodeMgmt *pMgmt) {
     return -1;
   }
 
+  SSingleWorkerCfg fCfg = {
+      .min = tsNumOfMnodeFetchThreads,
+      .max = tsNumOfMnodeFetchThreads,
+      .name = "mnode-fetch",
+      .fp = (FItem)mmProcessRpcMsg,
+      .param = pMgmt,
+  };
+  if (tSingleWorkerInit(&pMgmt->fetchWorker, &fCfg) != 0) {
+    dError("failed to start mnode-fetch worker since %s", terrstr());
+    return -1;
+  }
+
   SSingleWorkerCfg rCfg = {
       .min = tsNumOfMnodeReadThreads,
       .max = tsNumOfMnodeReadThreads,
@@ -220,13 +243,11 @@ int32_t mmStartWorker(SMnodeMgmt *pMgmt) {
 }
 
 void mmStopWorker(SMnodeMgmt *pMgmt) {
-  taosThreadRwlockWrlock(&pMgmt->lock);
-  pMgmt->stopped = 1;
-  taosThreadRwlockUnlock(&pMgmt->lock);
   while (pMgmt->refCount > 0) taosMsleep(10);
 
   tSingleWorkerCleanup(&pMgmt->monitorWorker);
   tSingleWorkerCleanup(&pMgmt->queryWorker);
+  tSingleWorkerCleanup(&pMgmt->fetchWorker);
   tSingleWorkerCleanup(&pMgmt->readWorker);
   tSingleWorkerCleanup(&pMgmt->writeWorker);
   tSingleWorkerCleanup(&pMgmt->syncWorker);

source/dnode/mgmt/mgmt_snode/src/smHandle.c

@@ -95,9 +95,12 @@ SArray *smGetMsgHandles() {
   if (dmSetMgmtHandle(pArray, TDMT_MON_SM_INFO, smPutNodeMsgToMonitorQueue, 0) == NULL) goto _OVER;
 
   if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DEPLOY, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_DROP, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RUN, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DISPATCH, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DISPATCH_RSP, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RECOVER, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RECOVER_RSP, smPutNodeMsgToMgmtQueue, 1) == NULL) goto _OVER;
 
   code = 0;
 _OVER:

source/dnode/mgmt/mgmt_snode/src/smWorker.c

@@ -55,7 +55,9 @@ static void smProcessUniqueQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t num
     taosGetQitem(qall, (void **)&pMsg);
 
     dTrace("msg:%p, get from snode-unique queue", pMsg);
-    sndProcessUMsg(pMgmt->pSnode, pMsg);
+    if (sndProcessUMsg(pMgmt->pSnode, pMsg) < 0) {
+      ASSERT(0);
+    }
 
     dTrace("msg:%p, is freed", pMsg);
     rpcFreeCont(pMsg->pCont);
@@ -67,7 +69,9 @@ static void smProcessSharedQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
   SSnodeMgmt *pMgmt = pInfo->ahandle;
 
   dTrace("msg:%p, get from snode-shared queue", pMsg);
-  sndProcessSMsg(pMgmt->pSnode, pMsg);
+  if (sndProcessSMsg(pMgmt->pSnode, pMsg) < 0) {
+    ASSERT(0);
+  }
 
   dTrace("msg:%p, is freed", pMsg);
   rpcFreeCont(pMsg->pCont);

source/dnode/mgmt/mgmt_vnode/src/vmHandle.c

@@ -228,7 +228,7 @@ int32_t vmProcessCreateVnodeReq(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) {
     vmReleaseVnode(pMgmt, pVnode);
     terrno = TSDB_CODE_NODE_ALREADY_DEPLOYED;
     code = terrno;
-    goto _OVER;
+    return 0;
   }
 
   snprintf(path, TSDB_FILENAME_LEN, "vnode%svnode%d", TD_DIRSEP, vnodeCfg.vgId);
@@ -334,6 +334,7 @@ SArray *vmGetMsgHandles() {
   if (dmSetMgmtHandle(pArray, TDMT_VND_CANCEL_TASK, vmPutMsgToFetchQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_TASK, vmPutMsgToFetchQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_CREATE_STB, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
+  if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_TTL_TABLE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_ALTER_STB, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_DROP_STB, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
   if (dmSetMgmtHandle(pArray, TDMT_VND_CREATE_TABLE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;

source/dnode/mgmt/mgmt_vnode/src/vmWorker.c

@@ -31,7 +31,8 @@ static void vmProcessMgmtQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
   SVnodeMgmt *pMgmt = pInfo->ahandle;
   int32_t     code = -1;
 
-  dTrace("msg:%p, get from vnode-mgmt queue", pMsg);
+  STraceId *trace = &pMsg->info.traceId;
+  dGTrace("msg:%p, get from vnode-mgmt queue", pMsg);
   switch (pMsg->msgType) {
     case TDMT_MON_VM_INFO:
       code = vmProcessGetMonitorInfoReq(pMgmt, pMsg);
@@ -97,7 +98,7 @@ static void vmProcessFetchQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
 
 static void vmProcessSyncQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
   SVnodeObj *pVnode = pInfo->ahandle;
-  SRpcMsg   *pMsg = NULL;
+  SRpcMsg *  pMsg = NULL;
 
   for (int32_t i = 0; i < numOfMsgs; ++i) {
     if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
@@ -118,7 +119,7 @@ static void vmProcessSyncQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOf
 
 static void vmProcessMergeQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
   SVnodeObj *pVnode = pInfo->ahandle;
-  SRpcMsg   *pMsg = NULL;
+  SRpcMsg *  pMsg = NULL;
 
   for (int32_t i = 0; i < numOfMsgs; ++i) {
     if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
@@ -169,10 +170,6 @@ static int32_t vmPutMsgToQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg, EQueueType qtyp
       dTrace("vgId:%d, msg:%p put into vnode-sync queue", pVnode->vgId, pMsg);
       taosWriteQitem(pVnode->pSyncQ, pMsg);
       break;
-    case MERGE_QUEUE:
-      dTrace("vgId:%d, msg:%p put into vnode-merge queue", pVnode->vgId, pMsg);
-      taosWriteQitem(pVnode->pMergeQ, pMsg);
-      break;
     case APPLY_QUEUE:
       dTrace("vgId:%d, msg:%p put into vnode-apply queue", pVnode->vgId, pMsg);
       taosWriteQitem(pVnode->pApplyQ, pMsg);
@@ -195,8 +192,6 @@ int32_t vmPutMsgToQueryQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMsg
 
 int32_t vmPutMsgToFetchQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMsgToQueue(pMgmt, pMsg, FETCH_QUEUE); }
 
-int32_t vmPutMsgToMergeQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMsgToQueue(pMgmt, pMsg, MERGE_QUEUE); }
-
 int32_t vmPutMsgToMgmtQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) {
   dTrace("msg:%p, put into vnode-mgmt queue", pMsg);
   taosWriteQitem(pMgmt->mgmtWorker.queue, pMsg);
@@ -242,9 +237,6 @@ int32_t vmGetQueueSize(SVnodeMgmt *pMgmt, int32_t vgId, EQueueType qtype) {
       case FETCH_QUEUE:
         size = taosQueueItemSize(pVnode->pFetchQ);
         break;
-      case MERGE_QUEUE:
-        size = taosQueueItemSize(pVnode->pMergeQ);
-        break;
       default:
         break;
     }

source/dnode/mgmt/node_mgmt/src/dmTransport.c

@@ -21,21 +21,6 @@ static void dmSendRedirectRsp(SRpcMsg *pMsg, const SEpSet *pNewEpSet);
 static void dmSendRsp(SRpcMsg *pMsg);
 static void dmBuildMnodeRedirectRsp(SDnode *pDnode, SRpcMsg *pMsg);
 
-static inline int32_t dmBuildNodeMsg(SRpcMsg *pMsg, SRpcMsg *pRpc) {
-  SRpcConnInfo *pConnInfo = &(pRpc->info.connInfo);
-  // if (IsReq(pRpc)) {
-  //  terrno = TSDB_CODE_MND_NO_USER_FROM_CONN;
-  //  dError("failed to build msg since %s, app:%p handle:%p", terrstr(), pRpc->info.ahandle, pRpc->info.handle);
-  //  return -1;
-  //}
-
-  memcpy(pMsg, pRpc, sizeof(SRpcMsg));
-  memcpy(pMsg->conn.user, pConnInfo->user, TSDB_USER_LEN);
-  pMsg->conn.clientIp = pConnInfo->clientIp;
-  pMsg->conn.clientPort = pConnInfo->clientPort;
-  return 0;
-}
-
 int32_t dmProcessNodeMsg(SMgmtWrapper *pWrapper, SRpcMsg *pMsg) {
   NodeMsgFp msgFp = pWrapper->msgFps[TMSG_INDEX(pMsg->msgType)];
   if (msgFp == NULL) {
@@ -55,7 +40,8 @@ static void dmProcessRpcMsg(SDnode *pDnode, SRpcMsg *pRpc, SEpSet *pEpSet) {
   SMgmtWrapper *pWrapper = NULL;
   SDnodeHandle *pHandle = &pTrans->msgHandles[TMSG_INDEX(pRpc->msgType)];
 
-  dTrace("msg:%s is received, handle:%p len:%d code:0x%x app:%p refId:%" PRId64, TMSG_INFO(pRpc->msgType),
+  STraceId *trace = &pRpc->info.traceId;
+  dGTrace("msg:%s is received, handle:%p len:%d code:0x%x app:%p refId:%" PRId64, TMSG_INFO(pRpc->msgType),
           pRpc->info.handle, pRpc->contLen, pRpc->code, pRpc->info.ahandle, pRpc->info.refId);
 
   if (pRpc->msgType == TDMT_DND_NET_TEST) {
@@ -116,14 +102,10 @@ static void dmProcessRpcMsg(SDnode *pDnode, SRpcMsg *pRpc, SEpSet *pEpSet) {
   }
 
   pMsg = taosAllocateQitem(sizeof(SRpcMsg), RPC_QITEM);
-  if (pMsg == NULL) {
+  if (pMsg == NULL) goto _OVER;
-    goto _OVER;
+  memcpy(pMsg, pRpc, sizeof(SRpcMsg));
-  }
-  dTrace("msg:%p, is created, type:%s", pMsg, TMSG_INFO(pRpc->msgType));
 
-  if (dmBuildNodeMsg(pMsg, pRpc) != 0) {
+  dTrace("msg:%p, is created, type:%s handle:%p", pMsg, TMSG_INFO(pRpc->msgType), pMsg->info.handle);
-    goto _OVER;
-  }
 
   if (InParentProc(pWrapper)) {
     code = dmPutToProcCQueue(&pWrapper->proc, pMsg, DND_FUNC_REQ);

source/dnode/mgmt/node_util/inc/dmUtil.h

@@ -34,13 +34,13 @@
 
 #include "dnode.h"
 #include "mnode.h"
-#include "qnode.h"
 #include "monitor.h"
+#include "qnode.h"
 #include "sync.h"
 #include "wal.h"
 
 #include "libs/function/function.h"
-
+// clang-format off
 #ifdef __cplusplus
 extern "C" {
 #endif
@@ -51,6 +51,7 @@ extern "C" {
 #define dInfo(...)  { if (dDebugFlag & DEBUG_INFO)  { taosPrintLog("DND ", DEBUG_INFO, 255, __VA_ARGS__); }}
 #define dDebug(...) { if (dDebugFlag & DEBUG_DEBUG) { taosPrintLog("DND ", DEBUG_DEBUG, dDebugFlag, __VA_ARGS__); }}
 #define dTrace(...) { if (dDebugFlag & DEBUG_TRACE) { taosPrintLog("DND ", DEBUG_TRACE, dDebugFlag, __VA_ARGS__); }}
+#define dGTrace(param, ...) do { char buf[40] = {0}; TRACE_TO_STR(trace, buf); dTrace(param ",GTID: %s", __VA_ARGS__, buf);} while(0)
 
 typedef enum {
   DNODE = 0,
@@ -184,3 +185,4 @@ void    dmSetMnodeEpSet(SDnodeData *pData, SEpSet *pEpSet);
 #endif
 
 #endif /*_TD_DM_INT_H_*/
+// clang-format on

source/dnode/mnode/impl/inc/mndDef.h

@@ -124,7 +124,8 @@ typedef struct {
   int32_t        lastErrorNo;
   tmsg_t         lastMsgType;
   SEpSet         lastEpset;
-  char           dbname[TSDB_DB_FNAME_LEN];
+  char           dbname1[TSDB_DB_FNAME_LEN];
+  char           dbname2[TSDB_DB_FNAME_LEN];
   int32_t        startFunc;
   int32_t        stopFunc;
   int32_t        paramLen;
@@ -540,26 +541,35 @@ typedef struct {
 
 typedef struct {
   char name[TSDB_STREAM_FNAME_LEN];
+  // ctl
+  SRWLatch lock;
+  // create info
+  int64_t createTime;
+  int64_t updateTime;
+  int32_t version;
+  int64_t smaId;  // 0 for unused
+  // info
+  int64_t uid;
+  int8_t  status;
+  // config
+  int8_t  dropPolicy;
+  int8_t  trigger;
+  int64_t triggerParam;
+  int64_t watermark;
+  // source and target
+  int64_t sourceDbUid;
+  int64_t targetDbUid;
   char    sourceDb[TSDB_DB_FNAME_LEN];
   char    targetDb[TSDB_DB_FNAME_LEN];
   char    targetSTbName[TSDB_TABLE_FNAME_LEN];
   int64_t targetStbUid;
-  int64_t        createTime;
-  int64_t        updateTime;
-  int64_t        uid;
-  int64_t        dbUid;
-  int32_t        version;
-  int32_t        vgNum;
-  SRWLatch       lock;
-  int8_t         status;
-  int8_t         createdBy;      // STREAM_CREATED_BY__USER or SMA
   int32_t fixedSinkVgId;  // 0 for shuffle
+  // fixedSinkVg is not applicable for encode and decode
   SVgObj fixedSinkVg;
-  int64_t        smaId;  // 0 for unused
+
-  int8_t         trigger;
+  // transformation
-  int64_t        triggerParam;
-  int64_t        watermark;
   char*          sql;
+  char*          ast;
   char*          physicalPlan;
   SArray*        tasks;  // SArray<SArray<SStreamTask>>
   SSchemaWrapper outputSchema;

source/dnode/mnode/impl/inc/mndInt.h

@@ -40,6 +40,8 @@ extern "C" {
 #define mInfo(...)  { if (mDebugFlag & DEBUG_INFO)  { taosPrintLog("MND ", DEBUG_INFO, 255, __VA_ARGS__); }}
 #define mDebug(...) { if (mDebugFlag & DEBUG_DEBUG) { taosPrintLog("MND ", DEBUG_DEBUG, mDebugFlag, __VA_ARGS__); }}
 #define mTrace(...) { if (mDebugFlag & DEBUG_TRACE) { taosPrintLog("MND ", DEBUG_TRACE, mDebugFlag, __VA_ARGS__); }}
+#define mGTrace(param, ...) do { char buf[40] = {0}; TRACE_TO_STR(trace, buf); mTrace(param ", GTID: %s", __VA_ARGS__, buf);} while(0)
+
 // clang-format on
 
 #define SYSTABLE_SCH_TABLE_NAME_LEN ((TSDB_TABLE_NAME_LEN - 1) + VARSTR_HEADER_SIZE)
@@ -54,7 +56,7 @@ typedef void (*ShowFreeIterFp)(SMnode *pMnode, void *pIter);
 typedef struct SQWorker SQHandle;
 
 typedef struct {
-  const char  *name;
+  const char * name;
   MndInitFp    initFp;
   MndCleanupFp cleanupFp;
 } SMnodeStep;
@@ -63,11 +65,12 @@ typedef struct {
   int64_t        showId;
   ShowRetrieveFp retrieveFps[TSDB_MGMT_TABLE_MAX];
   ShowFreeIterFp freeIterFps[TSDB_MGMT_TABLE_MAX];
-  SCacheObj     *cache;
+  SCacheObj *    cache;
 } SShowMgmt;
 
 typedef struct {
-  SCacheObj *cache;
+  SCacheObj *connCache;
+  SCacheObj *appCache;  
 } SProfileMgmt;
 
 typedef struct {
@@ -99,14 +102,14 @@ typedef struct SMnode {
   bool           stopped;
   bool           restored;
   bool           deploy;
-  char          *path;
+  char *         path;
   int64_t        checkTime;
-  SSdb          *pSdb;
+  SSdb *         pSdb;
-  SArray        *pSteps;
+  SArray *       pSteps;
-  SQHandle      *pQuery;
+  SQHandle *     pQuery;
-  SHashObj      *infosMeta;
+  SHashObj *     infosMeta;
-  SHashObj      *perfsMeta;
+  SHashObj *     perfsMeta;
-  SWal          *pWal;
+  SWal *         pWal;
   SShowMgmt      showMgmt;
   SProfileMgmt   profileMgmt;
   STelemMgmt     telemMgmt;

source/dnode/mnode/impl/inc/mndScheduler.h

@@ -32,6 +32,8 @@ int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream);
 int32_t mndConvertRsmaTask(char** pDst, int32_t* pDstLen, const char* ast, int64_t uid, int8_t triggerType,
                            int64_t watermark, double filesFactor);
 
+int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream);
+
 #ifdef __cplusplus
 }
 #endif

source/dnode/mnode/impl/inc/mndStb.h

@@ -27,8 +27,7 @@ void     mndCleanupStb(SMnode *pMnode);
 SStbObj *mndAcquireStb(SMnode *pMnode, char *stbName);
 void     mndReleaseStb(SMnode *pMnode, SStbObj *pStb);
 SSdbRaw *mndStbActionEncode(SStbObj *pStb);
-int32_t  mndValidateStbInfo(SMnode *pMnode, SSTableVersion *pStbs, int32_t numOfStbs, void **ppRsp,
+int32_t  mndValidateStbInfo(SMnode *pMnode, SSTableVersion *pStbs, int32_t numOfStbs, void **ppRsp, int32_t *pRspLen);
-                            int32_t *pRspLen);
 int32_t  mndGetNumOfStbs(SMnode *pMnode, char *dbName, int32_t *pNumOfStbs);
 
 int32_t mndCheckCreateStbReq(SMCreateStbReq *pCreate);
@@ -36,6 +35,9 @@ SDbObj *mndAcquireDbByStb(SMnode *pMnode, const char *stbName);
 int32_t mndBuildStbFromReq(SMnode *pMnode, SStbObj *pDst, SMCreateStbReq *pCreate, SDbObj *pDb);
 int32_t mndAddStbToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SStbObj *pStb);
 
+void mndExtractDbNameFromStbFullName(const char *stbFullName, char *dst);
+void mndExtractTbNameFromStbFullName(const char *stbFullName, char *dst, int32_t dstSize);
+
 #ifdef __cplusplus
 }
 #endif

source/dnode/mnode/impl/inc/mndStream.h

@@ -31,7 +31,8 @@ void        mndReleaseStream(SMnode *pMnode, SStreamObj *pStream);
 SSdbRaw *mndStreamActionEncode(SStreamObj *pStream);
 SSdbRow *mndStreamActionDecode(SSdbRaw *pRaw);
 
-int32_t mndAddStreamToTrans(SMnode *pMnode, SStreamObj *pStream, const char *ast, STrans *pTrans);
+int32_t mndDropStreamByDb(SMnode *pMnode, STrans *pTrans, SDbObj *pDb);
+int32_t mndPersistStream(SMnode *pMnode, STrans *pTrans, SStreamObj *pStream);
 
 #ifdef __cplusplus
 }

source/dnode/mnode/impl/inc/mndTrans.h

@@ -68,7 +68,7 @@ int32_t mndTransAppendRedoAction(STrans *pTrans, STransAction *pAction);
 int32_t mndTransAppendUndoAction(STrans *pTrans, STransAction *pAction);
 void    mndTransSetRpcRsp(STrans *pTrans, void *pCont, int32_t contLen);
 void    mndTransSetCb(STrans *pTrans, ETrnFunc startFunc, ETrnFunc stopFunc, void *param, int32_t paramLen);
-void    mndTransSetDbName(STrans *pTrans, const char *dbname);
+void    mndTransSetDbName(STrans *pTrans, const char *dbname1, const char *dbname2);
 void    mndTransSetSerial(STrans *pTrans);
 
 int32_t mndTransPrepare(SMnode *pMnode, STrans *pTrans);

source/dnode/mnode/impl/src/mndAuth.c

@@ -56,7 +56,7 @@ static int32_t mndProcessAuthReq(SRpcMsg *pReq) {
   memcpy(authRsp.user, authReq.user, TSDB_USER_LEN);
 
   int32_t code = mndRetriveAuth(pReq->info.node, &authRsp);
-  mTrace("user:%s, auth req received, spi:%d encrypt:%d ruser:%s", pReq->conn.user, authRsp.spi, authRsp.encrypt,
+  mTrace("user:%s, auth req received, spi:%d encrypt:%d ruser:%s", pReq->info.conn.user, authRsp.spi, authRsp.encrypt,
          authRsp.user);
 
   int32_t contLen = tSerializeSAuthReq(NULL, 0, &authRsp);

source/dnode/mnode/impl/src/mndBnode.c

@@ -292,7 +292,7 @@ static int32_t mndProcessCreateBnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_BNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_BNODE) != 0) {
     goto _OVER;
   }
 
@@ -394,7 +394,7 @@ static int32_t mndProcessDropBnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_BNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_BNODE) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndDb.c

@@ -477,7 +477,7 @@ static int32_t mndCreateDb(SMnode *pMnode, SRpcMsg *pReq, SCreateDbReq *pCreate,
 
   mDebug("trans:%d, used to create db:%s", pTrans->id, pCreate->db);
 
-  mndTransSetDbName(pTrans, dbObj.name);
+  mndTransSetDbName(pTrans, dbObj.name, NULL);
   if (mndSetCreateDbRedoLogs(pMnode, pTrans, &dbObj, pVgroups) != 0) goto _OVER;
   if (mndSetCreateDbUndoLogs(pMnode, pTrans, &dbObj, pVgroups) != 0) goto _OVER;
   if (mndSetCreateDbCommitLogs(pMnode, pTrans, &dbObj, pVgroups) != 0) goto _OVER;
@@ -521,12 +521,12 @@ static int32_t mndProcessCreateDbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  pUser = mndAcquireUser(pMnode, pReq->conn.user);
+  pUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pUser == NULL) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_DB, NULL) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_DB, NULL) != 0) {
     goto _OVER;
   }
 
@@ -668,7 +668,7 @@ static int32_t mndAlterDb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pOld, SDbObj *p
   mDebug("trans:%d, used to alter db:%s", pTrans->id, pOld->name);
 
   int32_t code = -1;
-  mndTransSetDbName(pTrans, pOld->name);
+  mndTransSetDbName(pTrans, pOld->name, NULL);
   if (mndSetAlterDbRedoLogs(pMnode, pTrans, pOld, pNew) != 0) goto _OVER;
   if (mndSetAlterDbCommitLogs(pMnode, pTrans, pOld, pNew) != 0) goto _OVER;
   if (mndSetAlterDbRedoActions(pMnode, pTrans, pOld, pNew) != 0) goto _OVER;
@@ -700,7 +700,7 @@ static int32_t mndProcessAlterDbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_ALTER_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_ALTER_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -921,7 +921,7 @@ static int32_t mndDropDb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb) {
   if (pTrans == NULL) goto _OVER;
 
   mDebug("trans:%d, used to drop db:%s", pTrans->id, pDb->name);
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
 
   if (mndSetDropDbRedoLogs(pMnode, pTrans, pDb) != 0) goto _OVER;
   if (mndSetDropDbCommitLogs(pMnode, pTrans, pDb) != 0) goto _OVER;
@@ -980,7 +980,7 @@ static int32_t mndProcessDropDbReq(SRpcMsg *pReq) {
     }
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_DROP_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -1127,7 +1127,7 @@ static int32_t mndProcessUseDbReq(SRpcMsg *pReq) {
 
       mError("db:%s, failed to process use db req since %s", usedbReq.db, terrstr());
     } else {
-      if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_USE_DB, pDb) != 0) {
+      if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_USE_DB, pDb) != 0) {
         goto _OVER;
       }
 
@@ -1252,7 +1252,7 @@ static int32_t mndProcessCompactDbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_COMPACT_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_COMPACT_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -1391,11 +1391,13 @@ static void dumpDbInfoData(SSDataBlock *pBlock, SDbObj *pDb, SShowObj *pShow, in
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, rows, (const char *)strict, false);
 
-    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-    colDataAppend(pColInfo, rows, (const char *)&pDb->cfg.daysPerFile, false);
-
     char    tmp[128] = {0};
     int32_t len = 0;
+    len = sprintf(&tmp[VARSTR_HEADER_SIZE], "%dm", pDb->cfg.daysPerFile);
+    varDataSetLen(tmp, len);
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, rows, (const char *)tmp, false);
+
     if (pDb->cfg.daysToKeep0 > pDb->cfg.daysToKeep1 || pDb->cfg.daysToKeep0 > pDb->cfg.daysToKeep2) {
       len = sprintf(&tmp[VARSTR_HEADER_SIZE], "%dm,%dm,%dm", pDb->cfg.daysToKeep1, pDb->cfg.daysToKeep2,
                     pDb->cfg.daysToKeep0);

source/dnode/mnode/impl/src/mndDef.c

@@ -18,34 +18,35 @@
 #include "mndConsumer.h"
 
 int32_t tEncodeSStreamObj(SEncoder *pEncoder, const SStreamObj *pObj) {
-  int32_t sz = 0;
-  /*int32_t outputNameSz = 0;*/
   if (tEncodeCStr(pEncoder, pObj->name) < 0) return -1;
+
+  if (tEncodeI64(pEncoder, pObj->createTime) < 0) return -1;
+  if (tEncodeI64(pEncoder, pObj->updateTime) < 0) return -1;
+  if (tEncodeI32(pEncoder, pObj->version) < 0) return -1;
+  if (tEncodeI64(pEncoder, pObj->smaId) < 0) return -1;
+
+  if (tEncodeI64(pEncoder, pObj->uid) < 0) return -1;
+  if (tEncodeI8(pEncoder, pObj->status) < 0) return -1;
+
+  if (tEncodeI8(pEncoder, pObj->dropPolicy) < 0) return -1;
+  if (tEncodeI8(pEncoder, pObj->trigger) < 0) return -1;
+  if (tEncodeI64(pEncoder, pObj->triggerParam) < 0) return -1;
+  if (tEncodeI64(pEncoder, pObj->watermark) < 0) return -1;
+
+  if (tEncodeI64(pEncoder, pObj->sourceDbUid) < 0) return -1;
+  if (tEncodeI64(pEncoder, pObj->targetDbUid) < 0) return -1;
   if (tEncodeCStr(pEncoder, pObj->sourceDb) < 0) return -1;
   if (tEncodeCStr(pEncoder, pObj->targetDb) < 0) return -1;
   if (tEncodeCStr(pEncoder, pObj->targetSTbName) < 0) return -1;
   if (tEncodeI64(pEncoder, pObj->targetStbUid) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->createTime) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->updateTime) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->uid) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->dbUid) < 0) return -1;
-  if (tEncodeI32(pEncoder, pObj->version) < 0) return -1;
-  if (tEncodeI8(pEncoder, pObj->status) < 0) return -1;
-  if (tEncodeI8(pEncoder, pObj->createdBy) < 0) return -1;
-  if (tEncodeI8(pEncoder, pObj->trigger) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->triggerParam) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->watermark) < 0) return -1;
   if (tEncodeI32(pEncoder, pObj->fixedSinkVgId) < 0) return -1;
-  if (tEncodeI64(pEncoder, pObj->smaId) < 0) return -1;
-  if (tEncodeCStr(pEncoder, pObj->sql) < 0) return -1;
-  /*if (tEncodeCStr(pEncoder, pObj->logicalPlan) < 0) return -1;*/
-  if (tEncodeCStr(pEncoder, pObj->physicalPlan) < 0) return -1;
-  // TODO encode tasks
-  if (pObj->tasks) {
-    sz = taosArrayGetSize(pObj->tasks);
-  }
-  if (tEncodeI32(pEncoder, sz) < 0) return -1;
 
+  if (tEncodeCStr(pEncoder, pObj->sql) < 0) return -1;
+  if (tEncodeCStr(pEncoder, pObj->ast) < 0) return -1;
+  if (tEncodeCStr(pEncoder, pObj->physicalPlan) < 0) return -1;
+
+  int32_t sz = taosArrayGetSize(pObj->tasks);
+  if (tEncodeI32(pEncoder, sz) < 0) return -1;
   for (int32_t i = 0; i < sz; i++) {
     SArray *pArray = taosArrayGetP(pObj->tasks, i);
     int32_t innerSz = taosArrayGetSize(pArray);
@@ -58,40 +59,37 @@ int32_t tEncodeSStreamObj(SEncoder *pEncoder, const SStreamObj *pObj) {
 
   if (tEncodeSSchemaWrapper(pEncoder, &pObj->outputSchema) < 0) return -1;
 
-#if 0
-  if (pObj->ColAlias != NULL) {
-    outputNameSz = taosArrayGetSize(pObj->ColAlias);
-  }
-  if (tEncodeI32(pEncoder, outputNameSz) < 0) return -1;
-  for (int32_t i = 0; i < outputNameSz; i++) {
-    char *name = taosArrayGetP(pObj->ColAlias, i);
-    if (tEncodeCStr(pEncoder, name) < 0) return -1;
-  }
-#endif
   return pEncoder->pos;
 }
 
 int32_t tDecodeSStreamObj(SDecoder *pDecoder, SStreamObj *pObj) {
   if (tDecodeCStrTo(pDecoder, pObj->name) < 0) return -1;
+
+  if (tDecodeI64(pDecoder, &pObj->createTime) < 0) return -1;
+  if (tDecodeI64(pDecoder, &pObj->updateTime) < 0) return -1;
+  if (tDecodeI32(pDecoder, &pObj->version) < 0) return -1;
+  if (tDecodeI64(pDecoder, &pObj->smaId) < 0) return -1;
+
+  if (tDecodeI64(pDecoder, &pObj->uid) < 0) return -1;
+  if (tDecodeI8(pDecoder, &pObj->status) < 0) return -1;
+
+  if (tDecodeI8(pDecoder, &pObj->dropPolicy) < 0) return -1;
+  if (tDecodeI8(pDecoder, &pObj->trigger) < 0) return -1;
+  if (tDecodeI64(pDecoder, &pObj->triggerParam) < 0) return -1;
+  if (tDecodeI64(pDecoder, &pObj->watermark) < 0) return -1;
+
+  if (tDecodeI64(pDecoder, &pObj->sourceDbUid) < 0) return -1;
+  if (tDecodeI64(pDecoder, &pObj->targetDbUid) < 0) return -1;
   if (tDecodeCStrTo(pDecoder, pObj->sourceDb) < 0) return -1;
   if (tDecodeCStrTo(pDecoder, pObj->targetDb) < 0) return -1;
   if (tDecodeCStrTo(pDecoder, pObj->targetSTbName) < 0) return -1;
   if (tDecodeI64(pDecoder, &pObj->targetStbUid) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->createTime) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->updateTime) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->uid) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->dbUid) < 0) return -1;
-  if (tDecodeI32(pDecoder, &pObj->version) < 0) return -1;
-  if (tDecodeI8(pDecoder, &pObj->status) < 0) return -1;
-  if (tDecodeI8(pDecoder, &pObj->createdBy) < 0) return -1;
-  if (tDecodeI8(pDecoder, &pObj->trigger) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->triggerParam) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->watermark) < 0) return -1;
   if (tDecodeI32(pDecoder, &pObj->fixedSinkVgId) < 0) return -1;
-  if (tDecodeI64(pDecoder, &pObj->smaId) < 0) return -1;
+
   if (tDecodeCStrAlloc(pDecoder, &pObj->sql) < 0) return -1;
-  /*if (tDecodeCStrAlloc(pDecoder, &pObj->logicalPlan) < 0) return -1;*/
+  if (tDecodeCStrAlloc(pDecoder, &pObj->ast) < 0) return -1;
   if (tDecodeCStrAlloc(pDecoder, &pObj->physicalPlan) < 0) return -1;
+
   pObj->tasks = NULL;
   int32_t sz;
   if (tDecodeI32(pDecoder, &sz) < 0) return -1;
@@ -112,21 +110,7 @@ int32_t tDecodeSStreamObj(SDecoder *pDecoder, SStreamObj *pObj) {
   }
 
   if (tDecodeSSchemaWrapper(pDecoder, &pObj->outputSchema) < 0) return -1;
-#if 0
+
-  int32_t outputNameSz;
-  if (tDecodeI32(pDecoder, &outputNameSz) < 0) return -1;
-  if (outputNameSz != 0) {
-    pObj->ColAlias = taosArrayInit(outputNameSz, sizeof(void *));
-    if (pObj->ColAlias == NULL) {
-      return -1;
-    }
-  }
-  for (int32_t i = 0; i < outputNameSz; i++) {
-    char *name;
-    if (tDecodeCStrAlloc(pDecoder, &name) < 0) return -1;
-    taosArrayPush(pObj->ColAlias, &name);
-  }
-#endif
   return 0;
 }
 

source/dnode/mnode/impl/src/mndDnode.c

@@ -370,6 +370,7 @@ static int32_t mndProcessStatusReq(SRpcMsg *pReq) {
   SMnodeObj *pObj = mndAcquireMnode(pMnode, pDnode->id);
   if (pObj != NULL) {
     if (pObj->state != statusReq.mload.syncState) {
+      mInfo("dnode:%d, mnode syncstate from %s to %s", pObj->id, syncStr(pObj->state), syncStr(statusReq.mload.syncState));
       pObj->state = statusReq.mload.syncState;
       pObj->stateStartTime = taosGetTimestampMs();
     }
@@ -521,7 +522,7 @@ static int32_t mndProcessCreateDnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_DNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_DNODE) != 0) {
     goto _OVER;
   }
 
@@ -622,7 +623,7 @@ static int32_t mndProcessDropDnodeReq(SRpcMsg *pReq) {
     }
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_MNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_MNODE) != 0) {
     goto _OVER;
   }
 
@@ -760,11 +761,12 @@ static int32_t mndRetrieveDnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, (const char *)&pDnode->createdTime, false);
 
-    char b[tListLen(offlineReason) + VARSTR_HEADER_SIZE] = {0};
+    char *b = taosMemoryCalloc(VARSTR_HEADER_SIZE + strlen(offlineReason[pDnode->offlineReason]) + 1, 1);
     STR_TO_VARSTR(b, online ? "" : offlineReason[pDnode->offlineReason]);
 
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, b, false);
+    taosMemoryFreeClear(b);
 
     numOfRows++;
     sdbRelease(pSdb, pDnode);

source/dnode/mnode/impl/src/mndFunc.c

@@ -318,7 +318,7 @@ static int32_t mndProcessCreateFuncReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_FUNC) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_FUNC) != 0) {
     goto _OVER;
   }
 
@@ -365,7 +365,7 @@ static int32_t mndProcessDropFuncReq(SRpcMsg *pReq) {
     }
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_FUNC) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_FUNC) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndMain.c

@@ -58,31 +58,69 @@ static void *mndBuildTimerMsg(int32_t *pContLen) {
 
 static void mndPullupTrans(SMnode *pMnode) {
   int32_t contLen = 0;
-  void   *pReq = mndBuildTimerMsg(&contLen);
+  void *  pReq = mndBuildTimerMsg(&contLen);
   SRpcMsg rpcMsg = {.msgType = TDMT_MND_TRANS_TIMER, .pCont = pReq, .contLen = contLen};
   tmsgPutToQueue(&pMnode->msgCb, WRITE_QUEUE, &rpcMsg);
 }
 
 static void mndCalMqRebalance(SMnode *pMnode) {
   int32_t contLen = 0;
-  void   *pReq = mndBuildTimerMsg(&contLen);
+  void *  pReq = mndBuildTimerMsg(&contLen);
   SRpcMsg rpcMsg = {.msgType = TDMT_MND_MQ_TIMER, .pCont = pReq, .contLen = contLen};
   tmsgPutToQueue(&pMnode->msgCb, READ_QUEUE, &rpcMsg);
 }
 
 static void mndPullupTelem(SMnode *pMnode) {
   int32_t contLen = 0;
-  void   *pReq = mndBuildTimerMsg(&contLen);
+  void *  pReq = mndBuildTimerMsg(&contLen);
   SRpcMsg rpcMsg = {.msgType = TDMT_MND_TELEM_TIMER, .pCont = pReq, .contLen = contLen};
   tmsgPutToQueue(&pMnode->msgCb, READ_QUEUE, &rpcMsg);
 }
 
+static void mndPushTtlTime(SMnode *pMnode) {
+  SSdb   *pSdb = pMnode->pSdb;
+  SVgObj *pVgroup = NULL;
+  void   *pIter = NULL;
+
+  while (1) {
+    pIter = sdbFetch(pSdb, SDB_VGROUP, pIter, (void **)&pVgroup);
+    if (pIter == NULL) break;
+
+    int32_t contLen = sizeof(SMsgHead) + sizeof(int32_t);
+    SMsgHead   *pHead = rpcMallocCont(contLen);
+    if (pHead == NULL) {
+      mError("ttl time malloc err. contLen:%d", contLen);
+      sdbRelease(pSdb, pVgroup);
+      continue;
+    }
+    pHead->contLen = htonl(contLen);
+    pHead->vgId = htonl(pVgroup->vgId);
+
+    int32_t t = taosGetTimestampSec();
+    *(int32_t*)(POINTER_SHIFT(pHead, sizeof(SMsgHead))) = htonl(t);
+
+    SRpcMsg rpcMsg = {.msgType = TDMT_VND_DROP_TTL_TABLE, .pCont = pHead, .contLen = contLen};
+
+    SEpSet epSet = mndGetVgroupEpset(pMnode, pVgroup);
+    int32_t code = tmsgSendReq(&epSet, &rpcMsg);
+    if(code != 0){
+      mError("ttl time seed err. code:%d", code);
+    }
+    mError("ttl time seed succ. time:%d", t);
+    sdbRelease(pSdb, pVgroup);
+  }
+}
+
 static void *mndThreadFp(void *param) {
   SMnode *pMnode = param;
   int64_t lastTime = 0;
   setThreadName("mnode-timer");
 
   while (1) {
+    if (lastTime % (864000) == 0) {   // sleep 1 day for ttl
+      mndPushTtlTime(pMnode);
+    }
+
     lastTime++;
     taosMsleep(100);
     if (mndGetStop(pMnode)) break;
@@ -378,7 +416,7 @@ void mndStop(SMnode *pMnode) {
 }
 
 int32_t mndProcessSyncMsg(SRpcMsg *pMsg) {
-  SMnode    *pMnode = pMsg->info.node;
+  SMnode *   pMnode = pMsg->info.node;
   SSyncMgmt *pMgmt = &pMnode->syncMgmt;
   int32_t    code = 0;
 
@@ -397,7 +435,7 @@ int32_t mndProcessSyncMsg(SRpcMsg *pMsg) {
 
   char  logBuf[512] = {0};
   char *syncNodeStr = sync2SimpleStr(pMgmt->sync);
-  snprintf(logBuf, sizeof(logBuf), "==vnodeProcessSyncReq== msgType:%d, syncNode: %s", pMsg->msgType, syncNodeStr);
+  snprintf(logBuf, sizeof(logBuf), "==mndProcessSyncMsg== msgType:%d, syncNode: %s", pMsg->msgType, syncNodeStr);
   static int64_t mndTick = 0;
   if (++mndTick % 10 == 1) {
     mTrace("sync trace msg:%s, %s", TMSG_INFO(pMsg->msgType), syncNodeStr);
@@ -505,12 +543,14 @@ int32_t mndProcessSyncMsg(SRpcMsg *pMsg) {
 }
 
 static int32_t mndCheckMnodeState(SRpcMsg *pMsg) {
+  if (!IsReq(pMsg)) return 0;
   if (mndAcquireRpcRef(pMsg->info.node) == 0) return 0;
 
   if (IsReq(pMsg) && pMsg->msgType != TDMT_MND_MQ_TIMER && pMsg->msgType != TDMT_MND_TELEM_TIMER &&
       pMsg->msgType != TDMT_MND_TRANS_TIMER) {
-    mError("msg:%p, failed to check mnode state since %s, app:%p type:%s", pMsg, terrstr(), pMsg->info.ahandle,
+    mError("msg:%p, failed to check mnode state since %s, type:%s", pMsg, terrstr(), TMSG_INFO(pMsg->msgType));
-           TMSG_INFO(pMsg->msgType));
+
+    mndAbortPreprocessMsg(pMsg);
 
     SEpSet epSet = {0};
     mndGetMnodeEpSet(pMsg->info.node, &epSet);
@@ -533,13 +573,14 @@ static int32_t mndCheckMsgContent(SRpcMsg *pMsg) {
   if (!IsReq(pMsg)) return 0;
   if (pMsg->contLen != 0 && pMsg->pCont != NULL) return 0;
 
-  mError("msg:%p, failed to check msg content, app:%p type:%s", pMsg, pMsg->info.ahandle, TMSG_INFO(pMsg->msgType));
+  mError("msg:%p, failed to check msg, cont:%p contLen:%d, app:%p type:%s", pMsg, pMsg->pCont, pMsg->contLen,
+         pMsg->info.ahandle, TMSG_INFO(pMsg->msgType));
   terrno = TSDB_CODE_INVALID_MSG_LEN;
   return -1;
 }
 
 int32_t mndProcessRpcMsg(SRpcMsg *pMsg) {
-  SMnode  *pMnode = pMsg->info.node;
+  SMnode * pMnode = pMsg->info.node;
   MndMsgFp fp = pMnode->msgFp[TMSG_INDEX(pMsg->msgType)];
   if (fp == NULL) {
     mError("msg:%p, failed to get msg handle, app:%p type:%s", pMsg, pMsg->info.ahandle, TMSG_INFO(pMsg->msgType));
@@ -550,14 +591,15 @@ int32_t mndProcessRpcMsg(SRpcMsg *pMsg) {
   if (mndCheckMsgContent(pMsg) != 0) return -1;
   if (mndCheckMnodeState(pMsg) != 0) return -1;
 
-  mTrace("msg:%p, start to process in mnode, app:%p type:%s", pMsg, pMsg->info.ahandle, TMSG_INFO(pMsg->msgType));
+  STraceId *trace = &pMsg->info.traceId;
+  mGTrace("msg:%p, start to process in mnode, app:%p type:%s", pMsg, pMsg->info.ahandle, TMSG_INFO(pMsg->msgType));
   int32_t code = (*fp)(pMsg);
   mndReleaseRpcRef(pMnode);
 
   if (code == TSDB_CODE_ACTION_IN_PROGRESS) {
     mTrace("msg:%p, won't response immediately since in progress", pMsg);
   } else if (code == 0) {
-    mTrace("msg:%p, successfully processed and response", pMsg);
+    mTrace("msg:%p, successfully processed", pMsg);
   } else {
     mError("msg:%p, failed to process since %s, app:%p type:%s", pMsg, terrstr(), pMsg->info.ahandle,
            TMSG_INFO(pMsg->msgType));
@@ -591,7 +633,7 @@ int32_t mndGetMonitorInfo(SMnode *pMnode, SMonClusterInfo *pClusterInfo, SMonVgr
                           SMonGrantInfo *pGrantInfo) {
   if (mndAcquireRpcRef(pMnode) != 0) return -1;
 
-  SSdb   *pSdb = pMnode->pSdb;
+  SSdb *  pSdb = pMnode->pSdb;
   int64_t ms = taosGetTimestampMs();
 
   pClusterInfo->dnodes = taosArrayInit(sdbGetSize(pSdb, SDB_DNODE), sizeof(SMonDnodeDesc));
@@ -667,7 +709,7 @@ int32_t mndGetMonitorInfo(SMnode *pMnode, SMonClusterInfo *pClusterInfo, SMonVgr
     pGrantInfo->timeseries_used += pVgroup->numOfTimeSeries;
     tstrncpy(desc.status, "unsynced", sizeof(desc.status));
     for (int32_t i = 0; i < pVgroup->replica; ++i) {
-      SVnodeGid     *pVgid = &pVgroup->vnodeGid[i];
+      SVnodeGid *    pVgid = &pVgroup->vnodeGid[i];
       SMonVnodeDesc *pVnDesc = &desc.vnodes[i];
       pVnDesc->dnode_id = pVgid->dnodeId;
       tstrncpy(pVnDesc->vnode_role, syncStr(pVgid->role), sizeof(pVnDesc->vnode_role));
@@ -699,6 +741,7 @@ int32_t mndGetMonitorInfo(SMnode *pMnode, SMonClusterInfo *pClusterInfo, SMonVgr
 
 int32_t mndGetLoad(SMnode *pMnode, SMnodeLoad *pLoad) {
   pLoad->syncState = syncGetMyRole(pMnode->syncMgmt.sync);
+  mTrace("mnode current syncstate is %s", syncStr(pLoad->syncState));
   return 0;
 }
 

source/dnode/mnode/impl/src/mndMnode.c

@@ -414,7 +414,7 @@ static int32_t mndProcessCreateMnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_MNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_MNODE) != 0) {
     goto _OVER;
   }
 
@@ -621,7 +621,7 @@ static int32_t mndProcessDropMnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_MNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_MNODE) != 0) {
     goto _OVER;
   }
 
@@ -667,6 +667,10 @@ static int32_t mndRetrieveMnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
     }
     if (pObj->pDnode && mndIsDnodeOnline(pObj->pDnode, curMs)) {
       roles = syncStr(pObj->state);
+      if (pObj->state == TAOS_SYNC_STATE_LEADER && pObj->id != pMnode->selfDnodeId) {
+        roles = syncStr(TAOS_SYNC_STATE_ERROR);
+        mError("mnode:%d, is leader too", pObj->id);
+      }
     }
     char b2[12 + VARSTR_HEADER_SIZE] = {0};
     STR_WITH_MAXSIZE_TO_VARSTR(b2, roles, pShow->pMeta->pSchemas[cols].bytes);
@@ -738,7 +742,7 @@ static int32_t mndProcessAlterMnodeReq(SRpcMsg *pReq) {
     pMgmt->errCode = 0;
     pMgmt->transId = -1;
     tsem_wait(&pMgmt->syncSem);
-    mInfo("alter mnode sync result:%s", tstrerror(pMgmt->errCode));
+    mInfo("alter mnode sync result:0x%x %s", pMgmt->errCode, tstrerror(pMgmt->errCode));
     terrno = pMgmt->errCode;
     return pMgmt->errCode;
   }

source/dnode/mnode/impl/src/mndProfile.c

@@ -43,6 +43,16 @@ typedef struct {
   SArray  *pQueries;  // SArray<SQueryDesc>
 } SConnObj;
 
+typedef struct {
+  int64_t            appId;
+  uint32_t           ip;
+  int32_t            pid;
+  char               name[TSDB_APP_NAME_LEN];
+  int64_t            startTime;
+  SAppClusterSummary summary;
+  int64_t            lastAccessTimeMs;
+} SAppObj;
+
 static SConnObj *mndCreateConn(SMnode *pMnode, const char *user, int8_t connType, uint32_t ip, uint16_t port,
                                int32_t pid, const char *app, int64_t startTime);
 static void      mndFreeConn(SConnObj *pConn);
@@ -57,14 +67,24 @@ static int32_t   mndProcessKillConnReq(SRpcMsg *pReq);
 static int32_t   mndRetrieveConns(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows);
 static int32_t   mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows);
 static void      mndCancelGetNextQuery(SMnode *pMnode, void *pIter);
+static void      mndFreeApp(SAppObj *pApp);
+static int32_t   mndRetrieveApps(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows);
+static void      mndCancelGetNextApp(SMnode *pMnode, void *pIter);
 
 int32_t mndInitProfile(SMnode *pMnode) {
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
 
   // in ms
-  int32_t connCheckTime = tsShellActivityTimer * 2 * 1000;
+  int32_t checkTime = tsShellActivityTimer * 2 * 1000;
-  pMgmt->cache = taosCacheInit(TSDB_DATA_TYPE_INT, connCheckTime, true, (__cache_free_fn_t)mndFreeConn, "conn");
+  pMgmt->connCache = taosCacheInit(TSDB_DATA_TYPE_UINT, checkTime, true, (__cache_free_fn_t)mndFreeConn, "conn");
-  if (pMgmt->cache == NULL) {
+  if (pMgmt->connCache == NULL) {
+    terrno = TSDB_CODE_OUT_OF_MEMORY;
+    mError("failed to alloc profile cache since %s", terrstr());
+    return -1;
+  }
+
+  pMgmt->appCache = taosCacheInit(TSDB_DATA_TYPE_BIGINT, checkTime, true, (__cache_free_fn_t)mndFreeApp, "app");
+  if (pMgmt->appCache == NULL) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
     mError("failed to alloc profile cache since %s", terrstr());
     return -1;
@@ -79,15 +99,22 @@ int32_t mndInitProfile(SMnode *pMnode) {
   mndAddShowFreeIterHandle(pMnode, TSDB_MGMT_TABLE_CONNS, mndCancelGetNextConn);
   mndAddShowRetrieveHandle(pMnode, TSDB_MGMT_TABLE_QUERIES, mndRetrieveQueries);
   mndAddShowFreeIterHandle(pMnode, TSDB_MGMT_TABLE_QUERIES, mndCancelGetNextQuery);
+  mndAddShowRetrieveHandle(pMnode, TSDB_MGMT_TABLE_APPS, mndRetrieveApps);
+  mndAddShowFreeIterHandle(pMnode, TSDB_MGMT_TABLE_APPS, mndCancelGetNextApp);
 
   return 0;
 }
 
 void mndCleanupProfile(SMnode *pMnode) {
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
-  if (pMgmt->cache != NULL) {
+  if (pMgmt->connCache != NULL) {
-    taosCacheCleanup(pMgmt->cache);
+    taosCacheCleanup(pMgmt->connCache);
-    pMgmt->cache = NULL;
+    pMgmt->connCache = NULL;
+  }
+
+  if (pMgmt->appCache != NULL) {
+    taosCacheCleanup(pMgmt->appCache);
+    pMgmt->appCache = NULL;
   }
 }
 
@@ -97,7 +124,7 @@ static SConnObj *mndCreateConn(SMnode *pMnode, const char *user, int8_t connType
 
   char    connStr[255] = {0};
   int32_t len = snprintf(connStr, sizeof(connStr), "%s%d%d%d%s", user, ip, port, pid, app);
-  int32_t connId = mndGenerateUid(connStr, len);
+  uint32_t connId = mndGenerateUid(connStr, len);
   if (startTime == 0) startTime = taosGetTimestampMs();
 
   SConnObj connObj = {.id = connId,
@@ -118,7 +145,7 @@ static SConnObj *mndCreateConn(SMnode *pMnode, const char *user, int8_t connType
   tstrncpy(connObj.app, app, TSDB_APP_NAME_LEN);
 
   int32_t   keepTime = tsShellActivityTimer * 3;
-  SConnObj *pConn = taosCachePut(pMgmt->cache, &connId, sizeof(int32_t), &connObj, sizeof(connObj), keepTime * 1000);
+  SConnObj *pConn = taosCachePut(pMgmt->connCache, &connId, sizeof(uint32_t), &connObj, sizeof(connObj), keepTime * 1000);
   if (pConn == NULL) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
     mError("conn:%d, failed to put into cache since %s, user:%s", connId, user, terrstr());
@@ -140,14 +167,13 @@ static void mndFreeConn(SConnObj *pConn) {
 static SConnObj *mndAcquireConn(SMnode *pMnode, uint32_t connId) {
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
 
-  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->cache, &connId, sizeof(connId));
+  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->connCache, &connId, sizeof(connId));
   if (pConn == NULL) {
     mDebug("conn:%u, already destroyed", connId);
     return NULL;
   }
 
-  int32_t keepTime = tsShellActivityTimer * 3;
+  pConn->lastAccessTimeMs = taosGetTimestampMs();
-  pConn->lastAccessTimeMs = keepTime * 1000 + (uint64_t)taosGetTimestampMs();
 
   mTrace("conn:%u, acquired from cache, data:%p", pConn->id, pConn);
   return pConn;
@@ -158,7 +184,7 @@ static void mndReleaseConn(SMnode *pMnode, SConnObj *pConn) {
   mTrace("conn:%u, released from cache, data:%p", pConn->id, pConn);
 
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
-  taosCacheRelease(pMgmt->cache, (void **)&pConn, false);
+  taosCacheRelease(pMgmt->connCache, (void **)&pConn, false);
 }
 
 void *mndGetNextConn(SMnode *pMnode, SCacheIter *pIter) {
@@ -194,15 +220,15 @@ static int32_t mndProcessConnectReq(SRpcMsg *pReq) {
     goto CONN_OVER;
   }
 
-  taosIp2String(pReq->conn.clientIp, ip);
+  taosIp2String(pReq->info.conn.clientIp, ip);
 
-  pUser = mndAcquireUser(pMnode, pReq->conn.user);
+  pUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pUser == NULL) {
-    mError("user:%s, failed to login while acquire user since %s", pReq->conn.user, terrstr());
+    mError("user:%s, failed to login while acquire user since %s", pReq->info.conn.user, terrstr());
     goto CONN_OVER;
   }
   if (0 != strncmp(connReq.passwd, pUser->pass, TSDB_PASSWORD_LEN - 1)) {
-    mError("user:%s, failed to auth while acquire user, input:%s", pReq->conn.user, connReq.passwd);
+    mError("user:%s, failed to auth while acquire user, input:%s", pReq->info.conn.user, connReq.passwd);
     code = TSDB_CODE_RPC_AUTH_FAILURE;
     goto CONN_OVER;
   }
@@ -213,15 +239,16 @@ static int32_t mndProcessConnectReq(SRpcMsg *pReq) {
     pDb = mndAcquireDb(pMnode, db);
     if (pDb == NULL) {
       terrno = TSDB_CODE_MND_INVALID_DB;
-      mError("user:%s, failed to login from %s while use db:%s since %s", pReq->conn.user, ip, connReq.db, terrstr());
+      mError("user:%s, failed to login from %s while use db:%s since %s", pReq->info.conn.user, ip, connReq.db,
+             terrstr());
       goto CONN_OVER;
     }
   }
 
-  pConn = mndCreateConn(pMnode, pReq->conn.user, connReq.connType, pReq->conn.clientIp, pReq->conn.clientPort,
+  pConn = mndCreateConn(pMnode, pReq->info.conn.user, connReq.connType, pReq->info.conn.clientIp,
-                        connReq.pid, connReq.app, connReq.startTime);
+                        pReq->info.conn.clientPort, connReq.pid, connReq.app, connReq.startTime);
   if (pConn == NULL) {
-    mError("user:%s, failed to login from %s while create connection since %s", pReq->conn.user, ip, terrstr());
+    mError("user:%s, failed to login from %s while create connection since %s", pReq->info.conn.user, ip, terrstr());
     goto CONN_OVER;
   }
 
@@ -246,7 +273,7 @@ static int32_t mndProcessConnectReq(SRpcMsg *pReq) {
   pReq->info.rspLen = contLen;
   pReq->info.rsp = pRsp;
 
-  mDebug("user:%s, login from %s:%d, conn:%u, app:%s", pReq->conn.user, ip, pConn->port, pConn->id, connReq.app);
+  mDebug("user:%s, login from %s:%d, conn:%u, app:%s", pReq->info.conn.user, ip, pConn->port, pConn->id, connReq.app);
 
   code = 0;
 
@@ -275,6 +302,77 @@ static int32_t mndSaveQueryList(SConnObj *pConn, SQueryHbReqBasic *pBasic) {
   return TSDB_CODE_SUCCESS;
 }
 
+static SAppObj *mndCreateApp(SMnode *pMnode, uint32_t clientIp, SAppHbReq* pReq) {
+  SProfileMgmt *pMgmt = &pMnode->profileMgmt;
+
+  SAppObj app;
+  app.appId = pReq->appId;
+  app.ip = clientIp;
+  app.pid = pReq->pid;
+  strcpy(app.name, pReq->name);
+  app.startTime = pReq->startTime;
+  memcpy(&app.summary, &pReq->summary, sizeof(pReq->summary));
+  app.lastAccessTimeMs = taosGetTimestampMs();
+
+  int32_t   keepTime = tsShellActivityTimer * 3;
+  SAppObj *pApp = taosCachePut(pMgmt->appCache, &pReq->appId, sizeof(pReq->appId), &app, sizeof(app), keepTime * 1000);
+  if (pApp == NULL) {
+    terrno = TSDB_CODE_OUT_OF_MEMORY;
+    mError("failed to app %" PRIx64 " into cache since %s", pReq->appId, terrstr());
+    return NULL;
+  }
+  
+  mTrace("app %" PRIx64 " is put into cache", pReq->appId);
+  return pApp;
+}
+
+static void mndFreeApp(SAppObj *pApp) {
+  mTrace("app %" PRIx64 " is destroyed", pApp->appId);
+}
+
+
+static SAppObj *mndAcquireApp(SMnode *pMnode, int64_t appId) {
+  SProfileMgmt *pMgmt = &pMnode->profileMgmt;
+
+  SAppObj *pApp = taosCacheAcquireByKey(pMgmt->appCache, &appId, sizeof(appId));
+  if (pApp == NULL) {
+    mDebug("app %" PRIx64 " not in cache", appId);
+    return NULL;
+  }
+
+  pApp->lastAccessTimeMs = (uint64_t)taosGetTimestampMs();
+
+  mTrace("app %" PRIx64 " acquired from cache", appId);
+  return pApp;
+}
+
+static void mndReleaseApp(SMnode *pMnode, SAppObj *pApp) {
+  if (pApp == NULL) return;
+  mTrace("release app %" PRIx64 " to cache", pApp->appId);
+
+  SProfileMgmt *pMgmt = &pMnode->profileMgmt;
+  taosCacheRelease(pMgmt->appCache, (void **)&pApp, false);
+}
+
+void *mndGetNextApp(SMnode *pMnode, SCacheIter *pIter) {
+  SAppObj *pApp = NULL;
+  bool      hasNext = taosCacheIterNext(pIter);
+  if (hasNext) {
+    size_t dataLen = 0;
+    pApp = taosCacheIterGetData(pIter, &dataLen);
+  } else {
+    taosCacheDestroyIter(pIter);
+  }
+
+  return pApp;
+}
+
+static void mndCancelGetNextApp(SMnode *pMnode, void *pIter) {
+  if (pIter != NULL) {
+    taosCacheDestroyIter(pIter);
+  }
+}
+
 static SClientHbRsp *mndMqHbBuildRsp(SMnode *pMnode, SClientHbReq *pReq) {
 #if 0
   SClientHbRsp* pRsp = taosMemoryMalloc(sizeof(SClientHbRsp));
@@ -340,25 +438,48 @@ static SClientHbRsp *mndMqHbBuildRsp(SMnode *pMnode, SClientHbReq *pReq) {
   return NULL;
 }
 
+static int32_t mndUpdateAppInfo(SMnode *pMnode, SClientHbReq *pHbReq, SRpcConnInfo *connInfo) {
+  SAppHbReq* pReq = &pHbReq->app;
+  SAppObj *pApp = mndAcquireApp(pMnode, pReq->appId);
+  if (pApp == NULL) {
+    pApp = mndCreateApp(pMnode, connInfo->clientIp, pReq);
+    if (pApp == NULL) {
+      mError("failed to create new app %" PRIx64 " since %s", pReq->appId, terrstr());
+      return -1;
+    } else {
+      mDebug("a new app %" PRIx64 "created", pReq->appId);
+      mndReleaseApp(pMnode, pApp);      
+      return TSDB_CODE_SUCCESS;
+    }
+  }
+
+  memcpy(&pApp->summary, &pReq->summary, sizeof(pReq->summary));
+
+  mndReleaseApp(pMnode, pApp);
+
+  return TSDB_CODE_SUCCESS;
+}
+
 static int32_t mndProcessQueryHeartBeat(SMnode *pMnode, SRpcMsg *pMsg, SClientHbReq *pHbReq,
                                         SClientHbBatchRsp *pBatchRsp) {
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
   SClientHbRsp  hbRsp = {.connKey = pHbReq->connKey, .status = 0, .info = NULL, .query = NULL};
+  SRpcConnInfo connInfo = pMsg->info.conn;
+
+  mndUpdateAppInfo(pMnode, pHbReq, &connInfo);
 
   if (pHbReq->query) {
     SQueryHbReqBasic *pBasic = pHbReq->query;
 
-    SRpcConnInfo connInfo = pMsg->conn;
-
     SConnObj *pConn = mndAcquireConn(pMnode, pBasic->connId);
     if (pConn == NULL) {
       pConn = mndCreateConn(pMnode, connInfo.user, CONN_TYPE__QUERY, connInfo.clientIp, connInfo.clientPort,
-                            pBasic->pid, pBasic->app, 0);
+                            pHbReq->app.pid, pHbReq->app.name, 0);
       if (pConn == NULL) {
         mError("user:%s, conn:%u is freed and failed to create new since %s", connInfo.user, pBasic->connId, terrstr());
         return -1;
       } else {
-        mDebug("user:%s, conn:%u is freed and create a new conn:%u", connInfo.user, pBasic->connId, pConn->id);
+        mDebug("user:%s, conn:%u is freed, will create a new conn:%u", connInfo.user, pBasic->connId, pConn->id);
       }
     }
 
@@ -500,7 +621,7 @@ static int32_t mndProcessKillQueryReq(SRpcMsg *pReq) {
   SMnode       *pMnode = pReq->info.node;
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
 
-  SUserObj *pUser = mndAcquireUser(pMnode, pReq->conn.user);
+  SUserObj *pUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pUser == NULL) return 0;
   if (!pUser->superUser) {
     mndReleaseUser(pMnode, pUser);
@@ -515,17 +636,28 @@ static int32_t mndProcessKillQueryReq(SRpcMsg *pReq) {
     return -1;
   }
 
-  mInfo("kill query msg is received, queryId:%d", killReq.queryId);
+  mInfo("kill query msg is received, queryId:%s", killReq.queryStrId);
+  int32_t connId = 0;
+  uint64_t queryId = 0;
+  char* p = strchr(killReq.queryStrId, ':');
+  if (NULL == p) {
+    mError("invalid query id %s", killReq.queryStrId);
+    terrno = TSDB_CODE_MND_INVALID_QUERY_ID;
+    return -1;
+  }
+  *p = 0;
+  connId = taosStr2Int32(killReq.queryStrId, NULL, 16);
+  queryId = taosStr2UInt64(p + 1, NULL, 16);
 
-  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->cache, &killReq.connId, sizeof(int32_t));
+  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->connCache, &connId, sizeof(int32_t));
   if (pConn == NULL) {
-    mError("connId:%d, failed to kill queryId:%d, conn not exist", killReq.connId, killReq.queryId);
+    mError("connId:%x, failed to kill queryId:%" PRIx64 ", conn not exist", connId, queryId);
     terrno = TSDB_CODE_MND_INVALID_CONN_ID;
     return -1;
   } else {
-    mInfo("connId:%d, queryId:%d is killed by user:%s", killReq.connId, killReq.queryId, pReq->conn.user);
+    mInfo("connId:%x, queryId:%" PRIx64 " is killed by user:%s", connId, queryId, pReq->info.conn.user);
-    pConn->killId = killReq.queryId;
+    pConn->killId = queryId;
-    taosCacheRelease(pMgmt->cache, (void **)&pConn, false);
+    taosCacheRelease(pMgmt->connCache, (void **)&pConn, false);
     return 0;
   }
 }
@@ -534,7 +666,7 @@ static int32_t mndProcessKillConnReq(SRpcMsg *pReq) {
   SMnode       *pMnode = pReq->info.node;
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
 
-  SUserObj *pUser = mndAcquireUser(pMnode, pReq->conn.user);
+  SUserObj *pUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pUser == NULL) return 0;
   if (!pUser->superUser) {
     mndReleaseUser(pMnode, pUser);
@@ -549,15 +681,15 @@ static int32_t mndProcessKillConnReq(SRpcMsg *pReq) {
     return -1;
   }
 
-  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->cache, &killReq.connId, sizeof(int32_t));
+  SConnObj *pConn = taosCacheAcquireByKey(pMgmt->connCache, &killReq.connId, sizeof(uint32_t));
   if (pConn == NULL) {
-    mError("connId:%d, failed to kill connection, conn not exist", killReq.connId);
+    mError("connId:%u, failed to kill connection, conn not exist", killReq.connId);
     terrno = TSDB_CODE_MND_INVALID_CONN_ID;
     return -1;
   } else {
-    mInfo("connId:%d, is killed by user:%s", killReq.connId, pReq->conn.user);
+    mInfo("connId:%u, is killed by user:%s", killReq.connId, pReq->info.conn.user);
     pConn->killed = 1;
-    taosCacheRelease(pMgmt->cache, (void **)&pConn, false);
+    taosCacheRelease(pMgmt->connCache, (void **)&pConn, false);
     return TSDB_CODE_SUCCESS;
   }
 }
@@ -571,7 +703,7 @@ static int32_t mndRetrieveConns(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBl
 
   if (pShow->pIter == NULL) {
     SProfileMgmt *pMgmt = &pMnode->profileMgmt;
-    pShow->pIter = taosCacheCreateIter(pMgmt->cache);
+    pShow->pIter = taosCacheCreateIter(pMgmt->connCache);
   }
 
   while (numOfRows < rows) {
@@ -627,7 +759,7 @@ static int32_t mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *p
 
   if (pShow->pIter == NULL) {
     SProfileMgmt *pMgmt = &pMnode->profileMgmt;
-    pShow->pIter = taosCacheCreateIter(pMgmt->cache);
+    pShow->pIter = taosCacheCreateIter(pMgmt->connCache);
   }
 
   while (numOfRows < rows) {
@@ -645,7 +777,7 @@ static int32_t mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *p
 
     int32_t numOfQueries = taosArrayGetSize(pConn->pQueries);
     for (int32_t i = 0; i < numOfQueries; ++i) {
-      SQueryDesc* pQuery = taosArrayGet(pConn->pQueries, i);
+      SQueryDesc *pQuery = taosArrayGet(pConn->pQueries, i);
       cols = 0;
 
       char queryId[26 + VARSTR_HEADER_SIZE] = {0};
@@ -666,7 +798,7 @@ static int32_t mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *p
       colDataAppend(pColInfo, numOfRows, (const char *)app, false);
 
       pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-      colDataAppend(pColInfo, numOfRows, (const char *)&pQuery->pid, false);
+      colDataAppend(pColInfo, numOfRows, (const char *)&pConn->pid, false);
 
       char user[TSDB_USER_LEN + VARSTR_HEADER_SIZE] = {0};
       STR_TO_VARSTR(user, pConn->user);
@@ -698,7 +830,7 @@ static int32_t mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *p
         if (i) {
           offset += snprintf(subStatus + offset, strSize - offset - 1, ",");
         }
-        SQuerySubDesc* pDesc = taosArrayGet(pQuery->subDesc, i);
+        SQuerySubDesc *pDesc = taosArrayGet(pQuery->subDesc, i);
         offset += snprintf(subStatus + offset, strSize - offset - 1, "%" PRIu64 ":%s", pDesc->tid, pDesc->status);
       }
       varDataLen(subStatus) = strlen(&subStatus[VARSTR_HEADER_SIZE]);
@@ -720,6 +852,86 @@ static int32_t mndRetrieveQueries(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *p
   return numOfRows;
 }
 
+static int32_t mndRetrieveApps(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows) {
+  SMnode    *pMnode = pReq->info.node;
+  SSdb      *pSdb = pMnode->pSdb;
+  int32_t    numOfRows = 0;
+  int32_t    cols = 0;
+  SAppObj   *pApp = NULL;
+  
+  if (pShow->pIter == NULL) {
+    SProfileMgmt *pMgmt = &pMnode->profileMgmt;
+    pShow->pIter = taosCacheCreateIter(pMgmt->appCache);
+  }
+
+  while (numOfRows < rows) {
+    pApp = mndGetNextApp(pMnode, pShow->pIter);
+    if (pApp == NULL) {
+      pShow->pIter = NULL;
+      break;
+    }
+
+    cols = 0;
+
+    SColumnInfoData *pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->appId, false);
+
+    char ip[TSDB_IPv4ADDR_LEN + 6 + VARSTR_HEADER_SIZE] = {0};
+    sprintf(&ip[VARSTR_HEADER_SIZE], "%s", taosIpStr(pApp->ip));
+    varDataLen(ip) = strlen(&ip[VARSTR_HEADER_SIZE]);
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)ip, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->pid, false);
+
+    char name[TSDB_APP_NAME_LEN + 6 + VARSTR_HEADER_SIZE] = {0};
+    sprintf(&name[VARSTR_HEADER_SIZE], "%s", pApp->name);
+    varDataLen(name) = strlen(&name[VARSTR_HEADER_SIZE]);
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)name, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->startTime, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.numOfInsertsReq, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.numOfInsertRows, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.insertElapsedTime, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.insertBytes, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.fetchBytes, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.queryElapsedTime, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.numOfSlowQueries, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.totalRequests, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->summary.currentRequests, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)&pApp->lastAccessTimeMs, false);
+
+    numOfRows++;
+  }
+
+  pShow->numOfRows += numOfRows;
+  return numOfRows;
+}
+
+
 static void mndCancelGetNextQuery(SMnode *pMnode, void *pIter) {
   if (pIter != NULL) {
     taosCacheDestroyIter(pIter);
@@ -728,5 +940,5 @@ static void mndCancelGetNextQuery(SMnode *pMnode, void *pIter) {
 
 int32_t mndGetNumOfConnections(SMnode *pMnode) {
   SProfileMgmt *pMgmt = &pMnode->profileMgmt;
-  return taosCacheGetNumOfObj(pMgmt->cache);
+  return taosCacheGetNumOfObj(pMgmt->connCache);
 }

source/dnode/mnode/impl/src/mndQnode.c

@@ -294,7 +294,7 @@ static int32_t mndProcessCreateQnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_QNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_QNODE) != 0) {
     goto _OVER;
   }
 
@@ -396,7 +396,7 @@ static int32_t mndProcessDropQnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_QNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_QNODE) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndQuery.c

@@ -25,6 +25,13 @@ int32_t mndPreProcessMsg(SRpcMsg *pMsg) {
   return qWorkerPreprocessQueryMsg(pMnode->pQuery, pMsg);
 }
 
+void mndAbortPreprocessMsg(SRpcMsg *pMsg) {
+  if (TDMT_VND_QUERY != pMsg->msgType) return;
+
+  SMnode *pMnode = pMsg->info.node;
+  qWorkerAbortPreprocessQueryMsg(pMnode->pQuery, pMsg);
+}
+
 int32_t mndProcessQueryMsg(SRpcMsg *pMsg) {
   int32_t     code = -1;
   SMnode     *pMnode = pMsg->info.node;

source/dnode/mnode/impl/src/mndScheduler.c

@@ -63,9 +63,8 @@ int32_t mndConvertRsmaTask(char** pDst, int32_t* pDstLen, const char* ast, int64
       .topicQuery = false,
       .streamQuery = true,
       .rSmaQuery = true,
-      .triggerType = STREAM_TRIGGER_AT_ONCE,
+      .triggerType = triggerType,
       .watermark = watermark,
-      /*.filesFactor = filesFactor,*/
   };
 
   if (qCreateQueryPlan(&cxt, &pPlan, NULL) < 0) {
@@ -105,7 +104,7 @@ int32_t mndPersistTaskDeployReq(STrans* pTrans, SStreamTask* pTask, const SEpSet
   int32_t size = encoder.pos;
   int32_t tlen = sizeof(SMsgHead) + size;
   tEncoderClear(&encoder);
-  void* buf = taosMemoryMalloc(tlen);
+  void* buf = taosMemoryCalloc(1, tlen);
   if (buf == NULL) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
     return -1;
@@ -131,7 +130,7 @@ int32_t mndPersistTaskDeployReq(STrans* pTrans, SStreamTask* pTask, const SEpSet
 int32_t mndAddSinkToTask(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream, SStreamTask* pTask) {
   pTask->dispatchType = TASK_DISPATCH__NONE;
   // sink
-  if (pStream->createdBy == STREAM_CREATED_BY__SMA) {
+  if (pStream->smaId != 0) {
     pTask->sinkType = TASK_SINK__SMA;
     pTask->smaSink.smaId = pStream->smaId;
   } else {
@@ -157,6 +156,7 @@ int32_t mndAddDispatcherToInnerTask(SMnode* pMnode, STrans* pTrans, SStreamObj*
     }
     sdbRelease(pMnode->pSdb, pDb);
 
+    memcpy(pTask->shuffleDispatcher.stbFullName, pStream->targetSTbName, TSDB_TABLE_FNAME_LEN);
     SArray* pVgs = pTask->shuffleDispatcher.dbInfo.pVgroupInfos;
     int32_t sz = taosArrayGetSize(pVgs);
     SArray* sinkLv = taosArrayGetP(pStream->tasks, 0);
@@ -166,6 +166,7 @@ int32_t mndAddDispatcherToInnerTask(SMnode* pMnode, STrans* pTrans, SStreamObj*
       for (int32_t j = 0; j < sinkLvSize; j++) {
         SStreamTask* pLastLevelTask = taosArrayGetP(sinkLv, j);
         if (pLastLevelTask->nodeId == pVgInfo->vgId) {
+          ASSERT(pVgInfo->vgId > 0);
           pVgInfo->taskId = pLastLevelTask->taskId;
           ASSERT(pVgInfo->taskId != 0);
           break;
@@ -268,14 +269,13 @@ int32_t mndAddShuffleSinkTasksToStream(SMnode* pMnode, STrans* pTrans, SStreamOb
     pTask->epSet = mndGetVgroupEpset(pMnode, pVgroup);
 
     // source
-    pTask->sourceType = TASK_SOURCE__MERGE;
     pTask->inputType = TASK_INPUT_TYPE__DATA_BLOCK;
 
     // exec
     pTask->execType = TASK_EXEC__NONE;
 
     // sink
-    if (pStream->createdBy == STREAM_CREATED_BY__SMA) {
+    if (pStream->smaId != 0) {
       pTask->sinkType = TASK_SINK__SMA;
       pTask->smaSink.smaId = pStream->smaId;
     } else {
@@ -314,14 +314,13 @@ int32_t mndAddFixedSinkTaskToStream(SMnode* pMnode, STrans* pTrans, SStreamObj*
 #endif
   pTask->epSet = mndGetVgroupEpset(pMnode, &pStream->fixedSinkVg);
   // source
-  pTask->sourceType = TASK_SOURCE__MERGE;
   pTask->inputType = TASK_INPUT_TYPE__DATA_BLOCK;
 
   // exec
   pTask->execType = TASK_EXEC__NONE;
 
   // sink
-  if (pStream->createdBy == STREAM_CREATED_BY__SMA) {
+  if (pStream->smaId != 0) {
     pTask->sinkType = TASK_SINK__SMA;
     pTask->smaSink.smaId = pStream->smaId;
   } else {
@@ -346,8 +345,6 @@ int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream) {
     terrno = TSDB_CODE_QRY_INVALID_INPUT;
     return -1;
   }
-  ASSERT(pStream->vgNum == 0);
-
   int32_t totLevel = LIST_LENGTH(pPlan->pSubplans);
   ASSERT(totLevel <= 2);
   pStream->tasks = taosArrayInit(totLevel, sizeof(void*));
@@ -399,7 +396,7 @@ int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream) {
 
       // exec
       pFinalTask->execType = TASK_EXEC__PIPE;
-      SVgObj* pVgroup = mndSchedFetchOneVg(pMnode, pStream->dbUid);
+      SVgObj* pVgroup = mndSchedFetchOneVg(pMnode, pStream->sourceDbUid);
       if (mndAssignTaskToVg(pMnode, pTrans, pFinalTask, plan, pVgroup) < 0) {
         sdbRelease(pSdb, pVgroup);
         qDestroyQueryPlan(pPlan);
@@ -420,13 +417,15 @@ int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream) {
       SVgObj* pVgroup;
       pIter = sdbFetch(pSdb, SDB_VGROUP, pIter, (void**)&pVgroup);
       if (pIter == NULL) break;
-      if (pVgroup->dbUid != pStream->dbUid) {
+      if (pVgroup->dbUid != pStream->sourceDbUid) {
         sdbRelease(pSdb, pVgroup);
         continue;
       }
       SStreamTask* pTask = tNewSStreamTask(pStream->uid);
       mndAddTaskToTaskSet(taskSourceLevel, pTask);
 
+      pTask->dataScan = 1;
+
       // input
       pTask->inputType = TASK_INPUT_TYPE__SUMBIT_BLOCK;
 
@@ -463,13 +462,15 @@ int32_t mndScheduleStream(SMnode* pMnode, STrans* pTrans, SStreamObj* pStream) {
       SVgObj* pVgroup;
       pIter = sdbFetch(pSdb, SDB_VGROUP, pIter, (void**)&pVgroup);
       if (pIter == NULL) break;
-      if (pVgroup->dbUid != pStream->dbUid) {
+      if (pVgroup->dbUid != pStream->sourceDbUid) {
         sdbRelease(pSdb, pVgroup);
         continue;
       }
       SStreamTask* pTask = tNewSStreamTask(pStream->uid);
       mndAddTaskToTaskSet(taskOneLevel, pTask);
 
+      pTask->dataScan = 1;
+
       // input
       pTask->inputType = TASK_INPUT_TYPE__SUMBIT_BLOCK;
 

source/dnode/mnode/impl/src/mndShow.c

@@ -104,6 +104,8 @@ static int32_t convertToRetrieveType(char *name, int32_t len) {
     type = TSDB_MGMT_TABLE_TOPICS;
   } else if (strncasecmp(name, TSDB_PERFS_TABLE_STREAMS, len) == 0) {
     type = TSDB_MGMT_TABLE_STREAMS;
+  } else if (strncasecmp(name, TSDB_PERFS_TABLE_APPS, len) == 0) {
+    type = TSDB_MGMT_TABLE_APPS;
   } else {
     //    ASSERT(0);
   }
@@ -229,7 +231,7 @@ static int32_t mndProcessRetrieveSysTableReq(SRpcMsg *pReq) {
 
   mDebug("show:0x%" PRIx64 ", start retrieve data, type:%d", pShow->id, pShow->type);
 
-  // if (mndCheckShowAuth(pMnode, pReq->conn.user, pShow->type) != 0) return -1;
+  // if (mndCheckShowAuth(pMnode, pReq->info.conn.user, pShow->type) != 0) return -1;
 
   int32_t      numOfCols = pShow->pMeta->numOfColumns;
   SSDataBlock *pBlock = taosMemoryCalloc(1, sizeof(SSDataBlock));

source/dnode/mnode/impl/src/mndSma.c

@@ -20,6 +20,7 @@
 #include "mndDnode.h"
 #include "mndInfoSchema.h"
 #include "mndMnode.h"
+#include "mndScheduler.h"
 #include "mndShow.h"
 #include "mndStb.h"
 #include "mndStream.h"
@@ -520,6 +521,7 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   memcpy(smaObj.db, pDb->name, TSDB_DB_FNAME_LEN);
   smaObj.createdTime = taosGetTimestampMs();
   smaObj.uid = mndGenerateUid(pCreate->name, TSDB_TABLE_FNAME_LEN);
+  ASSERT(smaObj.uid != 0);
   char resultTbName[TSDB_TABLE_FNAME_LEN + 16] = {0};
   snprintf(resultTbName, TSDB_TABLE_FNAME_LEN + 16, "td.tsma.rst.tb.%s", pCreate->name);
   memcpy(smaObj.dstTbName, resultTbName, TSDB_TABLE_FNAME_LEN);
@@ -528,7 +530,10 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   smaObj.dbUid = pStb->dbUid;
   smaObj.intervalUnit = pCreate->intervalUnit;
   smaObj.slidingUnit = pCreate->slidingUnit;
+#if 0
   smaObj.timezone = pCreate->timezone;
+#endif
+  smaObj.timezone = tsTimezone; // use timezone of server
   smaObj.interval = pCreate->interval;
   smaObj.offset = pCreate->offset;
   smaObj.sliding = pCreate->sliding;
@@ -556,13 +561,13 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   streamObj.createTime = taosGetTimestampMs();
   streamObj.updateTime = streamObj.createTime;
   streamObj.uid = mndGenerateUid(pCreate->name, strlen(pCreate->name));
-  streamObj.dbUid = pDb->uid;
+  streamObj.sourceDbUid = pDb->uid;
   streamObj.version = 1;
   streamObj.sql = pCreate->sql;
-  streamObj.createdBy = STREAM_CREATED_BY__SMA;
   streamObj.smaId = smaObj.uid;
   streamObj.watermark = 0;
   streamObj.trigger = STREAM_TRIGGER_AT_ONCE;
+  streamObj.ast = strdup(smaObj.ast);
 
   if (mndAllocSmaVgroup(pMnode, pDb, &streamObj.fixedSinkVg) != 0) {
     mError("sma:%s, failed to create since %s", smaObj.name, terrstr());
@@ -571,10 +576,43 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   smaObj.dstVgId = streamObj.fixedSinkVg.vgId;
   streamObj.fixedSinkVgId = smaObj.dstVgId;
 
+  SNode *pAst = NULL;
+  if (nodesStringToNode(streamObj.ast, &pAst) < 0) {
+    ASSERT(0);
+    return -1;
+  }
+
+  // extract output schema from ast
+  if (qExtractResultSchema(pAst, (int32_t *)&streamObj.outputSchema.nCols, &streamObj.outputSchema.pSchema) != 0) {
+    ASSERT(0);
+    return -1;
+  }
+
+  SQueryPlan  *pPlan = NULL;
+  SPlanContext cxt = {
+      .pAstRoot = pAst,
+      .topicQuery = false,
+      .streamQuery = true,
+      .triggerType = streamObj.trigger,
+      .watermark = streamObj.watermark,
+  };
+
+  if (qCreateQueryPlan(&cxt, &pPlan, NULL) < 0) {
+    ASSERT(0);
+    return -1;
+  }
+
+  // save physcial plan
+  if (nodesNodeToString((SNode *)pPlan, false, &streamObj.physicalPlan, NULL) != 0) {
+    ASSERT(0);
+    return -1;
+  }
+  if (pAst != NULL) nodesDestroyNode(pAst);
+
   int32_t code = -1;
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_DB, pReq);
   if (pTrans == NULL) goto _OVER;
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
   mndTransSetSerial(pTrans);
   mDebug("trans:%d, used to create sma:%s", pTrans->id, pCreate->name);
 
@@ -585,7 +623,8 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   if (mndSetUpdateSmaStbCommitLogs(pMnode, pTrans, pStb) != 0) goto _OVER;
   // if (mndSetCreateSmaRedoActions(pMnode, pTrans, pDb, &smaObj) != 0) goto _OVER;
   if (mndSetCreateSmaVgroupRedoActions(pMnode, pTrans, pDb, &streamObj.fixedSinkVg, &smaObj) != 0) goto _OVER;
-  if (mndAddStreamToTrans(pMnode, &streamObj, pCreate->ast, pTrans) != 0) goto _OVER;
+  if (mndScheduleStream(pMnode, pTrans, &streamObj) != 0) goto _OVER;
+  if (mndPersistStream(pMnode, pTrans, &streamObj) != 0) goto _OVER;
   if (mndTransPrepare(pMnode, pTrans) != 0) goto _OVER;
 
   code = 0;
@@ -674,7 +713,7 @@ static int32_t mndProcessCreateSmaReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -816,7 +855,7 @@ static int32_t mndDropSma(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SSmaObj *p
   if (pTrans == NULL) goto _OVER;
 
   mDebug("trans:%d, used to drop sma:%s", pTrans->id, pSma->name);
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
 
   if (mndSetDropSmaRedoLogs(pMnode, pTrans, pSma) != 0) goto _OVER;
   if (mndSetDropSmaVgroupRedoLogs(pMnode, pTrans, pVgroup) != 0) goto _OVER;
@@ -935,7 +974,7 @@ static int32_t mndProcessDropSmaReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndSnode.c

@@ -300,7 +300,7 @@ static int32_t mndProcessCreateSnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_SNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_SNODE) != 0) {
     goto _OVER;
   }
 
@@ -404,7 +404,7 @@ static int32_t mndProcessDropSnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_SNODE) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_SNODE) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndStb.c

@@ -28,7 +28,6 @@
 #include "mndTrans.h"
 #include "mndUser.h"
 #include "mndVgroup.h"
-#include "mndSma.h"
 #include "tname.h"
 
 #define STB_VER_NUMBER   1
@@ -118,7 +117,7 @@ SSdbRaw *mndStbActionEncode(SStbObj *pStb) {
   }
 
   if (pStb->commentLen > 0) {
-    SDB_SET_BINARY(pRaw, dataPos, pStb->comment, pStb->commentLen, _OVER)
+    SDB_SET_BINARY(pRaw, dataPos, pStb->comment, pStb->commentLen + 1, _OVER)
   }
   if (pStb->ast1Len > 0) {
     SDB_SET_BINARY(pRaw, dataPos, pStb->pAst1, pStb->ast1Len, _OVER)
@@ -205,9 +204,9 @@ static SSdbRow *mndStbActionDecode(SSdbRaw *pRaw) {
   }
 
   if (pStb->commentLen > 0) {
-    pStb->comment = taosMemoryCalloc(pStb->commentLen, 1);
+    pStb->comment = taosMemoryCalloc(pStb->commentLen + 1, 1);
     if (pStb->comment == NULL) goto _OVER;
-    SDB_GET_BINARY(pRaw, dataPos, pStb->comment, pStb->commentLen, _OVER)
+    SDB_GET_BINARY(pRaw, dataPos, pStb->comment, pStb->commentLen + 1, _OVER)
   }
   if (pStb->ast1Len > 0) {
     pStb->pAst1 = taosMemoryCalloc(pStb->ast1Len, 1);
@@ -281,8 +280,8 @@ static int32_t mndStbActionUpdate(SSdb *pSdb, SStbObj *pOld, SStbObj *pNew) {
     }
   }
 
-  if (pOld->commentLen < pNew->commentLen) {
+  if (pOld->commentLen < pNew->commentLen && pNew->commentLen > 0) {
-    void *comment = taosMemoryMalloc(pNew->commentLen);
+    void *comment = taosMemoryMalloc(pNew->commentLen + 1);
     if (comment != NULL) {
       taosMemoryFree(pOld->comment);
       pOld->comment = comment;
@@ -292,6 +291,7 @@ static int32_t mndStbActionUpdate(SSdb *pSdb, SStbObj *pOld, SStbObj *pNew) {
       taosWUnLockLatch(&pOld->lock);
     }
   }
+  pOld->commentLen = pNew->commentLen;
 
   if (pOld->ast1Len < pNew->ast1Len) {
     void *pAst1 = taosMemoryMalloc(pNew->ast1Len);
@@ -331,8 +331,8 @@ static int32_t mndStbActionUpdate(SSdb *pSdb, SStbObj *pOld, SStbObj *pNew) {
     pOld->numOfTags = pNew->numOfTags;
     memcpy(pOld->pTags, pNew->pTags, pOld->numOfTags * sizeof(SSchema));
   }
-  if (pNew->commentLen != 0) {
+  if (pNew->commentLen > 0) {
-    memcpy(pOld->comment, pNew->comment, pNew->commentLen);
+    memcpy(pOld->comment, pNew->comment, pNew->commentLen + 1);
   }
   if (pNew->ast1Len != 0) {
     memcpy(pOld->pAst1, pNew->pAst1, pNew->ast1Len);
@@ -403,13 +403,13 @@ static void *mndBuildVCreateStbReq(SMnode *pMnode, SVgObj *pVgroup, SStbObj *pSt
     req.pRSmaParam.delay = pStb->delay;
     if (pStb->ast1Len > 0) {
       if (mndConvertRsmaTask(&req.pRSmaParam.qmsg1, &req.pRSmaParam.qmsg1Len, pStb->pAst1, pStb->uid,
-                             STREAM_TRIGGER_AT_ONCE, 0, req.pRSmaParam.xFilesFactor) != TSDB_CODE_SUCCESS) {
+                             STREAM_TRIGGER_WINDOW_CLOSE, 0, req.pRSmaParam.xFilesFactor) != TSDB_CODE_SUCCESS) {
         return NULL;
       }
     }
     if (pStb->ast2Len > 0) {
       if (mndConvertRsmaTask(&req.pRSmaParam.qmsg2, &req.pRSmaParam.qmsg2Len, pStb->pAst2, pStb->uid,
-                             STREAM_TRIGGER_AT_ONCE, 0, req.pRSmaParam.xFilesFactor) != TSDB_CODE_SUCCESS) {
+                             STREAM_TRIGGER_WINDOW_CLOSE, 0, req.pRSmaParam.xFilesFactor) != TSDB_CODE_SUCCESS) {
         return NULL;
       }
     }
@@ -670,19 +670,19 @@ int32_t mndBuildStbFromReq(SMnode *pMnode, SStbObj *pDst, SMCreateStbReq *pCreat
   pDst->tagVer = 1;
   pDst->colVer = 1;
   pDst->nextColId = 1;
-  pDst->xFilesFactor = pCreate->xFilesFactor;
+  // pDst->xFilesFactor = pCreate->xFilesFactor;
-  pDst->delay = pCreate->delay;
+  // pDst->delay = pCreate->delay;
   pDst->ttl = pCreate->ttl;
   pDst->numOfColumns = pCreate->numOfColumns;
   pDst->numOfTags = pCreate->numOfTags;
   pDst->commentLen = pCreate->commentLen;
   if (pDst->commentLen > 0) {
-    pDst->comment = taosMemoryCalloc(pDst->commentLen, 1);
+    pDst->comment = taosMemoryCalloc(pDst->commentLen + 1, 1);
     if (pDst->comment == NULL) {
       terrno = TSDB_CODE_OUT_OF_MEMORY;
       return -1;
     }
-    memcpy(pDst->comment, pCreate->comment, pDst->commentLen);
+    memcpy(pDst->comment, pCreate->comment, pDst->commentLen + 1);
   }
 
   pDst->ast1Len = pCreate->ast1Len;
@@ -759,7 +759,7 @@ _OVER:
 }
 
 int32_t mndAddStbToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SStbObj *pStb) {
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
   if (mndSetCreateStbRedoLogs(pMnode, pTrans, pDb, pStb) != 0) return -1;
   if (mndSetCreateStbUndoLogs(pMnode, pTrans, pDb, pStb) != 0) return -1;
   if (mndSetCreateStbCommitLogs(pMnode, pTrans, pDb, pStb) != 0) return -1;
@@ -806,7 +806,7 @@ static int32_t mndProcessCreateStbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -836,7 +836,7 @@ _OVER:
 }
 
 static int32_t mndCheckAlterStbReq(SMAlterStbReq *pAlter) {
-  if (pAlter->commentLen != 0 || pAlter->ttl != 0) return 0;
+  if (pAlter->commentLen >= 0 || pAlter->ttl != 0) return 0;
 
   if (pAlter->numOfFields < 1 || pAlter->numOfFields != (int32_t)taosArrayGetSize(pAlter->pFields)) {
     terrno = TSDB_CODE_MND_INVALID_STB_OPTION;
@@ -891,13 +891,16 @@ static int32_t mndUpdateStbCommentAndTTL(const SStbObj *pOld, SStbObj *pNew, cha
                                          int32_t ttl) {
   if (commentLen > 0) {
     pNew->commentLen = commentLen;
-    pNew->comment = taosMemoryCalloc(1, commentLen);
+    pNew->comment = taosMemoryCalloc(1, commentLen + 1);
     if (pNew->comment == NULL) {
       terrno = TSDB_CODE_OUT_OF_MEMORY;
       return -1;
     }
-    memcpy(pNew->comment, pComment, commentLen);
+    memcpy(pNew->comment, pComment, commentLen + 1);
+  } else if(commentLen == 0){
+    pNew->commentLen = 0;
   }
+
   if (ttl >= 0) {
     pNew->ttl = ttl;
   }
@@ -1207,7 +1210,6 @@ static int32_t mndSetAlterStbRedoActions(SMnode *pMnode, STrans *pTrans, SDbObj
       sdbRelease(pSdb, pVgroup);
       return -1;
     }
-
     STransAction action = {0};
     action.epSet = mndGetVgroupEpset(pMnode, pVgroup);
     action.pCont = pReq;
@@ -1271,7 +1273,8 @@ static int32_t mndBuildStbSchemaImp(SDbObj *pDb, SStbObj *pStb, const char *tbNa
   return 0;
 }
 
-static int32_t mndBuildStbSchema(SMnode *pMnode, const char *dbFName, const char *tbName, STableMetaRsp *pRsp, int32_t *smaVer) {
+static int32_t mndBuildStbSchema(SMnode *pMnode, const char *dbFName, const char *tbName, STableMetaRsp *pRsp,
+                                 int32_t *smaVer) {
   char tbFName[TSDB_TABLE_FNAME_LEN] = {0};
   snprintf(tbFName, sizeof(tbFName), "%s.%s", dbFName, tbName);
 
@@ -1396,7 +1399,7 @@ static int32_t mndAlterStb(SMnode *pMnode, SRpcMsg *pReq, const SMAlterStbReq *p
   if (pTrans == NULL) goto _OVER;
 
   mDebug("trans:%d, used to alter stb:%s", pTrans->id, pAlter->name);
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
 
   if (needRsp) {
     void   *pCont = NULL;
@@ -1454,7 +1457,7 @@ static int32_t mndProcessAlterStbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -1537,7 +1540,7 @@ static int32_t mndDropStb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SStbObj *p
   if (pTrans == NULL) goto _OVER;
 
   mDebug("trans:%d, used to drop stb:%s", pTrans->id, pStb->name);
-  mndTransSetDbName(pTrans, pDb->name);
+  mndTransSetDbName(pTrans, pDb->name, NULL);
 
   if (mndSetDropStbRedoLogs(pMnode, pTrans, pStb) != 0) goto _OVER;
   if (mndSetDropStbCommitLogs(pMnode, pTrans, pStb) != 0) goto _OVER;
@@ -1584,7 +1587,7 @@ static int32_t mndProcessDropStbReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -1769,16 +1772,23 @@ int32_t mndGetNumOfStbs(SMnode *pMnode, char *dbName, int32_t *pNumOfStbs) {
   return 0;
 }
 
-static void mndExtractTableName(char *tableId, char *name) {
+void mndExtractDbNameFromStbFullName(const char *stbFullName, char *dst) {
+  SName name = {0};
+  tNameFromString(&name, stbFullName, T_NAME_ACCT | T_NAME_DB | T_NAME_TABLE);
+
+  tNameGetFullDbName(&name, dst);
+}
+
+void mndExtractTbNameFromStbFullName(const char *stbFullName, char *dst, int32_t dstSize) {
   int32_t pos = -1;
   int32_t num = 0;
-  for (pos = 0; tableId[pos] != 0; ++pos) {
+  for (pos = 0; stbFullName[pos] != 0; ++pos) {
-    if (tableId[pos] == TS_PATH_DELIMITER[0]) num++;
+    if (stbFullName[pos] == TS_PATH_DELIMITER[0]) num++;
     if (num == 2) break;
   }
 
   if (num == 2) {
-    strcpy(name, tableId + pos + 1);
+    tstrncpy(dst, stbFullName + pos + 1, dstSize);
   }
 }
 
@@ -1808,7 +1818,7 @@ static int32_t mndRetrieveStb(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBloc
 
     SName name = {0};
     char  stbName[TSDB_TABLE_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
-    mndExtractTableName(pStb->name, &stbName[VARSTR_HEADER_SIZE]);
+    mndExtractTbNameFromStbFullName(pStb->name, &stbName[VARSTR_HEADER_SIZE], TSDB_TABLE_NAME_LEN);
     varDataSetLen(stbName, strlen(&stbName[VARSTR_HEADER_SIZE]));
 
     SColumnInfoData *pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
@@ -1834,17 +1844,17 @@ static int32_t mndRetrieveStb(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBloc
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, (const char *)&pStb->updateTime, false);  // number of tables
 
-    char *p = taosMemoryCalloc(1, pStb->commentLen + 1 + VARSTR_HEADER_SIZE);  // check malloc failures
-    if (p != NULL) {
-      if (pStb->commentLen != 0) {
-        STR_TO_VARSTR(p, pStb->comment);
-      } else {
-        STR_TO_VARSTR(p, "");
-      }
-
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols);
-      colDataAppend(pColInfo, numOfRows, (const char *)p, false);
+    if (pStb->commentLen > 0) {
-      taosMemoryFree(p);
+      char comment[TSDB_TB_COMMENT_LEN + VARSTR_HEADER_SIZE] = {0};
+      STR_TO_VARSTR(comment, pStb->comment);
+      colDataAppend(pColInfo, numOfRows, comment, false);
+    } else if(pStb->commentLen == 0) {
+      char comment[VARSTR_HEADER_SIZE + VARSTR_HEADER_SIZE] = {0};
+      STR_TO_VARSTR(comment, "");
+      colDataAppend(pColInfo, numOfRows, comment, false);
+    } else {
+      colDataAppendNULL(pColInfo, numOfRows);
     }
 
     numOfRows++;

source/dnode/mnode/impl/src/mndStream.c

@@ -196,7 +196,8 @@ void mndReleaseStream(SMnode *pMnode, SStreamObj *pStream) {
 }
 
 static int32_t mndCheckCreateStreamReq(SCMCreateStreamReq *pCreate) {
-  if (pCreate->name[0] == 0 || pCreate->sql == NULL || pCreate->sql[0] == 0) {
+  if (pCreate->name[0] == 0 || pCreate->sql == NULL || pCreate->sql[0] == 0 || pCreate->sourceDB[0] == 0 ||
+      pCreate->targetStbFullName[0] == 0) {
     terrno = TSDB_CODE_MND_INVALID_STREAM_OPTION;
     return -1;
   }
@@ -232,6 +233,114 @@ static int32_t mndStreamGetPlanString(const char *ast, int8_t triggerType, int64
   return code;
 }
 
+static int32_t mndBuildStreamObjFromCreateReq(SMnode *pMnode, SStreamObj *pObj, SCMCreateStreamReq *pCreate) {
+  SNode *pAst = NULL;
+
+  mDebug("stream:%s to create", pCreate->name);
+  memcpy(pObj->name, pCreate->name, TSDB_STREAM_FNAME_LEN);
+  pObj->createTime = taosGetTimestampMs();
+  pObj->updateTime = pObj->createTime;
+  pObj->version = 1;
+  pObj->smaId = 0;
+
+  pObj->uid = mndGenerateUid(pObj->name, strlen(pObj->name));
+  pObj->status = 0;
+
+  // TODO
+  pObj->dropPolicy = 0;
+  pObj->trigger = pCreate->triggerType;
+  pObj->triggerParam = pCreate->maxDelay;
+  pObj->watermark = pCreate->watermark;
+
+  memcpy(pObj->sourceDb, pCreate->sourceDB, TSDB_DB_FNAME_LEN);
+  SDbObj *pSourceDb = mndAcquireDb(pMnode, pCreate->sourceDB);
+  if (pSourceDb == NULL) {
+    /*ASSERT(0);*/
+    mDebug("stream:%s failed to create, source db %s not exist", pCreate->name, pObj->sourceDb);
+    terrno = TSDB_CODE_MND_DB_NOT_EXIST;
+    return -1;
+  }
+  pObj->sourceDbUid = pSourceDb->uid;
+
+  memcpy(pObj->targetSTbName, pCreate->targetStbFullName, TSDB_TABLE_FNAME_LEN);
+
+  SDbObj *pTargetDb = mndAcquireDbByStb(pMnode, pObj->targetSTbName);
+  if (pTargetDb == NULL) {
+    mDebug("stream:%s failed to create, target db %s not exist", pCreate->name, pObj->targetDb);
+    terrno = TSDB_CODE_MND_DB_NOT_EXIST;
+    return -1;
+  }
+  tstrncpy(pObj->targetDb, pTargetDb->name, TSDB_DB_FNAME_LEN);
+
+  pObj->targetStbUid = mndGenerateUid(pObj->targetSTbName, TSDB_TABLE_FNAME_LEN);
+  pObj->targetDbUid = pTargetDb->uid;
+
+  pObj->sql = pCreate->sql;
+  pObj->ast = pCreate->ast;
+
+  pCreate->sql = NULL;
+  pCreate->ast = NULL;
+
+  // deserialize ast
+  if (nodesStringToNode(pObj->ast, &pAst) < 0) {
+    /*ASSERT(0);*/
+    goto FAIL;
+  }
+
+  // extract output schema from ast
+  if (qExtractResultSchema(pAst, (int32_t *)&pObj->outputSchema.nCols, &pObj->outputSchema.pSchema) != 0) {
+    /*ASSERT(0);*/
+    goto FAIL;
+  }
+
+  SQueryPlan  *pPlan = NULL;
+  SPlanContext cxt = {
+      .pAstRoot = pAst,
+      .topicQuery = false,
+      .streamQuery = true,
+      .triggerType = pObj->trigger == STREAM_TRIGGER_MAX_DELAY ? STREAM_TRIGGER_WINDOW_CLOSE : pObj->trigger,
+      .watermark = pObj->watermark,
+  };
+
+  // using ast and param to build physical plan
+  if (qCreateQueryPlan(&cxt, &pPlan, NULL) < 0) {
+    /*ASSERT(0);*/
+    goto FAIL;
+  }
+
+  // save physcial plan
+  if (nodesNodeToString((SNode *)pPlan, false, &pObj->physicalPlan, NULL) != 0) {
+    /*ASSERT(0);*/
+    goto FAIL;
+  }
+
+FAIL:
+  if (pAst != NULL) nodesDestroyNode(pAst);
+  return 0;
+}
+
+int32_t mndPersistStream(SMnode *pMnode, STrans *pTrans, SStreamObj *pStream) {
+  SSdbRaw *pCommitRaw = mndStreamActionEncode(pStream);
+  if (pCommitRaw == NULL || mndTransAppendCommitlog(pTrans, pCommitRaw) != 0) {
+    mError("trans:%d, failed to append commit log since %s", pTrans->id, terrstr());
+    mndTransDrop(pTrans);
+    return -1;
+  }
+  sdbSetRawStatus(pCommitRaw, SDB_STATUS_READY);
+  return 0;
+}
+
+int32_t mndPersistDropStreamLog(SMnode *pMnode, STrans *pTrans, SStreamObj *pStream) {
+  SSdbRaw *pCommitRaw = mndStreamActionEncode(pStream);
+  if (pCommitRaw == NULL || mndTransAppendCommitlog(pTrans, pCommitRaw) != 0) {
+    mError("trans:%d, failed to append commit log since %s", pTrans->id, terrstr());
+    mndTransDrop(pTrans);
+    return -1;
+  }
+  sdbSetRawStatus(pCommitRaw, SDB_STATUS_DROPPED);
+  return 0;
+}
+
 int32_t mndAddStreamToTrans(SMnode *pMnode, SStreamObj *pStream, const char *ast, STrans *pTrans) {
   SNode *pAst = NULL;
 
@@ -321,7 +430,6 @@ static int32_t mndCreateStbForStream(SMnode *pMnode, STrans *pTrans, const SStre
     goto _OVER;
   }
 
-
   if (mndCheckDbAuth(pMnode, user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
@@ -357,7 +465,7 @@ static int32_t mndPersistTaskDropReq(STrans *pTrans, SStreamTask *pTask) {
   ASSERT(pTask->nodeId != 0);
 
   // vnode
-  if (pTask->nodeId > 0) {
+  /*if (pTask->nodeId > 0) {*/
   SVDropStreamTaskReq *pReq = taosMemoryCalloc(1, sizeof(SVDropStreamTaskReq));
   if (pReq == NULL) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
@@ -374,7 +482,7 @@ static int32_t mndPersistTaskDropReq(STrans *pTrans, SStreamTask *pTask) {
     taosMemoryFree(pReq);
     return -1;
   }
-  }
+  /*}*/
 
   return 0;
 }
@@ -404,10 +512,9 @@ static int32_t mndCreateStream(SMnode *pMnode, SRpcMsg *pReq, SCMCreateStreamReq
   streamObj.updateTime = streamObj.createTime;
   streamObj.uid = mndGenerateUid(pCreate->name, strlen(pCreate->name));
   streamObj.targetStbUid = mndGenerateUid(pCreate->targetStbFullName, TSDB_TABLE_FNAME_LEN);
-  streamObj.dbUid = pDb->uid;
+  streamObj.sourceDbUid = pDb->uid;
   streamObj.version = 1;
   streamObj.sql = pCreate->sql;
-  streamObj.createdBy = STREAM_CREATED_BY__USER;
   // TODO
   streamObj.fixedSinkVgId = 0;
   streamObj.smaId = 0;
@@ -438,7 +545,7 @@ static int32_t mndCreateStream(SMnode *pMnode, SRpcMsg *pReq, SCMCreateStreamReq
     return -1;
   }
 
-  if (streamObj.targetSTbName[0] && mndCreateStbForStream(pMnode, pTrans, &streamObj, pReq->conn.user) < 0) {
+  if (streamObj.targetSTbName[0] && mndCreateStbForStream(pMnode, pTrans, &streamObj, pReq->info.conn.user) < 0) {
     mError("trans:%d, failed to create stb for stream since %s", pTrans->id, terrstr());
     mndTransDrop(pTrans);
     return -1;
@@ -487,19 +594,71 @@ static int32_t mndProcessCreateStreamReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  // TODO check auth
+  // TODO check read auth for source and write auth for target
+#if 0
   pDb = mndAcquireDb(pMnode, createStreamReq.sourceDB);
   if (pDb == NULL) {
     terrno = TSDB_CODE_MND_DB_NOT_SELECTED;
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+    goto _OVER;
+  }
+#endif
+
+  STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_DB_INSIDE, pReq);
+  if (pTrans == NULL) {
+    mError("stream:%s, failed to create since %s", createStreamReq.name, terrstr());
     goto _OVER;
   }
 
-  code = mndCreateStream(pMnode, pReq, &createStreamReq, pDb);
+  mndTransSetDbName(pTrans, createStreamReq.sourceDB, NULL);
-  if (code == 0) code = TSDB_CODE_ACTION_IN_PROGRESS;
+  // TODO
+  /*mndTransSetDbName(pTrans, streamObj.targetDb, NULL);*/
+  mDebug("trans:%d, used to create stream:%s", pTrans->id, createStreamReq.name);
+
+  // build stream obj from request
+  SStreamObj streamObj = {0};
+  if (mndBuildStreamObjFromCreateReq(pMnode, &streamObj, &createStreamReq) < 0) {
+    ASSERT(0);
+    mError("stream:%s, failed to create since %s", createStreamReq.name, terrstr());
+    goto _OVER;
+  }
+
+  // create stb for stream
+  if (mndCreateStbForStream(pMnode, pTrans, &streamObj, pReq->info.conn.user) < 0) {
+    mError("trans:%d, failed to create stb for stream %s since %s", pTrans->id, createStreamReq.name, terrstr());
+    mndTransDrop(pTrans);
+    goto _OVER;
+  }
+
+  // schedule stream task for stream obj
+  if (mndScheduleStream(pMnode, pTrans, &streamObj) < 0) {
+    mError("stream:%s, failed to schedule since %s", createStreamReq.name, terrstr());
+    mndTransDrop(pTrans);
+    goto _OVER;
+  }
+
+  // add stream to trans
+  if (mndPersistStream(pMnode, pTrans, &streamObj) < 0) {
+    mError("stream:%s, failed to schedule since %s", createStreamReq.name, terrstr());
+    mndTransDrop(pTrans);
+    goto _OVER;
+  }
+
+  // execute creation
+  if (mndTransPrepare(pMnode, pTrans) != 0) {
+    mError("trans:%d, failed to prepare since %s", pTrans->id, terrstr());
+    mndTransDrop(pTrans);
+    goto _OVER;
+  }
+
+  mndTransDrop(pTrans);
+
+  /*code = mndCreateStream(pMnode, pReq, &createStreamReq, pDb);*/
+  /*if (code == 0) code = TSDB_CODE_ACTION_IN_PROGRESS;*/
+  code = TSDB_CODE_ACTION_IN_PROGRESS;
 
 _OVER:
   if (code != 0 && code != TSDB_CODE_ACTION_IN_PROGRESS) {
@@ -520,18 +679,24 @@ static int32_t mndProcessDropStreamReq(SRpcMsg *pReq) {
   /*SDbObj     *pDb = NULL;*/
   /*SUserObj   *pUser = NULL;*/
 
-  SMDropStreamTaskReq dropStreamReq = *(SMDropStreamTaskReq *)pReq->pCont;
+  SMDropStreamReq dropReq = *(SMDropStreamReq *)pReq->pCont;
 
-  pStream = mndAcquireStream(pMnode, dropStreamReq.name);
+  pStream = mndAcquireStream(pMnode, dropReq.name);
 
   if (pStream == NULL) {
+    if (dropReq.igNotExists) {
+      mDebug("stream:%s, not exist, ignore not exist is set", dropReq.name);
+      code = 0;
+      goto DROP_STREAM_OVER;
+    } else {
       terrno = TSDB_CODE_MND_STREAM_NOT_EXIST;
       return -1;
     }
+  }
 
 #if 0
   // todo check auth
-  pUser = mndAcquireUser(pMnode, pReq->conn.user);
+  pUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pUser == NULL) {
     goto DROP_STREAM_OVER;
   }
@@ -539,18 +704,62 @@ static int32_t mndProcessDropStreamReq(SRpcMsg *pReq) {
 
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_NOTHING, pReq);
   if (pTrans == NULL) {
-    mError("stream:%s, failed to drop since %s", dropStreamReq.name, terrstr());
+    mError("stream:%s, failed to drop since %s", dropReq.name, terrstr());
-    return -1;
+    return code;
   }
-  mDebug("trans:%d, used to drop stream:%s", pTrans->id, dropStreamReq.name);
+  mDebug("trans:%d, used to drop stream:%s", pTrans->id, dropReq.name);
 
   // drop all tasks
   if (mndDropStreamTasks(pMnode, pTrans, pStream) < 0) {
-    mError("stream:%s, failed to drop task since %s", dropStreamReq.name, terrstr());
+    mError("stream:%s, failed to drop task since %s", dropReq.name, terrstr());
+    return code;
+  }
+
+  // drop stream
+  if (mndPersistDropStreamLog(pMnode, pTrans, pStream) < 0) {
+    sdbRelease(pMnode->pSdb, pStream);
     return -1;
   }
 
 DROP_STREAM_OVER:
+  return code;
+}
+
+int32_t mndDropStreamByDb(SMnode *pMnode, STrans *pTrans, SDbObj *pDb) {
+  SSdb *pSdb = pMnode->pSdb;
+
+  void       *pIter = NULL;
+  SStreamObj *pStream = NULL;
+  while (1) {
+    pIter = sdbFetch(pSdb, SDB_STREAM, pIter, (void **)&pStream);
+    if (pIter == NULL) break;
+
+    if (pStream->sourceDbUid == pDb->uid || pStream->targetDbUid == pDb->uid) {
+      if (pStream->sourceDbUid != pStream->targetDbUid) {
+        sdbRelease(pSdb, pStream);
+        return -1;
+      } else {
+        // TODO drop all task on snode
+        if (mndPersistDropStreamLog(pMnode, pTrans, pStream) < 0) {
+          sdbRelease(pSdb, pStream);
+          return -1;
+        }
+      }
+    } else {
+      sdbRelease(pSdb, pStream);
+      continue;
+    }
+
+#if 0
+    if (mndSetDropOffsetStreamLogs(pMnode, pTrans, pStream) < 0) {
+      sdbRelease(pSdb, pStream);
+      goto END;
+    }
+#endif
+
+    sdbRelease(pSdb, pStream);
+  }
+
   return 0;
 }
 
@@ -569,7 +778,7 @@ static int32_t mndGetNumOfStreams(SMnode *pMnode, char *dbName, int32_t *pNumOfS
     pIter = sdbFetch(pSdb, SDB_STREAM, pIter, (void **)&pStream);
     if (pIter == NULL) break;
 
-    if (pStream->dbUid == pDb->uid) {
+    if (pStream->sourceDbUid == pDb->uid) {
       numOfStreams++;
     }
 

source/dnode/mnode/impl/src/mndSubscribe.c

@@ -403,7 +403,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR
 
 static int32_t mndPersistRebResult(SMnode *pMnode, SRpcMsg *pMsg, const SMqRebOutputObj *pOutput) {
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_DB_INSIDE, pMsg);
-  mndTransSetDbName(pTrans, pOutput->pSub->dbName);
+  mndTransSetDbName(pTrans, pOutput->pSub->dbName, NULL);
   if (pTrans == NULL) return -1;
 
   // make txn:

source/dnode/mnode/impl/src/mndSync.c

@@ -46,19 +46,21 @@ void mndSyncCommitMsg(struct SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbM
 
   int32_t transId = sdbGetIdFromRaw(pMnode->pSdb, pRaw);
   pMgmt->errCode = cbMeta.code;
-  mDebug("trans:%d, is proposed, saved:%d code:0x%x, index:%" PRId64 " term:%" PRId64 " role:%s raw:%p", transId,
+  mDebug("trans:%d, is proposed, saved:%d code:0x%x, apply index:%" PRId64 " term:%" PRIu64 " config:%" PRId64
-         pMgmt->transId, cbMeta.code, cbMeta.index, cbMeta.term, syncStr(cbMeta.state), pRaw);
+         " role:%s raw:%p",
+         transId, pMgmt->transId, cbMeta.code, cbMeta.index, cbMeta.term, cbMeta.lastConfigIndex, syncStr(cbMeta.state),
+         pRaw);
 
   if (pMgmt->errCode == 0) {
     sdbWriteWithoutFree(pMnode->pSdb, pRaw);
-    sdbSetApplyIndex(pMnode->pSdb, cbMeta.index);
+    sdbSetApplyInfo(pMnode->pSdb, cbMeta.index, cbMeta.term, cbMeta.lastConfigIndex);
-    sdbSetApplyTerm(pMnode->pSdb, cbMeta.term);
   }
 
   if (pMgmt->transId == transId) {
     if (pMgmt->errCode != 0) {
       mError("trans:%d, failed to propose since %s", transId, tstrerror(pMgmt->errCode));
     }
+    pMgmt->transId = 0;
     tsem_post(&pMgmt->syncSem);
   } else {
     STrans *pTrans = mndAcquireTrans(pMnode, transId);
@@ -67,32 +69,29 @@ void mndSyncCommitMsg(struct SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbM
       mndReleaseTrans(pMnode, pTrans);
     }
 
-    if (cbMeta.index - sdbGetApplyIndex(pMnode->pSdb) > 100) {
+    sdbWriteFile(pMnode->pSdb, SDB_WRITE_DELTA);
-      SSnapshotMeta sMeta = {0};
-      if (syncGetSnapshotMeta(pMnode->syncMgmt.sync, &sMeta) == 0) {
-        sdbSetCurConfig(pMnode->pSdb, sMeta.lastConfigIndex);
-      }
-      sdbWriteFile(pMnode->pSdb);
-    }
   }
 }
 
-int32_t mndSyncGetSnapshot(struct SSyncFSM *pFsm, SSnapshot *pSnapshot) {
+int32_t mndSyncGetSnapshot(struct SSyncFSM *pFsm, SSnapshot *pSnapshot, void *pReaderParam, void **ppReader) {
+  // TODO:
+
+  // atomic operation
+  // step1. sdbGetCommitInfo
+  // step2. create ppReader with pReaderParam
+
+  return 0;
+}
+
+int32_t mndSyncGetSnapshotInfo(struct SSyncFSM *pFsm, SSnapshot *pSnapshot) {
   SMnode *pMnode = pFsm->data;
-  pSnapshot->lastApplyIndex = sdbGetCommitIndex(pMnode->pSdb);
+  sdbGetCommitInfo(pMnode->pSdb, &pSnapshot->lastApplyIndex, &pSnapshot->lastApplyTerm, &pSnapshot->lastConfigIndex);
-  pSnapshot->lastApplyTerm = sdbGetCommitTerm(pMnode->pSdb);
-  pSnapshot->lastConfigIndex = sdbGetCurConfig(pMnode->pSdb);
   return 0;
 }
 
 void mndRestoreFinish(struct SSyncFSM *pFsm) {
   SMnode *pMnode = pFsm->data;
 
-  SSnapshotMeta sMeta = {0};
-  if (syncGetSnapshotMeta(pMnode->syncMgmt.sync, &sMeta) == 0) {
-    sdbSetCurConfig(pMnode->pSdb, sMeta.lastConfigIndex);
-  }
-
   if (!pMnode->deploy) {
     mInfo("mnode sync restore finished, and will handle outstanding transactions");
     mndTransPullup(pMnode);
@@ -122,6 +121,7 @@ void mndReConfig(struct SSyncFSM *pFsm, const SRpcMsg *pMsg, SReConfigCbMeta cbM
     if (pMgmt->errCode != 0) {
       mError("trans:-1, failed to propose sync reconfig since %s", tstrerror(pMgmt->errCode));
     }
+    pMgmt->transId = 0;
     tsem_post(&pMgmt->syncSem);
   }
 }
@@ -169,6 +169,7 @@ SSyncFSM *mndSyncMakeFsm(SMnode *pMnode) {
   pFsm->FpRestoreFinishCb = mndRestoreFinish;
   pFsm->FpReConfigCb = mndReConfig;
   pFsm->FpGetSnapshot = mndSyncGetSnapshot;
+  pFsm->FpGetSnapshotInfo = mndSyncGetSnapshotInfo;
   pFsm->FpSnapshotStartRead = mndSnapshotStartRead;
   pFsm->FpSnapshotStopRead = mndSnapshotStopRead;
   pFsm->FpSnapshotDoRead = mndSnapshotDoRead;
@@ -221,17 +222,17 @@ void mndCleanupSync(SMnode *pMnode) {
 
 int32_t mndSyncPropose(SMnode *pMnode, SSdbRaw *pRaw, int32_t transId) {
   SSyncMgmt *pMgmt = &pMnode->syncMgmt;
-  SRpcMsg    rsp = {.code = TDMT_MND_APPLY_MSG, .contLen = sdbGetRawTotalSize(pRaw)};
+  SRpcMsg    req = {.msgType = TDMT_MND_APPLY_MSG, .contLen = sdbGetRawTotalSize(pRaw)};
-  rsp.pCont = rpcMallocCont(rsp.contLen);
+  req.pCont = rpcMallocCont(req.contLen);
-  if (rsp.pCont == NULL) return -1;
+  if (req.pCont == NULL) return -1;
-  memcpy(rsp.pCont, pRaw, rsp.contLen);
+  memcpy(req.pCont, pRaw, req.contLen);
 
   pMgmt->errCode = 0;
   pMgmt->transId = transId;
   mTrace("trans:%d, will be proposed", pMgmt->transId);
 
   const bool isWeak = false;
-  int32_t    code = syncPropose(pMgmt->sync, &rsp, isWeak);
+  int32_t    code = syncPropose(pMgmt->sync, &req, isWeak);
   if (code == 0) {
     tsem_wait(&pMgmt->syncSem);
   } else if (code == -1 && terrno == TSDB_CODE_SYN_NOT_LEADER) {
@@ -242,7 +243,7 @@ int32_t mndSyncPropose(SMnode *pMnode, SSdbRaw *pRaw, int32_t transId) {
     terrno = TSDB_CODE_APP_ERROR;
   }
 
-  rpcFreeCont(rsp.pCont);
+  rpcFreeCont(req.pCont);
   if (code != 0) {
     mError("trans:%d, failed to propose, code:0x%x", pMgmt->transId, code);
     return code;
@@ -256,23 +257,19 @@ void mndSyncStart(SMnode *pMnode) {
   syncSetMsgCb(pMgmt->sync, &pMnode->msgCb);
   syncStart(pMgmt->sync);
   mDebug("mnode sync started, id:%" PRId64 " standby:%d", pMgmt->sync, pMgmt->standby);
-
-  /*
-    if (pMgmt->standby) {
-      syncStartStandBy(pMgmt->sync);
-    } else {
-      syncStart(pMgmt->sync);
-    }
-  */
 }
 
-void mndSyncStop(SMnode *pMnode) {}
+void mndSyncStop(SMnode *pMnode) {
+  if (pMnode->syncMgmt.transId != 0) {
+    pMnode->syncMgmt.transId = 0;
+    tsem_post(&pMnode->syncMgmt.syncSem);
+  }
+}
 
 bool mndIsMaster(SMnode *pMnode) {
   SSyncMgmt *pMgmt = &pMnode->syncMgmt;
 
-  ESyncState state = syncGetMyRole(pMgmt->sync);
+  if (!syncIsReady(pMgmt->sync)) {
-  if (state != TAOS_SYNC_STATE_LEADER) {
     terrno = TSDB_CODE_SYN_NOT_LEADER;
     return false;
   }

source/dnode/mnode/impl/src/mndTopic.c

@@ -474,7 +474,7 @@ static int32_t mndProcessCreateTopicReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckDbAuth(pMnode, pReq->conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
+  if (mndCheckDbAuth(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pDb) != 0) {
     goto _OVER;
   }
 
@@ -566,7 +566,7 @@ static int32_t mndProcessDropTopicReq(SRpcMsg *pReq) {
 #endif
 
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_DB_INSIDE, pReq);
-  mndTransSetDbName(pTrans, pTopic->db);
+  mndTransSetDbName(pTrans, pTopic->db, NULL);
   if (pTrans == NULL) {
     mError("topic:%s, failed to drop since %s", pTopic->name, terrstr());
     return -1;

source/dnode/mnode/impl/src/mndTrans.c

@@ -52,7 +52,7 @@ static bool    mndTransPerformCommitActionStage(SMnode *pMnode, STrans *pTrans);
 static bool    mndTransPerformCommitStage(SMnode *pMnode, STrans *pTrans);
 static bool    mndTransPerformRollbackStage(SMnode *pMnode, STrans *pTrans);
 static bool    mndTransPerfromFinishedStage(SMnode *pMnode, STrans *pTrans);
-static bool    mndCantExecuteTransAction(SMnode *pMnode) { return !pMnode->deploy && !mndIsMaster(pMnode); }
+static bool    mndCannotExecuteTransAction(SMnode *pMnode) { return !pMnode->deploy && !mndIsMaster(pMnode); }
 
 static void    mndTransSendRpcRsp(SMnode *pMnode, STrans *pTrans);
 static int32_t mndProcessTransReq(SRpcMsg *pReq);
@@ -122,7 +122,8 @@ static SSdbRaw *mndTransActionEncode(STrans *pTrans) {
   SDB_SET_INT8(pRaw, dataPos, pTrans->conflict, _OVER)
   SDB_SET_INT8(pRaw, dataPos, pTrans->exec, _OVER)
   SDB_SET_INT64(pRaw, dataPos, pTrans->createdTime, _OVER)
-  SDB_SET_BINARY(pRaw, dataPos, pTrans->dbname, TSDB_DB_FNAME_LEN, _OVER)
+  SDB_SET_BINARY(pRaw, dataPos, pTrans->dbname1, TSDB_DB_FNAME_LEN, _OVER)
+  SDB_SET_BINARY(pRaw, dataPos, pTrans->dbname2, TSDB_DB_FNAME_LEN, _OVER)
   SDB_SET_INT32(pRaw, dataPos, pTrans->redoActionPos, _OVER)
 
   int32_t redoActionNum = taosArrayGetSize(pTrans->redoActions);
@@ -270,7 +271,8 @@ static SSdbRow *mndTransActionDecode(SSdbRaw *pRaw) {
   pTrans->conflict = conflict;
   pTrans->exec = exec;
   SDB_GET_INT64(pRaw, dataPos, &pTrans->createdTime, _OVER)
-  SDB_GET_BINARY(pRaw, dataPos, pTrans->dbname, TSDB_DB_FNAME_LEN, _OVER)
+  SDB_GET_BINARY(pRaw, dataPos, pTrans->dbname1, TSDB_DB_FNAME_LEN, _OVER)
+  SDB_GET_BINARY(pRaw, dataPos, pTrans->dbname2, TSDB_DB_FNAME_LEN, _OVER)
   SDB_GET_INT32(pRaw, dataPos, &pTrans->redoActionPos, _OVER)
   SDB_GET_INT32(pRaw, dataPos, &redoActionNum, _OVER)
   SDB_GET_INT32(pRaw, dataPos, &undoActionNum, _OVER)
@@ -521,9 +523,10 @@ static int32_t mndTransActionUpdate(SSdb *pSdb, STrans *pOld, STrans *pNew) {
   }
 
   if (pOld->stage == TRN_STAGE_ROLLBACK) {
-    pOld->stage = TRN_STAGE_FINISHED;
+    pOld->stage = TRN_STAGE_REDO_ACTION;
-    mTrace("trans:%d, stage from rollback to finished since perform update action", pNew->id);
+    mTrace("trans:%d, stage from rollback to undoAction since perform update action", pNew->id);
   }
+
   return 0;
 }
 
@@ -649,7 +652,14 @@ void mndTransSetCb(STrans *pTrans, ETrnFunc startFunc, ETrnFunc stopFunc, void *
   pTrans->paramLen = paramLen;
 }
 
-void mndTransSetDbName(STrans *pTrans, const char *dbname) { memcpy(pTrans->dbname, dbname, TSDB_DB_FNAME_LEN); }
+void mndTransSetDbName(STrans *pTrans, const char *dbname1, const char *dbname2) {
+  if (dbname1 != NULL) {
+    memcpy(pTrans->dbname1, dbname1, TSDB_DB_FNAME_LEN);
+  }
+  if (dbname2 != NULL) {
+    memcpy(pTrans->dbname2, dbname2, TSDB_DB_FNAME_LEN);
+  }
+}
 
 void mndTransSetSerial(STrans *pTrans) { pTrans->exec = TRN_EXEC_SERIAL; }
 
@@ -661,7 +671,7 @@ static int32_t mndTransSync(SMnode *pMnode, STrans *pTrans) {
   }
   sdbSetRawStatus(pRaw, SDB_STATUS_READY);
 
-  mDebug("trans:%d, sync to other mnodes", pTrans->id);
+  mDebug("trans:%d, sync to other mnodes, stage:%s", pTrans->id, mndTransStr(pTrans->stage));
   int32_t code = mndSyncPropose(pMnode, pRaw, pTrans->id);
   if (code != 0) {
     mError("trans:%d, failed to sync since %s", pTrans->id, terrstr());
@@ -674,6 +684,12 @@ static int32_t mndTransSync(SMnode *pMnode, STrans *pTrans) {
   return 0;
 }
 
+static bool mndCheckDbConflict(const char *db, STrans *pTrans) {
+  if (db[0] == 0) return false;
+  if (strcmp(db, pTrans->dbname1) == 0 || strcmp(db, pTrans->dbname2) == 0) return true;
+  return false;
+}
+
 static bool mndCheckTransConflict(SMnode *pMnode, STrans *pNew) {
   STrans *pTrans = NULL;
   void   *pIter = NULL;
@@ -688,14 +704,21 @@ static bool mndCheckTransConflict(SMnode *pMnode, STrans *pNew) {
     if (pNew->conflict == TRN_CONFLICT_GLOBAL) conflict = true;
     if (pNew->conflict == TRN_CONFLICT_DB) {
       if (pTrans->conflict == TRN_CONFLICT_GLOBAL) conflict = true;
-      if (pTrans->conflict == TRN_CONFLICT_DB && strcmp(pNew->dbname, pTrans->dbname) == 0) conflict = true;
+      if (pTrans->conflict == TRN_CONFLICT_DB || pTrans->conflict == TRN_CONFLICT_DB_INSIDE) {
-      if (pTrans->conflict == TRN_CONFLICT_DB_INSIDE && strcmp(pNew->dbname, pTrans->dbname) == 0) conflict = true;
+        if (mndCheckDbConflict(pNew->dbname1, pTrans)) conflict = true;
+        if (mndCheckDbConflict(pNew->dbname2, pTrans)) conflict = true;
+      }
     }
     if (pNew->conflict == TRN_CONFLICT_DB_INSIDE) {
       if (pTrans->conflict == TRN_CONFLICT_GLOBAL) conflict = true;
-      if (pTrans->conflict == TRN_CONFLICT_DB && strcmp(pNew->dbname, pTrans->dbname) == 0) conflict = true;
+      if (pTrans->conflict == TRN_CONFLICT_DB) {
+        if (mndCheckDbConflict(pNew->dbname1, pTrans)) conflict = true;
+        if (mndCheckDbConflict(pNew->dbname2, pTrans)) conflict = true;
       }
-    mError("trans:%d, can't execute since conflict with trans:%d, db:%s", pNew->id, pTrans->id, pTrans->dbname);
+    }
+
+    mError("trans:%d, can't execute since conflict with trans:%d, db1:%s db2:%s", pNew->id, pTrans->id, pTrans->dbname1,
+           pTrans->dbname2);
     sdbRelease(pMnode->pSdb, pTrans);
   }
 
@@ -704,7 +727,7 @@ static bool mndCheckTransConflict(SMnode *pMnode, STrans *pNew) {
 
 int32_t mndTransPrepare(SMnode *pMnode, STrans *pTrans) {
   if (pTrans->conflict == TRN_CONFLICT_DB || pTrans->conflict == TRN_CONFLICT_DB_INSIDE) {
-    if (strlen(pTrans->dbname) == 0) {
+    if (strlen(pTrans->dbname1) == 0 && strlen(pTrans->dbname2) == 0) {
       terrno = TSDB_CODE_MND_TRANS_CONFLICT;
       mError("trans:%d, failed to prepare conflict db not set", pTrans->id);
       return -1;
@@ -915,7 +938,7 @@ static int32_t mndTransWriteSingleLog(SMnode *pMnode, STrans *pTrans, STransActi
 
 static int32_t mndTransSendSingleMsg(SMnode *pMnode, STrans *pTrans, STransAction *pAction) {
   if (pAction->msgSent) return 0;
-  if (mndCantExecuteTransAction(pMnode)) return -1;
+  if (mndCannotExecuteTransAction(pMnode)) return -1;
 
   int64_t signature = pTrans->id;
   signature = (signature << 32);
@@ -1115,7 +1138,7 @@ static int32_t mndTransExecuteRedoActionsSerial(SMnode *pMnode, STrans *pTrans)
       pTrans->lastEpset = pAction->epSet;
     }
 
-    if (mndCantExecuteTransAction(pMnode)) break;
+    if (mndCannotExecuteTransAction(pMnode)) break;
 
     if (code == 0) {
       pTrans->code = 0;
@@ -1158,7 +1181,7 @@ static bool mndTransPerformRedoActionStage(SMnode *pMnode, STrans *pTrans) {
     code = mndTransExecuteRedoActions(pMnode, pTrans);
   }
 
-  if (mndCantExecuteTransAction(pMnode)) return false;
+  if (mndCannotExecuteTransAction(pMnode)) return false;
 
   if (code == 0) {
     pTrans->code = 0;
@@ -1171,8 +1194,8 @@ static bool mndTransPerformRedoActionStage(SMnode *pMnode, STrans *pTrans) {
   } else {
     pTrans->code = terrno;
     if (pTrans->policy == TRN_POLICY_ROLLBACK) {
-      pTrans->stage = TRN_STAGE_UNDO_ACTION;
+      pTrans->stage = TRN_STAGE_ROLLBACK;
-      mError("trans:%d, stage from redoAction to undoAction since %s", pTrans->id, terrstr());
+      mError("trans:%d, stage from redoAction to rollback since %s", pTrans->id, terrstr());
       continueExec = true;
     } else {
       pTrans->failedTimes++;
@@ -1185,7 +1208,7 @@ static bool mndTransPerformRedoActionStage(SMnode *pMnode, STrans *pTrans) {
 }
 
 static bool mndTransPerformCommitStage(SMnode *pMnode, STrans *pTrans) {
-  if (mndCantExecuteTransAction(pMnode)) return false;
+  if (mndCannotExecuteTransAction(pMnode)) return false;
 
   bool    continueExec = true;
   int32_t code = mndTransCommit(pMnode, pTrans);
@@ -1197,17 +1220,10 @@ static bool mndTransPerformCommitStage(SMnode *pMnode, STrans *pTrans) {
     continueExec = true;
   } else {
     pTrans->code = terrno;
-    if (pTrans->policy == TRN_POLICY_ROLLBACK) {
-      pTrans->stage = TRN_STAGE_UNDO_ACTION;
-      mError("trans:%d, stage from commit to undoAction since %s, failedTimes:%d", pTrans->id, terrstr(),
-             pTrans->failedTimes);
-      continueExec = true;
-    } else {
     pTrans->failedTimes++;
     mError("trans:%d, stage keep on commit since %s, failedTimes:%d", pTrans->id, terrstr(), pTrans->failedTimes);
     continueExec = false;
   }
-  }
 
   return continueExec;
 }
@@ -1235,11 +1251,9 @@ static bool mndTransPerformUndoActionStage(SMnode *pMnode, STrans *pTrans) {
   bool    continueExec = true;
   int32_t code = mndTransExecuteUndoActions(pMnode, pTrans);
 
-  if (mndCantExecuteTransAction(pMnode)) return false;
-
   if (code == 0) {
-    pTrans->stage = TRN_STAGE_ROLLBACK;
+    pTrans->stage = TRN_STAGE_FINISHED;
-    mDebug("trans:%d, stage from undoAction to rollback", pTrans->id);
+    mDebug("trans:%d, stage from undoAction to finished", pTrans->id);
     continueExec = true;
   } else if (code == TSDB_CODE_ACTION_IN_PROGRESS) {
     mDebug("trans:%d, stage keep on undoAction since %s", pTrans->id, tstrerror(code));
@@ -1254,14 +1268,14 @@ static bool mndTransPerformUndoActionStage(SMnode *pMnode, STrans *pTrans) {
 }
 
 static bool mndTransPerformRollbackStage(SMnode *pMnode, STrans *pTrans) {
-  if (mndCantExecuteTransAction(pMnode)) return false;
+  if (mndCannotExecuteTransAction(pMnode)) return false;
 
   bool    continueExec = true;
   int32_t code = mndTransRollback(pMnode, pTrans);
 
   if (code == 0) {
-    pTrans->stage = TRN_STAGE_FINISHED;
+    pTrans->stage = TRN_STAGE_UNDO_ACTION;
-    mDebug("trans:%d, stage from rollback to finished", pTrans->id);
+    mDebug("trans:%d, stage from rollback to undoAction", pTrans->id);
     continueExec = true;
   } else {
     pTrans->failedTimes++;
@@ -1309,12 +1323,12 @@ void mndTransExecute(SMnode *pMnode, STrans *pTrans) {
       case TRN_STAGE_COMMIT_ACTION:
         continueExec = mndTransPerformCommitActionStage(pMnode, pTrans);
         break;
-      case TRN_STAGE_UNDO_ACTION:
-        continueExec = mndTransPerformUndoActionStage(pMnode, pTrans);
-        break;
       case TRN_STAGE_ROLLBACK:
         continueExec = mndTransPerformRollbackStage(pMnode, pTrans);
         break;
+      case TRN_STAGE_UNDO_ACTION:
+        continueExec = mndTransPerformUndoActionStage(pMnode, pTrans);
+        break;
       case TRN_STAGE_FINISHED:
         continueExec = mndTransPerfromFinishedStage(pMnode, pTrans);
         break;
@@ -1369,7 +1383,7 @@ static int32_t mndProcessKillTransReq(SRpcMsg *pReq) {
 
   mInfo("trans:%d, start to kill", killReq.transId);
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_KILL_TRANS) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_KILL_TRANS) != 0) {
     goto _OVER;
   }
 
@@ -1416,11 +1430,8 @@ void mndTransPullup(SMnode *pMnode) {
     mndReleaseTrans(pMnode, pTrans);
   }
 
-  SSnapshotMeta sMeta = {0};
+  // todo, set to SDB_WRITE_DELTA
-  if (syncGetSnapshotMeta(pMnode->syncMgmt.sync, &sMeta) == 0) {
+  sdbWriteFile(pMnode->pSdb, 0);
-    sdbSetCurConfig(pMnode->pSdb, sMeta.lastConfigIndex);
-  }
-  sdbWriteFile(pMnode->pSdb);
   taosArrayDestroy(pArray);
 }
 
@@ -1449,10 +1460,15 @@ static int32_t mndRetrieveTrans(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBl
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, (const char *)stage, false);
 
-    char dbname[TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
+    char dbname1[TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
-    STR_WITH_MAXSIZE_TO_VARSTR(dbname, mndGetDbStr(pTrans->dbname), pShow->pMeta->pSchemas[cols].bytes);
+    STR_WITH_MAXSIZE_TO_VARSTR(dbname1, mndGetDbStr(pTrans->dbname1), pShow->pMeta->pSchemas[cols].bytes);
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-    colDataAppend(pColInfo, numOfRows, (const char *)dbname, false);
+    colDataAppend(pColInfo, numOfRows, (const char *)dbname1, false);
+
+    char dbname2[TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
+    STR_WITH_MAXSIZE_TO_VARSTR(dbname2, mndGetDbStr(pTrans->dbname2), pShow->pMeta->pSchemas[cols].bytes);
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, numOfRows, (const char *)dbname2, false);
 
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, (const char *)&pTrans->failedTimes, false);

source/dnode/mnode/impl/src/mndUser.c

@@ -354,13 +354,13 @@ static int32_t mndProcessCreateUserReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  pOperUser = mndAcquireUser(pMnode, pReq->conn.user);
+  pOperUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pOperUser == NULL) {
     terrno = TSDB_CODE_MND_NO_USER_FROM_CONN;
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_CREATE_USER) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_CREATE_USER) != 0) {
     goto _OVER;
   }
 
@@ -460,7 +460,7 @@ static int32_t mndProcessAlterUserReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  pOperUser = mndAcquireUser(pMnode, pReq->conn.user);
+  pOperUser = mndAcquireUser(pMnode, pReq->info.conn.user);
   if (pOperUser == NULL) {
     terrno = TSDB_CODE_MND_NO_USER_FROM_CONN;
     goto _OVER;
@@ -643,7 +643,7 @@ static int32_t mndProcessDropUserReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_DROP_USER) != 0) {
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_DROP_USER) != 0) {
     goto _OVER;
   }
 

source/dnode/mnode/impl/src/mndVgroup.c

@@ -1058,7 +1058,7 @@ int32_t mndSetMoveVgroupsInfoToTrans(SMnode *pMnode, STrans *pTrans, int32_t del
 
 static int32_t mndAddIncVgroupReplicaToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SVgObj *pVgroup,
                                              int32_t newDnodeId) {
-  mDebug("vgId:%d, will add 1 vnode, replica:%d, dnode:%d", pVgroup->vgId, pVgroup->replica, newDnodeId);
+  mDebug("vgId:%d, will add 1 vnode, replica:%d dnode:%d", pVgroup->vgId, pVgroup->replica, newDnodeId);
 
   SVnodeGid *pGid = &pVgroup->vnodeGid[pVgroup->replica];
   pVgroup->replica++;
@@ -1074,7 +1074,7 @@ static int32_t mndAddIncVgroupReplicaToTrans(SMnode *pMnode, STrans *pTrans, SDb
 
 static int32_t mndAddDecVgroupReplicaFromTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SVgObj *pVgroup,
                                                int32_t delDnodeId) {
-  mDebug("vgId:%d, will remove 1 vnode, replica:%d, dnode:%d", pVgroup->vgId, pVgroup->replica, delDnodeId);
+  mDebug("vgId:%d, will remove 1 vnode, replica:%d dnode:%d", pVgroup->vgId, pVgroup->replica, delDnodeId);
 
   SVnodeGid *pGid = NULL;
   SVnodeGid  delGid = {0};
@@ -1116,7 +1116,8 @@ static int32_t mndRedistributeVgroup(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb,
   memcpy(&newVg, pVgroup, sizeof(SVgObj));
   mInfo("vgId:%d, vgroup info before redistribute, replica:%d", newVg.vgId, newVg.replica);
   for (int32_t i = 0; i < newVg.replica; ++i) {
-    mInfo("vgId:%d, vnode:%d dnode:%d", newVg.vgId, i, newVg.vnodeGid[i].dnodeId);
+    mInfo("vgId:%d, vnode:%d dnode:%d role:%s", newVg.vgId, i, newVg.vnodeGid[i].dnodeId,
+          syncStr(newVg.vnodeGid[i].role));
   }
 
   if (pNew1 != NULL && pOld1 != NULL) {
@@ -1198,9 +1199,9 @@ static int32_t mndProcessRedistributeVgroupMsg(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  mInfo("vgId:%d, start to redistribute to dnode %d:%d:%d", req.vgId, req.dnodeId1, req.dnodeId2, req.dnodeId3);
+  mInfo("vgId:%d, start to redistribute vgroup to dnode %d:%d:%d", req.vgId, req.dnodeId1, req.dnodeId2, req.dnodeId3);
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_REDISTRIBUTE_VGROUP) != 0) goto _OVER;
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_REDISTRIBUTE_VGROUP) != 0) goto _OVER;
 
   pVgroup = mndAcquireVgroup(pMnode, req.vgId);
   if (pVgroup == NULL) goto _OVER;
@@ -1500,7 +1501,7 @@ static int32_t mndProcessSplitVgroupMsg(SRpcMsg *pReq) {
   pDb = mndAcquireDb(pMnode, pVgroup->dbName);
   if (pDb == NULL) goto _OVER;
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_SPLIT_VGROUP) != 0) goto _OVER;
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_SPLIT_VGROUP) != 0) goto _OVER;
 
   code = mndSplitVgroup(pMnode, pReq, pDb, pVgroup);
   if (code == 0) code = TSDB_CODE_ACTION_IN_PROGRESS;
@@ -1624,7 +1625,7 @@ static int32_t mndProcessBalanceVgroupMsg(SRpcMsg *pReq) {
 
   mDebug("start to balance vgroup");
 
-  if (mndCheckOperAuth(pMnode, pReq->conn.user, MND_OPER_BALANCE_VGROUP) != 0) goto _OVER;
+  if (mndCheckOperAuth(pMnode, pReq->info.conn.user, MND_OPER_BALANCE_VGROUP) != 0) goto _OVER;
 
   while (1) {
     SDnodeObj *pDnode = NULL;

source/dnode/mnode/impl/test/profile/profile.cpp

@@ -295,8 +295,7 @@ TEST_F(MndTestProfile, 07_KillQueryMsg) {
 
 TEST_F(MndTestProfile, 08_KillQueryMsg_InvalidConn) {
   SKillQueryReq killReq = {0};
-  killReq.connId = 2345;
+  strcpy(killReq.queryStrId, "2345:2345");
-  killReq.queryId = 2345;
 
   int32_t contLen = tSerializeSKillQueryReq(NULL, 0, &killReq);
   void*   pReq = rpcMallocCont(contLen);

source/dnode/mnode/impl/test/sdb/sdbTest.cpp

@@ -493,8 +493,11 @@ TEST_F(MndTestSdb, 01_Write_Str) {
   ASSERT_EQ(sdbGetSize(pSdb, SDB_USER), 2);
   ASSERT_EQ(sdbGetMaxId(pSdb, SDB_USER), -1);
   ASSERT_EQ(sdbGetTableVer(pSdb, SDB_USER), 2);
-  sdbSetApplyIndex(pSdb, -1);
+  sdbSetApplyInfo(pSdb, -1, -1, -1);
-  ASSERT_EQ(sdbGetApplyIndex(pSdb), -1);
+  // int64_t index, config;
+  // int64_t term;
+  // sdbGetCommitInfo(pSdb, &index, &term, &config);
+  // ASSERT_EQ(index, -1);
   ASSERT_EQ(mnode.insertTimes, 2);
   ASSERT_EQ(mnode.deleteTimes, 0);
 
@@ -700,11 +703,12 @@ TEST_F(MndTestSdb, 01_Write_Str) {
   }
 
   // write version
-  sdbSetApplyIndex(pSdb, 0);
+  sdbSetApplyInfo(pSdb, 0, 0, 0);
-  sdbSetApplyIndex(pSdb, 1);
+  sdbSetApplyInfo(pSdb, 1, 0, 0);
-  ASSERT_EQ(sdbGetApplyIndex(pSdb), 1);
+  // sdbGetApplyInfo(pSdb, &index, &term, &config);
-  ASSERT_EQ(sdbWriteFile(pSdb), 0);
+  // ASSERT_EQ(index, 1);
-  ASSERT_EQ(sdbWriteFile(pSdb), 0);
+  ASSERT_EQ(sdbWriteFile(pSdb, 0), 0);
+  ASSERT_EQ(sdbWriteFile(pSdb, 0), 0);
 
   sdbCleanup(pSdb);
   ASSERT_EQ(mnode.insertTimes, 7);
@@ -772,7 +776,11 @@ TEST_F(MndTestSdb, 01_Read_Str) {
   ASSERT_EQ(sdbGetSize(pSdb, SDB_USER), 2);
   ASSERT_EQ(sdbGetMaxId(pSdb, SDB_USER), -1);
   ASSERT_EQ(sdbGetTableVer(pSdb, SDB_USER), 5);
-  ASSERT_EQ(sdbGetApplyIndex(pSdb), 1);
+
+  int64_t index, config;
+  int64_t term;
+  sdbGetCommitInfo(pSdb, &index, &term, &config);
+  ASSERT_EQ(index, 1);
   ASSERT_EQ(mnode.insertTimes, 4);
   ASSERT_EQ(mnode.deleteTimes, 0);
 

source/dnode/mnode/impl/test/trans/trans2.cpp

@@ -128,7 +128,7 @@ class MndTestTrans2 : public ::testing::Test {
     mndTransSetCb(pTrans, TRANS_START_FUNC_TEST, TRANS_STOP_FUNC_TEST, param, strlen(param) + 1);
 
     if (pDb != NULL) {
-      mndTransSetDbName(pTrans, pDb->name);
+      mndTransSetDbName(pTrans, pDb->name, NULL);
     }
 
     int32_t code = mndTransPrepare(pMnode, pTrans);
@@ -201,7 +201,7 @@ class MndTestTrans2 : public ::testing::Test {
     }
 
     if (pDb != NULL) {
-      mndTransSetDbName(pTrans, pDb->name);
+      mndTransSetDbName(pTrans, pDb->name, NULL);
     }
 
     int32_t code = mndTransPrepare(pMnode, pTrans);

source/dnode/mnode/sdb/inc/sdb.h

@@ -37,6 +37,8 @@ extern "C" {
 #define mTrace(...) { if (mDebugFlag & DEBUG_TRACE) { taosPrintLog("MND ", DEBUG_TRACE, mDebugFlag, __VA_ARGS__); }}
 // clang-format on
 
+#define SDB_WRITE_DELTA 100
+
 #define SDB_GET_VAL(pData, dataPos, val, pos, func, type) \
   {                                                       \
     if (func(pRaw, dataPos, val) != 0) {                  \
@@ -169,11 +171,12 @@ typedef struct SSdb {
   SWal          *pWal;
   char          *currDir;
   char          *tmpDir;
-  int64_t        lastCommitVer;
+  int64_t        commitIndex;
-  int64_t        lastCommitTerm;
+  int64_t        commitTerm;
-  int64_t        curVer;
+  int64_t        commitConfig;
-  int64_t        curTerm;
+  int64_t        applyIndex;
-  int64_t        curConfig;
+  int64_t        applyTerm;
+  int64_t        applyConfig;
   int64_t        tableVer[SDB_MAX];
   int64_t        maxId[SDB_MAX];
   EKeyType       keyTypes[SDB_MAX];
@@ -257,7 +260,7 @@ int32_t sdbReadFile(SSdb *pSdb);
  * @param pSdb The sdb object.
  * @return int32_t 0 for success, -1 for failure.
  */
-int32_t sdbWriteFile(SSdb *pSdb);
+int32_t sdbWriteFile(SSdb *pSdb, int32_t delta);
 
 /**
  * @brief Parse and write raw data to sdb, then free the pRaw object
@@ -361,14 +364,8 @@ int64_t sdbGetTableVer(SSdb *pSdb, ESdbType type);
  * @param index The update value of the apply index.
  * @return int32_t The current index of sdb
  */
-void    sdbSetApplyIndex(SSdb *pSdb, int64_t index);
+void sdbSetApplyInfo(SSdb *pSdb, int64_t index, int64_t term, int64_t config);
-void    sdbSetApplyTerm(SSdb *pSdb, int64_t term);
+void sdbGetCommitInfo(SSdb *pSdb, int64_t *index, int64_t *term, int64_t *config);
-void    sdbSetCurConfig(SSdb *pSdb, int64_t config);
-int64_t sdbGetApplyIndex(SSdb *pSdb);
-int64_t sdbGetApplyTerm(SSdb *pSdb);
-int64_t sdbGetCommitIndex(SSdb *pSdb);
-int64_t sdbGetCommitTerm(SSdb *pSdb);
-int64_t sdbGetCurConfig(SSdb *pSdb);
 
 SSdbRaw *sdbAllocRaw(ESdbType type, int8_t sver, int32_t dataLen);
 void     sdbFreeRaw(SSdbRaw *pRaw);