Merge branch '3.0' of https://github.com/taosdata/TDengine into refact/submit_req_marks

cmake/cmake.define

@@ -1,6 +1,7 @@
 cmake_minimum_required(VERSION 3.0)
 
 set(CMAKE_VERBOSE_MAKEFILE OFF)
+set(TD_BUILD_TAOSA_INTERNAL FALSE)
 
 #set output directory
 SET(LIBRARY_OUTPUT_PATH ${PROJECT_BINARY_DIR}/build/lib)
@@ -141,13 +142,13 @@ ELSE ()
         SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -msse4.2")
         SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -msse4.2")
     ENDIF()
-    IF (COMPILER_SUPPORT_FMA)
-        SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mfma")
-        SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mfma")
-    ENDIF()
 
     IF ("${SIMD_SUPPORT}" MATCHES "true")
-        IF (COMPILER_SUPPORT_AVX)
+       IF (COMPILER_SUPPORT_FMA)
+           SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mfma")
+           SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mfma")
+       ENDIF()
+       IF (COMPILER_SUPPORT_AVX)
             SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mavx")
             SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mavx")
         ENDIF()

cmake/cmake.install

@@ -21,7 +21,7 @@ IF (TD_LINUX)
 ELSEIF (TD_WINDOWS)
   SET(TD_MAKE_INSTALL_SH "${TD_SOURCE_DIR}/packaging/tools/make_install.bat")
   INSTALL(CODE "MESSAGE(\"make install script: ${TD_MAKE_INSTALL_SH}\")")
-  INSTALL(CODE "execute_process(COMMAND ${TD_MAKE_INSTALL_SH} :needAdmin ${TD_SOURCE_DIR} ${PROJECT_BINARY_DIR} Windows ${TD_VER_NUMBER})")
+  INSTALL(CODE "execute_process(COMMAND ${TD_MAKE_INSTALL_SH} :needAdmin ${TD_SOURCE_DIR} ${PROJECT_BINARY_DIR} Windows ${TD_VER_NUMBER} ${TD_BUILD_TAOSA_INTERNAL})")
 ELSEIF (TD_DARWIN)
   SET(TD_MAKE_INSTALL_SH "${TD_SOURCE_DIR}/packaging/tools/make_install.sh")
   INSTALL(CODE "MESSAGE(\"make install script: ${TD_MAKE_INSTALL_SH}\")")

cmake/cmake.version

@@ -2,7 +2,7 @@
 IF (DEFINED VERNUMBER)
   SET(TD_VER_NUMBER ${VERNUMBER})
 ELSE ()
-  SET(TD_VER_NUMBER "3.0.2.0")
+  SET(TD_VER_NUMBER "3.0.2.1")
 ENDIF ()
 
 IF (DEFINED VERCOMPATIBLE)

cmake/taosadapter_CMakeLists.txt.in

@@ -2,7 +2,7 @@
 # taosadapter
 ExternalProject_Add(taosadapter
         GIT_REPOSITORY https://github.com/taosdata/taosadapter.git
-        GIT_TAG f0c1753
+        GIT_TAG a2e9920
         SOURCE_DIR "${TD_SOURCE_DIR}/tools/taosadapter"
         BINARY_DIR ""
         #BUILD_IN_SOURCE TRUE

docs/en/12-taos-sql/02-database.md

@@ -30,6 +30,8 @@ database_option: {
   | WAL_LEVEL {1 | 2}
   | VGROUPS value
   | SINGLE_STABLE {0 | 1}
+  | TABLE_PREFIX value
+  | TABLE_SUFFIX value
   | WAL_RETENTION_PERIOD value
   | WAL_ROLL_PERIOD value
   | WAL_RETENTION_SIZE value
@@ -67,6 +69,8 @@ database_option: {
 - SINGLE_STABLE: specifies whether the database can contain more than one supertable.
   - 0: The database can contain multiple supertables.
   - 1: The database can contain only one supertable.
+- TABLE_PREFIX：The prefix length in the table name that is ignored when distributing table to vnode based on table name.
+- TABLE_SUFFIX：The suffix length in the table name that is ignored when distributing table to vnode based on table name.
 - WAL_RETENTION_PERIOD: specifies the time after which WAL files are deleted. This parameter is used for data subscription. Enter a time in seconds. The default value of single copy is 0. A value of 0 indicates that each WAL file is deleted immediately after its contents are written to disk. -1: WAL files are never deleted. The default value of multiple copy is 4 days.
 - WAL_RETENTION_SIZE: specifies the size at which WAL files are deleted. This parameter is used for data subscription. Enter a size in KB. The default value of single copy is 0. A value of 0 indicates that each WAL file is deleted immediately after its contents are written to disk. -1: WAL files are never deleted. The default value of multiple copy is -1.
 - WAL_ROLL_PERIOD: specifies the time after which WAL files are rotated. After this period elapses, a new WAL file is created. The default value of single copy is 0. A value of 0 indicates that a new WAL file is created only after the previous WAL file was written to disk. The default values of multiple copy is 1 day.

docs/en/12-taos-sql/20-keywords.md

@@ -17,6 +17,7 @@ The following list shows all reserved keywords:
 - ADD
 - AFTER
 - AGGREGATE
+- ALIVE
 - ALL
 - ALTER
 - ANALYZE

docs/en/12-taos-sql/24-show.md

@@ -178,76 +178,138 @@ SHOW TABLE DISTRIBUTED table_name;
 
 Shows how table data is distributed.
 
-Examples： show table distributed d0\G;   Display the block distribution of table `d0` in detailed format.
+Examples： Below is an example of this command to display the block distribution of table `d0` in detailed format.
 
+```sql
+show table distributed d0\G;   
+```
+
+<details>
+ <summary> Show Example </summary>
+ <pre><code>
 *************************** 1.row ***************************
 _block_dist: Total_Blocks=[5] Total_Size=[93.65 Kb] Average_size=[18.73 Kb] Compression_Ratio=[23.98 %]
 
 Total_Blocks :  Table `d0` contains total 5 blocks
+
 Total_Size:  The total size of all the data blocks in table `d0` is 93.65 KB 
+
 Average_size:  The average size of each block is 18.73 KB
+
 Compression_Ratio: The data compression rate is 23.98%
  
 *************************** 2.row ***************************
 _block_dist: Total_Rows=[20000] Inmem_Rows=[0] MinRows=[3616] MaxRows=[4096] Average_Rows=[4000]
 
 Total_Rows: Table `d0` contains 20,000 rows
+
 Inmem_Rows： The rows still in memory, i.e. not committed in disk, is 0, i.e. none such rows
+
 MinRows：  The minimum number of rows in a block is 3,616 
+
 MaxRows： The maximum number of rows in a block is 4,096B
+
 Average_Rows： The average number of rows in a block is 4,000
 
 *************************** 3.row ***************************
 _block_dist: Total_Tables=[1] Total_Files=[2]
 
 Total_Tables:  The number of child tables, 1 in this example
+
 Total_Files：   The number of files storing the table's data, 2 in this example
 
 *************************** 4.row ***************************
+
 _block_dist: --------------------------------------------------------------------------------
+
 *************************** 5.row ***************************
+
 _block_dist: 0100 |
+
 *************************** 6.row ***************************
+
 _block_dist: 0299 |
+
 *************************** 7.row ***************************
+
 _block_dist: 0498 |
+
 *************************** 8.row ***************************
+
 _block_dist: 0697 |
+
 *************************** 9.row ***************************
+
 _block_dist: 0896 |
+
 *************************** 10.row ***************************
+
 _block_dist: 1095 |
+
 *************************** 11.row ***************************
+
 _block_dist: 1294 |
+
 *************************** 12.row ***************************
+
 _block_dist: 1493 |
+
 *************************** 13.row ***************************
+
 _block_dist: 1692 |
+
 *************************** 14.row ***************************
+
 _block_dist: 1891 |
+
 *************************** 15.row ***************************
+
 _block_dist: 2090 |
+
 *************************** 16.row ***************************
+
 _block_dist: 2289 |
+
 *************************** 17.row ***************************
+
 _block_dist: 2488 |
+
 *************************** 18.row ***************************
+
 _block_dist: 2687 |
+
 *************************** 19.row ***************************
+
 _block_dist: 2886 |
+
 *************************** 20.row ***************************
+
 _block_dist: 3085 |
+
 *************************** 21.row ***************************
+
 _block_dist: 3284 |
+
 *************************** 22.row ***************************
+
 _block_dist: 3483 |||||||||||||||||  1 (20.00%)
+
 *************************** 23.row ***************************
+
 _block_dist: 3682 |
+
 *************************** 24.row ***************************
+
 _block_dist: 3881 |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||  4 (80.00%)
+
 Query OK, 24 row(s) in set (0.002444s)
 
-  The above show the block distribution percentage according to the number of rows in each block. In the above example, `_block_dist: 3483 |||||||||||||||||  1 (20.00%)` means there is one block whose rows is between 3,483 and 3,681. `_block_dist: 3881 |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||  4 (80.00%)` means there are 4 blocks whose rows is between 3,881 and 4,096. The number of blocks whose rows fall in other range is zero.
+</code></pre>
+</details>
+
+  The above show the block distribution percentage according to the number of rows in each block. In the above example, we can get below information:
+  - `_block_dist: 3483 |||||||||||||||||  1 (20.00%)` means there is one block whose rows is between 3,483 and 3,681.
+  -  `_block_dist: 3881 |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||  4 (80.00%)` means there are 4 blocks whose rows is between 3,881 and 4,096.   -  The number of blocks whose rows fall in other range is zero.
 
 ## SHOW TAGS
 

docs/en/14-reference/04-taosadapter.md

@@ -21,6 +21,7 @@ taosAdapter provides the following features.
 - Seamless connection to collectd
 - Seamless connection to StatsD
 - Supports Prometheus remote_read and remote_write
+- Get table's VGroup ID
 
 ## taosAdapter architecture diagram
 
@@ -178,6 +179,7 @@ See [example/config/taosadapter.toml](https://github.com/taosdata/taosadapter/bl
   node_export is an exporter for machine metrics. Please visit [https://github.com/prometheus/node_exporter](https://github.com/prometheus/node_exporter) for more information.
 - Support for Prometheus remote_read and remote_write
   remote_read and remote_write are interfaces for Prometheus data read and write from/to other data storage solution. Please visit [https://prometheus.io/blog/2019/10/10/remote-read-meets-streaming/#remote-apis](https://prometheus.io/blog/2019/10/10/remote-read-meets-streaming/#remote-apis) for more information.
+- Get table's VGroup ID. For more information about VGroup, please refer to [primary-logic-unit](/tdinternal/arch/#primary-logic-unit).
 
 ## Interfaces
 
@@ -199,7 +201,7 @@ Support InfluxDB query parameters as follows.
 - `precision` The time precision used by TDengine
 - `u` TDengine user name
 - `p` TDengine password
-- `ttl` The time to live of automatically created sub-table. This value cannot be updated. TDengine will use the ttl value of the frist data of sub-table to create sub-table. For more information, please refer [Create Table](/taos-sql/table/#create-table)
+- `ttl` The time to live of automatically created sub-table. This value cannot be updated. TDengine will use the ttl value of the first data of sub-table to create sub-table. For more information, please refer [Create Table](/taos-sql/table/#create-table)
 
 Note:      InfluxDB token authorization is not supported at present. Only Basic authorization and query parameter validation are supported.
 Example:   curl --request POST http://127.0.0.1:6041/influxdb/v1/write?db=test --user "root:taosdata" --data-binary "measurement,host=host1 field1=2i,field2=2.0 1577836800000000000"
@@ -241,6 +243,10 @@ node_export is an exporter of hardware and OS metrics exposed by the \*NIX kerne
 
 <Prometheus />
 
+### Get table's VGroup ID
+
+You can call `http://<fqdn>:6041/rest/vgid?db=<db>&table=<table>` to get table's VGroup ID. For more information about VGroup, please refer to [primary-logic-unit](/tdinternal/arch/#primary-logic-unit).
+
 ## Memory usage optimization methods
 
 taosAdapter will monitor its memory usage during operation and adjust it with two thresholds. Valid values are integers between 1 to 100, and represent a percentage of the system's physical memory.

docs/en/28-releases/01-tdengine.md

@@ -10,6 +10,10 @@ For TDengine 2.x installation packages by version, please visit [here](https://w
 
 import Release from "/components/ReleaseV3";
 
+## 3.0.2.2
+
+<Release type="tdengine" version="3.0.2.2" />
+
 ## 3.0.2.1
 
 <Release type="tdengine" version="3.0.2.1" />

docs/en/28-releases/02-tools.md

@@ -10,6 +10,10 @@ For other historical version installers, please visit [here](https://www.taosdat
 
 import Release from "/components/ReleaseV3";
 
+## 2.4.0
+
+<Release type="tools" version="2.4.0" />
+
 ## 2.3.3
 
 <Release type="tools" version="2.3.3" />

docs/zh/12-taos-sql/02-database.md

@@ -30,6 +30,8 @@ database_option: {
   | WAL_LEVEL {1 | 2}
   | VGROUPS value
   | SINGLE_STABLE {0 | 1}
+  | TABLE_PREFIX value
+  | TABLE_SUFFIX value
   | WAL_RETENTION_PERIOD value
   | WAL_ROLL_PERIOD value
   | WAL_RETENTION_SIZE value
@@ -67,6 +69,8 @@ database_option: {
 - SINGLE_STABLE：表示此数据库中是否只可以创建一个超级表，用于超级表列非常多的情况。
   - 0：表示可以创建多张超级表。
   - 1：表示只可以创建一张超级表。
+- TABLE_PREFIX：内部存储引擎根据表名分配存储该表数据的 VNODE 时要忽略的前缀的长度。
+- TABLE_SUFFIX：内部存储引擎根据表名分配存储该表数据的 VNODE 时要忽略的后缀的长度。
 - WAL_RETENTION_PERIOD：wal 文件的额外保留策略，用于数据订阅。wal 的保存时长，单位为 s。单副本默认为 0，即落盘后立即删除。-1 表示不删除。多副本默认为 4 天。
 - WAL_RETENTION_SIZE：wal 文件的额外保留策略，用于数据订阅。wal 的保存的最大上限，单位为 KB。单副本默认为 0，即落盘后立即删除。多副本默认为-1，表示不删除。
 - WAL_ROLL_PERIOD：wal 文件切换时长，单位为 s。当 wal 文件创建并写入后，经过该时间，会自动创建一个新的 wal 文件。单副本默认为 0，即仅在落盘时创建新文件。多副本默认为 1 天。

docs/zh/12-taos-sql/20-keywords.md

@@ -18,6 +18,7 @@ description: TDengine 保留关键字的详细列表
 - ADD
 - AFTER
 - AGGREGATE
+- ALIVE
 - ALL
 - ALTER
 - ANALYZE

docs/zh/12-taos-sql/24-show.md

@@ -183,6 +183,10 @@ SHOW TABLE DISTRIBUTED table_name;
 
 语句： show table distributed d0\G;   竖行显示表 d0 的 BLOCK 分布情况
 
+<details>
+ <summary>显示示例</summary>
+ <pre><code>
+
 *************************** 1.row ***************************
 
 _block_dist: Total_Blocks=[5] Total_Size=[93.65 Kb] Average_size=[18.73 Kb] Compression_Ratio=[23.98 %]
@@ -244,6 +248,8 @@ _block_dist: 3881 ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 
 Query OK, 24 row(s) in set (0.002444s)
 
+</code></pre>
+ </details>
 
   上面是块中包含数据行数的块儿分布情况图，这里的 0100 0299 0498 … 表示的是每个块中包含的数据行数，上面的意思就是这个表的 5 个块，分布在 3483 ~3681 行的块有 1 个，占整个块的 20%，分布在 3881 ~ 4096（最大行数）的块数为 4 个，占整个块的 80%， 其它区域内分布块数为 0。
 

docs/zh/12-taos-sql/29-changes.md

@@ -94,6 +94,7 @@ description: "TDengine 3.0 版本的语法变更说明"
 | 9 | SAMPLE | 增强 | 可以直接用于超级表了。没有PARTITION BY时，超级表的数据会被合并成一条时间线。
 | 10 | STATECOUNT | 增强 | 可以直接用于超级表了。没有PARTITION BY时，超级表的数据会被合并成一条时间线。
 | 11 | STATEDURATION | 增强 | 可以直接用于超级表了。没有PARTITION BY时，超级表的数据会被合并成一条时间线。
+| 12 | TIMETRUNCATE | 增强 | 增加ignore_timezone参数，可选是否使用，默认值为1.
 
 
 ## SCHEMALESS 变更

docs/zh/14-reference/04-taosadapter.md

@@ -21,6 +21,7 @@ taosAdapter 提供以下功能：
 - 无缝连接到 collectd
 - 无缝连接到 StatsD
 - 支持 Prometheus remote_read 和 remote_write
+- 获取 table 所在的虚拟节点组（VGroup）的 VGroup ID
 
 ## taosAdapter 架构图
 
@@ -178,6 +179,7 @@ AllowWebSockets
   node_export 是一个机器指标的导出器。请访问 [https://github.com/prometheus/node_exporter](https://github.com/prometheus/node_exporter) 了解更多信息。
 - 支持 Prometheus remote_read 和 remote_write
   remote_read 和 remote_write 是 Prometheus 数据读写分离的集群方案。请访问[https://prometheus.io/blog/2019/10/10/remote-read-meets-streaming/#remote-apis](https://prometheus.io/blog/2019/10/10/remote-read-meets-streaming/#remote-apis) 了解更多信息。
+- 获取 table 所在的虚拟节点组（VGroup）的 VGroup ID。关于虚拟节点组（VGroup）的更多信息，请访问[整体架构文档](/tdinternal/arch/#主要逻辑单元) 。
 
 ## 接口
 
@@ -240,6 +242,10 @@ Prometheus 使用的由 \*NIX 内核暴露的硬件和操作系统指标的输
 
 <Prometheus />
 
+### 获取 table 的 VGroup ID
+
+可以访问 http 接口 `http://<fqdn>:6041/rest/vgid?db=<db>&table=<table>` 获取 table 的 VGroup ID。关于虚拟节点组（VGroup）的更多信息，请访问[整体架构文档](/tdinternal/arch/#主要逻辑单元) 。
+
 ## 内存使用优化方法
 
 taosAdapter 将监测自身运行过程中内存使用率并通过两个阈值进行调节。有效值范围为 -1 到 100 的整数，单位为系统物理内存的百分比。
@@ -282,7 +288,7 @@ http 返回内容：
 
 ## taosAdapter 监控指标
 
-taosAdapter 采集 http 相关指标、cpu 百分比和内存百分比。
+taosAdapter 采集 http 相关指标、CPU 百分比和内存百分比。
 
 ### http 接口
 
@@ -294,13 +300,13 @@ http://<fqdn>:6041/metrics
 
 ### 写入 TDengine
 
-taosAdapter 支持将 http 监控、cpu 百分比和内存百分比写入 TDengine。
+taosAdapter 支持将 http 监控、CPU 百分比和内存百分比写入 TDengine。
 
 有关配置参数
 
 | **配置项**                 | **描述**                                     | **默认值**  |
 |-------------------------|--------------------------------------------|----------|
-| monitor.collectDuration | cpu 和内存采集间隔                                | 3s       |
+| monitor.collectDuration | CPU 和内存采集间隔                                | 3s       |
 | monitor.identity        | 当前taosadapter 的标识符如果不设置将使用 'hostname:port' |          |
 | monitor.incgroup        | 是否是 cgroup 中运行(容器中运行设置为 true)              | false    |
 | monitor.writeToTD       | 是否写入到 TDengine                             | false    |

docs/zh/28-releases/01-tdengine.md

@@ -10,6 +10,10 @@ TDengine 2.x 各版本安装包请访问[这里](https://www.taosdata.com/all-do
 
 import Release from "/components/ReleaseV3";
 
+## 3.0.2.2
+
+<Release type="tdengine" version="3.0.2.2" />
+
 ## 3.0.2.1
 
 <Release type="tdengine" version="3.0.2.1" />

docs/zh/28-releases/02-tools.md

@@ -10,6 +10,10 @@ taosTools 各版本安装包下载链接如下：
 
 import Release from "/components/ReleaseV3";
 
+## 2.4.0
+
+<Release type="tools" version="2.4.0" />
+
 ## 2.3.3
 
 <Release type="tools" version="2.3.3" />

include/common/taosdef.h

@@ -30,6 +30,11 @@ typedef int64_t tb_uid_t;
 #define IS_TSWINDOW_SPECIFIED(win) (((win).skey != INT64_MIN) || ((win).ekey != INT64_MAX))
 #define TSWINDOW_IS_EQUAL(t1, t2)  (((t1).skey == (t2).skey) && ((t1).ekey == (t2).ekey))
 
+//define show cluster alive and show db.alive
+#define SHOW_STATUS_NOT_AVAILABLE  0
+#define SHOW_STATUS_AVAILABLE      1
+#define SHOW_STATUS_HALF_AVAILABLE 2
+
 typedef enum {
   TSDB_SUPER_TABLE = 1,   // super table
   TSDB_CHILD_TABLE = 2,   // table created from super table

include/common/tcommon.h

@@ -292,6 +292,7 @@ typedef struct STableBlockDistInfo {
   uint16_t numOfFiles;
   uint32_t numOfTables;
   uint32_t numOfBlocks;
+  uint32_t numOfVgroups;
   uint64_t totalSize;
   uint64_t totalRows;
   int32_t  maxRows;

include/common/tglobal.h

@@ -55,7 +55,7 @@ extern int32_t tsNumOfMnodeQueryThreads;
 extern int32_t tsNumOfMnodeFetchThreads;
 extern int32_t tsNumOfMnodeReadThreads;
 extern int32_t tsNumOfVnodeQueryThreads;
-extern int32_t tsNumOfVnodeStreamThreads;
+extern float   tsRatioOfVnodeStreamThreads;
 extern int32_t tsNumOfVnodeFetchThreads;
 extern int32_t tsNumOfVnodeRsmaThreads;
 extern int32_t tsNumOfQnodeQueryThreads;

include/common/tmsg.h

@@ -1751,6 +1751,8 @@ typedef struct {
 #define STREAM_FILL_HISTORY_ON        1
 #define STREAM_FILL_HISTORY_OFF       0
 #define STREAM_DEFAULT_FILL_HISTORY   STREAM_FILL_HISTORY_OFF
+#define STREAM_CREATE_STABLE_TRUE     1
+#define STREAM_CREATE_STABLE_FALSE    0
 
 typedef struct {
   char    name[TSDB_STREAM_FNAME_LEN];
@@ -1768,6 +1770,8 @@ typedef struct {
   SArray* pTags;  // array of SField
   // 3.0.20
   int64_t checkpointFreq;  // ms
+  // 3.0.2.3
+  int8_t createStb;
 } SCMCreateStreamReq;
 
 typedef struct {

include/common/ttokendef.h

@@ -175,171 +175,172 @@
 #define TK_CONSUMERS                      157
 #define TK_SUBSCRIPTIONS                  158
 #define TK_VNODES                         159
-#define TK_LIKE                           160
+#define TK_ALIVE                          160
-#define TK_TBNAME                         161
+#define TK_LIKE                           161
-#define TK_QTAGS                          162
+#define TK_TBNAME                         162
-#define TK_AS                             163
+#define TK_QTAGS                          163
-#define TK_INDEX                          164
+#define TK_AS                             164
-#define TK_FUNCTION                       165
+#define TK_INDEX                          165
-#define TK_INTERVAL                       166
+#define TK_FUNCTION                       166
-#define TK_COUNT                          167
+#define TK_INTERVAL                       167
-#define TK_LAST_ROW                       168
+#define TK_COUNT                          168
-#define TK_TOPIC                          169
+#define TK_LAST_ROW                       169
-#define TK_WITH                           170
+#define TK_TOPIC                          170
-#define TK_META                           171
+#define TK_WITH                           171
-#define TK_CONSUMER                       172
+#define TK_META                           172
-#define TK_GROUP                          173
+#define TK_CONSUMER                       173
-#define TK_DESC                           174
+#define TK_GROUP                          174
-#define TK_DESCRIBE                       175
+#define TK_DESC                           175
-#define TK_RESET                          176
+#define TK_DESCRIBE                       176
-#define TK_QUERY                          177
+#define TK_RESET                          177
-#define TK_CACHE                          178
+#define TK_QUERY                          178
-#define TK_EXPLAIN                        179
+#define TK_CACHE                          179
-#define TK_ANALYZE                        180
+#define TK_EXPLAIN                        180
-#define TK_VERBOSE                        181
+#define TK_ANALYZE                        181
-#define TK_NK_BOOL                        182
+#define TK_VERBOSE                        182
-#define TK_RATIO                          183
+#define TK_NK_BOOL                        183
-#define TK_NK_FLOAT                       184
+#define TK_RATIO                          184
-#define TK_OUTPUTTYPE                     185
+#define TK_NK_FLOAT                       185
-#define TK_AGGREGATE                      186
+#define TK_OUTPUTTYPE                     186
-#define TK_BUFSIZE                        187
+#define TK_AGGREGATE                      187
-#define TK_STREAM                         188
+#define TK_BUFSIZE                        188
-#define TK_INTO                           189
+#define TK_STREAM                         189
-#define TK_TRIGGER                        190
+#define TK_INTO                           190
-#define TK_AT_ONCE                        191
+#define TK_TRIGGER                        191
-#define TK_WINDOW_CLOSE                   192
+#define TK_AT_ONCE                        192
-#define TK_IGNORE                         193
+#define TK_WINDOW_CLOSE                   193
-#define TK_EXPIRED                        194
+#define TK_IGNORE                         194
-#define TK_FILL_HISTORY                   195
+#define TK_EXPIRED                        195
-#define TK_SUBTABLE                       196
+#define TK_FILL_HISTORY                   196
-#define TK_KILL                           197
+#define TK_SUBTABLE                       197
-#define TK_CONNECTION                     198
+#define TK_KILL                           198
-#define TK_TRANSACTION                    199
+#define TK_CONNECTION                     199
-#define TK_BALANCE                        200
+#define TK_TRANSACTION                    200
-#define TK_VGROUP                         201
+#define TK_BALANCE                        201
-#define TK_MERGE                          202
+#define TK_VGROUP                         202
-#define TK_REDISTRIBUTE                   203
+#define TK_MERGE                          203
-#define TK_SPLIT                          204
+#define TK_REDISTRIBUTE                   204
-#define TK_DELETE                         205
+#define TK_SPLIT                          205
-#define TK_INSERT                         206
+#define TK_DELETE                         206
-#define TK_NULL                           207
+#define TK_INSERT                         207
-#define TK_NK_QUESTION                    208
+#define TK_NULL                           208
-#define TK_NK_ARROW                       209
+#define TK_NK_QUESTION                    209
-#define TK_ROWTS                          210
+#define TK_NK_ARROW                       210
-#define TK_QSTART                         211
+#define TK_ROWTS                          211
-#define TK_QEND                           212
+#define TK_QSTART                         212
-#define TK_QDURATION                      213
+#define TK_QEND                           213
-#define TK_WSTART                         214
+#define TK_QDURATION                      214
-#define TK_WEND                           215
+#define TK_WSTART                         215
-#define TK_WDURATION                      216
+#define TK_WEND                           216
-#define TK_IROWTS                         217
+#define TK_WDURATION                      217
-#define TK_ISFILLED                       218
+#define TK_IROWTS                         218
-#define TK_CAST                           219
+#define TK_ISFILLED                       219
-#define TK_NOW                            220
+#define TK_CAST                           220
-#define TK_TODAY                          221
+#define TK_NOW                            221
-#define TK_TIMEZONE                       222
+#define TK_TODAY                          222
-#define TK_CLIENT_VERSION                 223
+#define TK_TIMEZONE                       223
-#define TK_SERVER_VERSION                 224
+#define TK_CLIENT_VERSION                 224
-#define TK_SERVER_STATUS                  225
+#define TK_SERVER_VERSION                 225
-#define TK_CURRENT_USER                   226
+#define TK_SERVER_STATUS                  226
-#define TK_CASE                           227
+#define TK_CURRENT_USER                   227
-#define TK_END                            228
+#define TK_CASE                           228
-#define TK_WHEN                           229
+#define TK_END                            229
-#define TK_THEN                           230
+#define TK_WHEN                           230
-#define TK_ELSE                           231
+#define TK_THEN                           231
-#define TK_BETWEEN                        232
+#define TK_ELSE                           232
-#define TK_IS                             233
+#define TK_BETWEEN                        233
-#define TK_NK_LT                          234
+#define TK_IS                             234
-#define TK_NK_GT                          235
+#define TK_NK_LT                          235
-#define TK_NK_LE                          236
+#define TK_NK_GT                          236
-#define TK_NK_GE                          237
+#define TK_NK_LE                          237
-#define TK_NK_NE                          238
+#define TK_NK_GE                          238
-#define TK_MATCH                          239
+#define TK_NK_NE                          239
-#define TK_NMATCH                         240
+#define TK_MATCH                          240
-#define TK_CONTAINS                       241
+#define TK_NMATCH                         241
-#define TK_IN                             242
+#define TK_CONTAINS                       242
-#define TK_JOIN                           243
+#define TK_IN                             243
-#define TK_INNER                          244
+#define TK_JOIN                           244
-#define TK_SELECT                         245
+#define TK_INNER                          245
-#define TK_DISTINCT                       246
+#define TK_SELECT                         246
-#define TK_WHERE                          247
+#define TK_DISTINCT                       247
-#define TK_PARTITION                      248
+#define TK_WHERE                          248
-#define TK_BY                             249
+#define TK_PARTITION                      249
-#define TK_SESSION                        250
+#define TK_BY                             250
-#define TK_STATE_WINDOW                   251
+#define TK_SESSION                        251
-#define TK_EVENT_WINDOW                   252
+#define TK_STATE_WINDOW                   252
-#define TK_START                          253
+#define TK_EVENT_WINDOW                   253
-#define TK_SLIDING                        254
+#define TK_START                          254
-#define TK_FILL                           255
+#define TK_SLIDING                        255
-#define TK_VALUE                          256
+#define TK_FILL                           256
-#define TK_NONE                           257
+#define TK_VALUE                          257
-#define TK_PREV                           258
+#define TK_NONE                           258
-#define TK_LINEAR                         259
+#define TK_PREV                           259
-#define TK_NEXT                           260
+#define TK_LINEAR                         260
-#define TK_HAVING                         261
+#define TK_NEXT                           261
-#define TK_RANGE                          262
+#define TK_HAVING                         262
-#define TK_EVERY                          263
+#define TK_RANGE                          263
-#define TK_ORDER                          264
+#define TK_EVERY                          264
-#define TK_SLIMIT                         265
+#define TK_ORDER                          265
-#define TK_SOFFSET                        266
+#define TK_SLIMIT                         266
-#define TK_LIMIT                          267
+#define TK_SOFFSET                        267
-#define TK_OFFSET                         268
+#define TK_LIMIT                          268
-#define TK_ASC                            269
+#define TK_OFFSET                         269
-#define TK_NULLS                          270
+#define TK_ASC                            270
-#define TK_ABORT                          271
+#define TK_NULLS                          271
-#define TK_AFTER                          272
+#define TK_ABORT                          272
-#define TK_ATTACH                         273
+#define TK_AFTER                          273
-#define TK_BEFORE                         274
+#define TK_ATTACH                         274
-#define TK_BEGIN                          275
+#define TK_BEFORE                         275
-#define TK_BITAND                         276
+#define TK_BEGIN                          276
-#define TK_BITNOT                         277
+#define TK_BITAND                         277
-#define TK_BITOR                          278
+#define TK_BITNOT                         278
-#define TK_BLOCKS                         279
+#define TK_BITOR                          279
-#define TK_CHANGE                         280
+#define TK_BLOCKS                         280
-#define TK_COMMA                          281
+#define TK_CHANGE                         281
-#define TK_COMPACT                        282
+#define TK_COMMA                          282
-#define TK_CONCAT                         283
+#define TK_COMPACT                        283
-#define TK_CONFLICT                       284
+#define TK_CONCAT                         284
-#define TK_COPY                           285
+#define TK_CONFLICT                       285
-#define TK_DEFERRED                       286
+#define TK_COPY                           286
-#define TK_DELIMITERS                     287
+#define TK_DEFERRED                       287
-#define TK_DETACH                         288
+#define TK_DELIMITERS                     288
-#define TK_DIVIDE                         289
+#define TK_DETACH                         289
-#define TK_DOT                            290
+#define TK_DIVIDE                         290
-#define TK_EACH                           291
+#define TK_DOT                            291
-#define TK_FAIL                           292
+#define TK_EACH                           292
-#define TK_FILE                           293
+#define TK_FAIL                           293
-#define TK_FOR                            294
+#define TK_FILE                           294
-#define TK_GLOB                           295
+#define TK_FOR                            295
-#define TK_ID                             296
+#define TK_GLOB                           296
-#define TK_IMMEDIATE                      297
+#define TK_ID                             297
-#define TK_IMPORT                         298
+#define TK_IMMEDIATE                      298
-#define TK_INITIALLY                      299
+#define TK_IMPORT                         299
-#define TK_INSTEAD                        300
+#define TK_INITIALLY                      300
-#define TK_ISNULL                         301
+#define TK_INSTEAD                        301
-#define TK_KEY                            302
+#define TK_ISNULL                         302
-#define TK_MODULES                        303
+#define TK_KEY                            303
-#define TK_NK_BITNOT                      304
+#define TK_MODULES                        304
-#define TK_NK_SEMI                        305
+#define TK_NK_BITNOT                      305
-#define TK_NOTNULL                        306
+#define TK_NK_SEMI                        306
-#define TK_OF                             307
+#define TK_NOTNULL                        307
-#define TK_PLUS                           308
+#define TK_OF                             308
-#define TK_PRIVILEGE                      309
+#define TK_PLUS                           309
-#define TK_RAISE                          310
+#define TK_PRIVILEGE                      310
-#define TK_REPLACE                        311
+#define TK_RAISE                          311
-#define TK_RESTRICT                       312
+#define TK_REPLACE                        312
-#define TK_ROW                            313
+#define TK_RESTRICT                       313
-#define TK_SEMI                           314
+#define TK_ROW                            314
-#define TK_STAR                           315
+#define TK_SEMI                           315
-#define TK_STATEMENT                      316
+#define TK_STAR                           316
-#define TK_STRICT                         317
+#define TK_STATEMENT                      317
-#define TK_STRING                         318
+#define TK_STRICT                         318
-#define TK_TIMES                          319
+#define TK_STRING                         319
-#define TK_UPDATE                         320
+#define TK_TIMES                          320
-#define TK_VALUES                         321
+#define TK_UPDATE                         321
-#define TK_VARIABLE                       322
+#define TK_VALUES                         322
-#define TK_VIEW                           323
+#define TK_VARIABLE                       323
-#define TK_WAL                            324
+#define TK_VIEW                           324
+#define TK_WAL                            325
 
 #define TK_NK_SPACE   600
 #define TK_NK_COMMENT 601

include/libs/catalog/catalog.h

@@ -44,6 +44,7 @@ typedef enum {
   AUTH_TYPE_READ = 1,
   AUTH_TYPE_WRITE,
   AUTH_TYPE_OTHER,
+  AUTH_TYPE_READ_OR_WRITE,
 } AUTH_TYPE;
 
 typedef struct SUserAuthInfo {

include/libs/command/command.h

@@ -22,7 +22,7 @@
 
 typedef struct SExplainCtx SExplainCtx;
 
-int32_t qExecCommand(bool sysInfoUser, SNode *pStmt, SRetrieveTableRsp **pRsp);
+int32_t qExecCommand(int64_t* pConnId, bool sysInfoUser, SNode *pStmt, SRetrieveTableRsp **pRsp);
 
 int32_t qExecStaticExplain(SQueryPlan *pDag, SRetrieveTableRsp **pRsp);
 int32_t qExecExplainBegin(SQueryPlan *pDag, SExplainCtx **pCtx, int64_t startTs);

include/libs/executor/executor.h

@@ -215,6 +215,7 @@ int32_t qStreamSourceRecoverStep1(qTaskInfo_t tinfo, int64_t ver);
 int32_t qStreamSourceRecoverStep2(qTaskInfo_t tinfo, int64_t ver);
 int32_t qStreamRecoverFinish(qTaskInfo_t tinfo);
 int32_t qStreamRestoreParam(qTaskInfo_t tinfo);
+bool    qStreamRecoverScanFinished(qTaskInfo_t tinfo);
 
 #ifdef __cplusplus
 }

include/libs/nodes/cmdnodes.h

@@ -40,6 +40,7 @@ extern "C" {
 #define SHOW_LOCAL_VARIABLES_RESULT_FIELD1_LEN (TSDB_CONFIG_OPTION_LEN + VARSTR_HEADER_SIZE)
 #define SHOW_LOCAL_VARIABLES_RESULT_FIELD2_LEN (TSDB_CONFIG_VALUE_LEN + VARSTR_HEADER_SIZE)
 
+#define SHOW_ALIVE_RESULT_COLS 1
 #define PRIVILEGE_TYPE_MASK(n) (1 << n)
 
 #define PRIVILEGE_TYPE_ALL       PRIVILEGE_TYPE_MASK(0)
@@ -262,6 +263,11 @@ typedef struct SShowCreateDatabaseStmt {
   void*     pCfg;  // SDbCfgInfo
 } SShowCreateDatabaseStmt;
 
+typedef struct SShowAliveStmt {
+  ENodeType type;
+  char      dbName[TSDB_DB_NAME_LEN];
+} SShowAliveStmt;
+
 typedef struct SShowCreateTableStmt {
   ENodeType type;
   char      dbName[TSDB_DB_NAME_LEN];
@@ -295,7 +301,7 @@ typedef struct SShowTableTagsStmt {
   SNodeList* pTags;
 } SShowTableTagsStmt;
 
-typedef enum EIndexType { INDEX_TYPE_SMA = 1, INDEX_TYPE_FULLTEXT } EIndexType;
+typedef enum EIndexType { INDEX_TYPE_SMA = 1, INDEX_TYPE_FULLTEXT, INDEX_TYPE_NORMAL } EIndexType;
 
 typedef struct SIndexOptions {
   ENodeType  type;
@@ -401,6 +407,7 @@ typedef struct SCreateStreamStmt {
   SNode*          pQuery;
   SNodeList*      pTags;
   SNode*          pSubtable;
+  SNodeList*      pCols;
 } SCreateStreamStmt;
 
 typedef struct SDropStreamStmt {

include/libs/nodes/nodes.h

@@ -208,6 +208,8 @@ typedef enum ENodeType {
   QUERY_NODE_DELETE_STMT,
   QUERY_NODE_INSERT_STMT,
   QUERY_NODE_QUERY,
+  QUERY_NODE_SHOW_DB_ALIVE_STMT,
+  QUERY_NODE_SHOW_CLUSTER_ALIVE_STMT,  
 
   // logic plan node
   QUERY_NODE_LOGIC_PLAN_SCAN = 1000,

include/libs/nodes/querynodes.h

@@ -51,6 +51,7 @@ typedef struct SExprNode {
   char      userAlias[TSDB_COL_NAME_LEN];
   SArray*   pAssociation;
   bool      orderAlias;
+  bool      asAlias;
 } SExprNode;
 
 typedef enum EColumnType {
@@ -127,8 +128,7 @@ typedef struct SLogicConditionNode {
 } SLogicConditionNode;
 
 typedef struct SNodeListNode {
-  ENodeType  type;  // QUERY_NODE_NODE_LIST
+  SExprNode  node;  // QUERY_NODE_NODE_LIST
-  SDataType  dataType;
   SNodeList* pNodeList;
 } SNodeListNode;
 

include/libs/qcom/query.h

@@ -286,7 +286,7 @@ extern int32_t (*queryProcessMsgRsp[TDMT_MAX])(void* output, char* msg, int32_t
   ((_code) == TSDB_CODE_SYN_NOT_LEADER || (_code) == TSDB_CODE_SYN_RESTORING || (_code) == TSDB_CODE_SYN_INTERNAL_ERROR)
 #define SYNC_OTHER_LEADER_REDIRECT_ERROR(_code) ((_code) == TSDB_CODE_MNODE_NOT_FOUND)
 
-#define NO_RET_REDIRECT_ERROR(_code) ((_code) == TSDB_CODE_RPC_BROKEN_LINK || (_code) == TSDB_CODE_RPC_NETWORK_UNAVAIL)
+#define NO_RET_REDIRECT_ERROR(_code) ((_code) == TSDB_CODE_RPC_BROKEN_LINK || (_code) == TSDB_CODE_RPC_NETWORK_UNAVAIL || (_code) == TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED)
 
 #define NEED_REDIRECT_ERROR(_code)                                              \
   (NO_RET_REDIRECT_ERROR(_code) || SYNC_UNKNOWN_LEADER_REDIRECT_ERROR(_code) || \

include/libs/stream/tstream.h

@@ -409,21 +409,6 @@ static FORCE_INLINE void streamTaskInputFail(SStreamTask* pTask) {
   atomic_store_8(&pTask->inputStatus, TASK_INPUT_STATUS__FAILED);
 }
 
-static FORCE_INLINE int32_t streamTaskOutput(SStreamTask* pTask, SStreamDataBlock* pBlock) {
-  if (pTask->outputType == TASK_OUTPUT__TABLE) {
-    pTask->tbSink.tbSinkFunc(pTask, pTask->tbSink.vnode, 0, pBlock->blocks);
-    taosArrayDestroyEx(pBlock->blocks, (FDelete)blockDataFreeRes);
-    taosFreeQitem(pBlock);
-  } else if (pTask->outputType == TASK_OUTPUT__SMA) {
-    pTask->smaSink.smaSink(pTask->smaSink.vnode, pTask->smaSink.smaId, pBlock->blocks);
-    taosArrayDestroyEx(pBlock->blocks, (FDelete)blockDataFreeRes);
-    taosFreeQitem(pBlock);
-  } else {
-    taosWriteQitem(pTask->outputQueue->queue, pBlock);
-  }
-  return 0;
-}
-
 typedef struct {
   SMsgHead head;
   int64_t  streamId;
@@ -601,6 +586,7 @@ int32_t streamProcessRetrieveRsp(SStreamTask* pTask, SStreamRetrieveRsp* pRsp);
 
 int32_t streamTryExec(SStreamTask* pTask);
 int32_t streamSchedExec(SStreamTask* pTask);
+int32_t streamTaskOutput(SStreamTask* pTask, SStreamDataBlock* pBlock);
 
 int32_t streamScanExec(SStreamTask* pTask, int32_t batchSz);
 

include/libs/sync/sync.h

@@ -153,10 +153,10 @@ typedef struct SSyncFSM {
   void (*FpBecomeFollowerCb)(const struct SSyncFSM* pFsm);
 
   int32_t (*FpGetSnapshot)(const struct SSyncFSM* pFsm, SSnapshot* pSnapshot, void* pReaderParam, void** ppReader);
-  int32_t (*FpGetSnapshotInfo)(const struct SSyncFSM* pFsm, SSnapshot* pSnapshot);
+  void (*FpGetSnapshotInfo)(const struct SSyncFSM* pFsm, SSnapshot* pSnapshot);
 
   int32_t (*FpSnapshotStartRead)(const struct SSyncFSM* pFsm, void* pReaderParam, void** ppReader);
-  int32_t (*FpSnapshotStopRead)(const struct SSyncFSM* pFsm, void* pReader);
+  void (*FpSnapshotStopRead)(const struct SSyncFSM* pFsm, void* pReader);
   int32_t (*FpSnapshotDoRead)(const struct SSyncFSM* pFsm, void* pReader, void** ppBuf, int32_t* len);
 
   int32_t (*FpSnapshotStartWrite)(const struct SSyncFSM* pFsm, void* pWriterParam, void** ppWriter);

include/os/osMemory.h

@@ -22,12 +22,16 @@ 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 sectio
+#if !defined(WINDOWS)
+
 #ifndef ALLOW_FORBID_FUNC
 #define malloc  MALLOC_FUNC_TAOS_FORBID
 #define calloc  CALLOC_FUNC_TAOS_FORBID
 #define realloc REALLOC_FUNC_TAOS_FORBID
 #define free    FREE_FUNC_TAOS_FORBID
-#endif
+#endif // ifndef ALLOW_FORBID_FUNC
+
+#endif // if !defined(WINDOWS)
 
 void   *taosMemoryMalloc(int64_t size);
 void   *taosMemoryCalloc(int64_t num, int64_t size);

include/os/osSysinfo.h

@@ -70,6 +70,7 @@ typedef struct {
 
 SysNameInfo taosGetSysNameInfo();
 bool        taosCheckCurrentInDll();
+int         taosGetlocalhostname(char *hostname, size_t maxLen);
 
 #ifdef __cplusplus
 }

include/os/osSystem.h

@@ -62,11 +62,38 @@ void    taosResetTerminalMode();
     taosMemoryFree(strings);                                               \
   }
 #else
+#include <windows.h>
+#include <dbghelp.h>
+
+#define STACKSIZE 64
 #define taosPrintTrace(flags, level, dflag)                                                                \
   {                                                                                                        \
-    taosPrintLog(flags, level, dflag,                                                                      \
+    unsigned int   i;                                                                                                 \
-                 "backtrace not implemented on windows, so detailed stack information cannot be printed"); \
+    void*          stack[STACKSIZE];                                                                                  \
-    }
+    unsigned short frames;                                                                                            \
+    SYMBOL_INFO*   symbol;                                                                                            \
+    HANDLE         process;                                                                                           \
+                                                                                                                      \
+    process = GetCurrentProcess();                                                                                    \
+                                                                                                                      \
+    SymInitialize(process, NULL, TRUE);                                                                               \
+                                                                                                                      \
+    frames = CaptureStackBackTrace(0, STACKSIZE, stack, NULL);                                                        \
+    symbol = (SYMBOL_INFO*)calloc(sizeof(SYMBOL_INFO) + 256 * sizeof(char), 1);                                       \
+    if (symbol != NULL) {                                                                                             \
+        symbol->MaxNameLen = 255;                                                                                     \
+        symbol->SizeOfStruct = sizeof(SYMBOL_INFO);                                                                   \
+                                                                                                                      \
+        if (frames > 0) {                                                                                             \
+          taosPrintLog(flags, level, dflag, "obtained %d stack frames", frames);                                      \
+          for (i = 0; i < frames; i++) {                                                                              \
+            SymFromAddr(process, (DWORD64)(stack[i]), 0, symbol);                                                     \
+            taosPrintLog(flags, level, dflag, "frame:%i: %s - 0x%0X", frames - i - 1, symbol->Name, symbol->Address); \
+          }                                                                                                           \
+        }                                                                                                             \
+        free(symbol);                                                                                                 \
+    }                                                                                                                 \
+  }
 #endif
 
 #ifdef __cplusplus

include/util/taoserror.h

@@ -65,6 +65,8 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_RPC_PORT_EADDRINUSE           TAOS_DEF_ERROR_CODE(0, 0x0017) //
 #define TSDB_CODE_RPC_BROKEN_LINK               TAOS_DEF_ERROR_CODE(0, 0x0018) //
 #define TSDB_CODE_RPC_TIMEOUT                   TAOS_DEF_ERROR_CODE(0, 0x0019) //
+#define TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED    TAOS_DEF_ERROR_CODE(0, 0x0020) // "Vgroup could not be connected"
+#define TSDB_CODE_RPC_SOMENODE_BROKEN_LINK      TAOS_DEF_ERROR_CODE(0, 0x0021) // 
 
 //common & util
 #define TSDB_CODE_OPS_NOT_SUPPORT               TAOS_DEF_ERROR_CODE(0, 0x0100) //
@@ -204,9 +206,15 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_MND_CANT_DROP_LEADER          TAOS_DEF_ERROR_CODE(0, 0x0333)
 #define TSDB_CODE_MND_NO_ENOUGH_DNODES          TAOS_DEF_ERROR_CODE(0, 0x0334)
 #define TSDB_CODE_MND_INVALID_CLUSTER_CFG       TAOS_DEF_ERROR_CODE(0, 0x0335)
+// #define TSDB_CODE_MND_INVALID_DNODE_CFG_...  TAOS_DEF_ERROR_CODE(0, 0x0336)  // 2.x
+// #define TSDB_CODE_MND_BALANCE_ENABLED        TAOS_DEF_ERROR_CODE(0, 0x0337)  // 2.x
 #define TSDB_CODE_MND_VGROUP_NOT_IN_DNODE       TAOS_DEF_ERROR_CODE(0, 0x0338)
 #define TSDB_CODE_MND_VGROUP_ALREADY_IN_DNODE   TAOS_DEF_ERROR_CODE(0, 0x0339)
+// #define TSDB_CODE_MND_DNODE_NOT_FREE         TAOS_DEF_ERROR_CODE(0, 0x033A)  // 2.x
 #define TSDB_CODE_MND_INVALID_CLUSTER_ID        TAOS_DEF_ERROR_CODE(0, 0x033B)
+// #define TSDB_CODE_MND_NOT_READY              TAOS_DEF_ERROR_CODE(0, 0x033C)  // 2.x
+// #define TSDB_CODE_MND_DNODE_ID_NOT_CONFIGUREDTAOS_DEF_ERROR_CODE(0, 0x033D)  // 2.x
+// #define TSDB_CODE_MND_DNODE_EP_NOT_CONFIGUREDTAOS_DEF_ERROR_CODE(0, 0x033E)  // 2.x
 
 // mnode-acct
 #define TSDB_CODE_MND_ACCT_ALREADY_EXIST        TAOS_DEF_ERROR_CODE(0, 0x0340)
@@ -298,6 +306,8 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_MND_VGROUP_UN_CHANGED         TAOS_DEF_ERROR_CODE(0, 0x03B5)
 #define TSDB_CODE_MND_HAS_OFFLINE_DNODE         TAOS_DEF_ERROR_CODE(0, 0x03B6)
 #define TSDB_CODE_MND_INVALID_REPLICA           TAOS_DEF_ERROR_CODE(0, 0x03B7)
+#define TSDB_CODE_MND_DNODE_IN_CREATING         TAOS_DEF_ERROR_CODE(0, 0x03B8)
+#define TSDB_CODE_MND_DNODE_IN_DROPPING         TAOS_DEF_ERROR_CODE(0, 0x03B9)
 
 // mnode-stable-part2
 #define TSDB_CODE_MND_NAME_CONFLICT_WITH_TOPIC  TAOS_DEF_ERROR_CODE(0, 0x03C0)
@@ -318,6 +328,7 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_MND_TRANS_CLOG_IS_NULL        TAOS_DEF_ERROR_CODE(0, 0x03D4)
 #define TSDB_CODE_MND_TRANS_NETWORK_UNAVAILL    TAOS_DEF_ERROR_CODE(0, 0x03D5)
 #define TSDB_CODE_MND_LAST_TRANS_NOT_FINISHED   TAOS_DEF_ERROR_CODE(0, 0x03D6) //internal
+#define TSDB_CODE_MND_TRNAS_SYNC_TIMEOUT        TAOS_DEF_ERROR_CODE(0, 0x03D7)
 #define TSDB_CODE_MND_TRANS_UNKNOW_ERROR        TAOS_DEF_ERROR_CODE(0, 0x03DF)
 
 // mnode-mq
@@ -509,6 +520,8 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_SYN_STANDBY_NOT_READY         TAOS_DEF_ERROR_CODE(0, 0x0912)
 #define TSDB_CODE_SYN_BATCH_ERROR               TAOS_DEF_ERROR_CODE(0, 0x0913)
 #define TSDB_CODE_SYN_RESTORING                 TAOS_DEF_ERROR_CODE(0, 0x0914)
+#define TSDB_CODE_SYN_INVALID_SNAPSHOT_MSG      TAOS_DEF_ERROR_CODE(0, 0x0915)  // internal
+#define TSDB_CODE_SYN_BUFFER_FULL               TAOS_DEF_ERROR_CODE(0, 0x0916)  //
 #define TSDB_CODE_SYN_INTERNAL_ERROR            TAOS_DEF_ERROR_CODE(0, 0x09FF)
 
 // tq

include/util/tdef.h

@@ -255,7 +255,7 @@ typedef enum ELogicConditionType {
 #define TSDB_EP_LEN             (TSDB_FQDN_LEN + 6)
 #define TSDB_IPv4ADDR_LEN       16
 #define TSDB_FILENAME_LEN       128
-#define TSDB_SHOW_SQL_LEN       1024
+#define TSDB_SHOW_SQL_LEN       2048
 #define TSDB_SLOW_QUERY_SQL_LEN 512
 #define TSDB_SHOW_SUBQUERY_LEN  1000
 

include/util/tworker.h

@@ -17,6 +17,7 @@
 #define _TD_UTIL_WORKER_H_
 
 #include "tqueue.h"
+#include "tarray.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -26,10 +27,10 @@ typedef struct SQWorkerPool SQWorkerPool;
 typedef struct SWWorkerPool SWWorkerPool;
 
 typedef struct SQWorker {
-  int32_t       id;      // worker id
+  int32_t  id;      // worker id
-  int64_t       pid;     // thread pid
+  int64_t  pid;     // thread pid
-  TdThread      thread;  // thread id
+  TdThread thread;  // thread id
-  SQWorkerPool *pool;
+  void    *pool;
 } SQWorker;
 
 typedef struct SQWorkerPool {
@@ -42,6 +43,14 @@ typedef struct SQWorkerPool {
   TdThreadMutex mutex;
 } SQWorkerPool;
 
+typedef struct SAutoQWorkerPool {
+  float         ratio;
+  STaosQset    *qset;
+  const char   *name;
+  SArray       *workers;
+  TdThreadMutex mutex;
+} SAutoQWorkerPool;
+
 typedef struct SWWorker {
   int32_t       id;      // worker id
   int64_t       pid;     // thread pid
@@ -65,6 +74,11 @@ void        tQWorkerCleanup(SQWorkerPool *pool);
 STaosQueue *tQWorkerAllocQueue(SQWorkerPool *pool, void *ahandle, FItem fp);
 void        tQWorkerFreeQueue(SQWorkerPool *pool, STaosQueue *queue);
 
+int32_t     tAutoQWorkerInit(SAutoQWorkerPool *pool);
+void        tAutoQWorkerCleanup(SAutoQWorkerPool *pool);
+STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem fp);
+void        tAutoQWorkerFreeQueue(SAutoQWorkerPool *pool, STaosQueue *queue);
+
 int32_t     tWWorkerInit(SWWorkerPool *pool);
 void        tWWorkerCleanup(SWWorkerPool *pool);
 STaosQueue *tWWorkerAllocQueue(SWWorkerPool *pool, void *ahandle, FItems fp);

packaging/tools/make_install.bat

@@ -14,6 +14,7 @@ set binary_dir=%3
 set binary_dir=%binary_dir:/=\\%
 set osType=%4
 set verNumber=%5
+set Enterprise=%6
 set target_dir=C:\\TDengine
 
 if not exist %target_dir% (
@@ -57,6 +58,32 @@ if exist %binary_dir%\\build\\lib\\taosws.dll (
 if exist %binary_dir%\\build\\bin\\taosdump.exe (
     copy %binary_dir%\\build\\bin\\taosdump.exe %target_dir% > nul
 )
+if %Enterprise% == TRUE (
+    if exist %binary_dir%\\build\\bin\\taosx.exe (
+        copy %binary_dir%\\build\\bin\\taosx.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\tmq_sim.exe (
+        copy %binary_dir%\\build\\bin\\tmq_sim.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\tsim.exe (
+        copy %binary_dir%\\build\\bin\\tsim.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\tmq_taosx_ci.exe (
+        copy %binary_dir%\\build\\bin\\tmq_taosx_ci.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\tmq_demo.exe (
+        copy %binary_dir%\\build\\bin\\tmq_demo.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\dumper.exe (
+        copy %binary_dir%\\build\\bin\\dumper.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\runUdf.exe (
+        copy %binary_dir%\\build\\bin\\runUdf.exe %target_dir% > nul
+    )
+    if exist %binary_dir%\\build\\bin\\create_table.exe (
+        copy %binary_dir%\\build\\bin\\create_table.exe %target_dir% > nul
+    )
+)
     
 copy %binary_dir%\\build\\bin\\taosd.exe %target_dir% > nul
 copy %binary_dir%\\build\\bin\\udfd.exe %target_dir% > nul

packaging/tools/makepkg.sh

@@ -348,7 +348,8 @@ cd ${release_dir}
 #  install_dir has been distinguishes  cluster from  edege, so comments this code
 pkg_name=${install_dir}-${osType}-${cpuType}
 
-taostools_pkg_name=${taostools_install_dir}-${osType}-${cpuType}
+versionCompFirst=$(echo ${versionComp} | awk -F '.' '{print $1}')
+taostools_pkg_name=${taostools_install_dir}-${osType}-${cpuType}-comp${versionCompFirst}
 
 # if [ "$verMode" == "cluster" ]; then
 #   pkg_name=${install_dir}-${osType}-${cpuType}

source/client/src/clientImpl.c

@@ -122,6 +122,12 @@ STscObj* taos_connect_internal(const char* ip, const char* user, const char* pas
 
   char* key = getClusterKey(user, secretEncrypt, ip, port);
 
+  tscInfo("connecting to server, numOfEps:%d inUse:%d user:%s db:%s key:%s", epSet.epSet.numOfEps, epSet.epSet.inUse,
+          user, db, key);
+  for (int32_t i = 0; i < epSet.epSet.numOfEps; ++i) {
+    tscInfo("ep:%d, %s:%u", i, epSet.epSet.eps[i].fqdn, epSet.epSet.eps[i].port);
+  }
+
   SAppInstInfo** pInst = NULL;
   taosThreadMutexLock(&appInfo.mutex);
 
@@ -142,7 +148,7 @@ STscObj* taos_connect_internal(const char* ip, const char* user, const char* pas
     taosHashPut(appInfo.pInstMap, key, strlen(key), &p, POINTER_BYTES);
     p->instKey = key;
     key = NULL;
-    tscDebug("new app inst mgr %p, user:%s, ip:%s, port:%d", p, user, ip, port);
+    tscDebug("new app inst mgr %p, user:%s, ip:%s, port:%d", p, user, epSet.epSet.eps[0].fqdn, epSet.epSet.eps[0].port);
 
     pInst = &p;
   }
@@ -266,7 +272,7 @@ int32_t parseSql(SRequestObj* pRequest, bool topicQuery, SQuery** pQuery, SStmtC
 
 int32_t execLocalCmd(SRequestObj* pRequest, SQuery* pQuery) {
   SRetrieveTableRsp* pRsp = NULL;
-  int32_t            code = qExecCommand(pRequest->pTscObj->sysInfo, pQuery->pRoot, &pRsp);
+  int32_t            code = qExecCommand(&pRequest->pTscObj->id, pRequest->pTscObj->sysInfo, pQuery->pRoot, &pRsp);
   if (TSDB_CODE_SUCCESS == code && NULL != pRsp) {
     code = setQueryResultFromRsp(&pRequest->body.resInfo, pRsp, false, true);
   }
@@ -304,7 +310,7 @@ void asyncExecLocalCmd(SRequestObj* pRequest, SQuery* pQuery) {
     return;
   }
 
-  int32_t code = qExecCommand(pRequest->pTscObj->sysInfo, pQuery->pRoot, &pRsp);
+  int32_t code = qExecCommand(&pRequest->pTscObj->id ,pRequest->pTscObj->sysInfo, pQuery->pRoot, &pRsp);
   if (TSDB_CODE_SUCCESS == code && NULL != pRsp) {
     code = setQueryResultFromRsp(&pRequest->body.resInfo, pRsp, false, true);
   }
@@ -1397,6 +1403,21 @@ void processMsgFromServer(void* parent, SRpcMsg* pMsg, SEpSet* pEpSet) {
     memcpy((void*)tEpSet, (void*)pEpSet, sizeof(SEpSet));
   }
 
+  // pMsg is response msg
+  if (pMsg->msgType == TDMT_MND_CONNECT + 1) {
+    // restore origin code
+    if (pMsg->code == TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED) {
+      pMsg->code = TSDB_CODE_RPC_NETWORK_UNAVAIL;
+    } else if (pMsg->code == TSDB_CODE_RPC_SOMENODE_BROKEN_LINK) {
+      pMsg->code = TSDB_CODE_RPC_BROKEN_LINK;
+    }
+  } else {
+    // uniform to one error code: TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED
+    if (pMsg->code == TSDB_CODE_RPC_SOMENODE_BROKEN_LINK) {
+      pMsg->code = TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED;
+    }
+  }
+
   AsyncArg* arg = taosMemoryCalloc(1, sizeof(AsyncArg));
   arg->msg = *pMsg;
   arg->pEpset = tEpSet;

source/client/src/clientRawBlockWrite.c

@@ -179,7 +179,7 @@ static char* buildAlterSTableJson(void* alterData, int32_t alterDataLen) {
   }
   string = cJSON_PrintUnformatted(json);
 
-end:
+  end:
   cJSON_Delete(json);
   tFreeSMAltertbReq(&req);
   return string;
@@ -200,7 +200,7 @@ static char* processCreateStb(SMqMetaRsp* metaRsp) {
   }
   string = buildCreateTableJson(&req.schemaRow, &req.schemaTag, req.name, req.suid, TSDB_SUPER_TABLE);
 
-_err:
+  _err:
   tDecoderClear(&coder);
   return string;
 }
@@ -220,7 +220,7 @@ static char* processAlterStb(SMqMetaRsp* metaRsp) {
   }
   string = buildAlterSTableJson(req.alterOriData, req.alterOriDataLen);
 
-_err:
+  _err:
   tDecoderClear(&coder);
   return string;
 }
@@ -302,7 +302,7 @@ static void buildChildElement(cJSON* json, SVCreateTbReq* pCreateReq) {
     cJSON_AddItemToArray(tags, tag);
   }
 
-end:
+  end:
   cJSON_AddItemToObject(json, "tags", tags);
   taosArrayDestroy(pTagVals);
 }
@@ -360,7 +360,7 @@ static char* processCreateTable(SMqMetaRsp* metaRsp) {
     }
   }
 
-_exit:
+  _exit:
   for (int32_t iReq = 0; iReq < req.nReqs; iReq++) {
     pCreateReq = req.pReqs + iReq;
     taosMemoryFreeClear(pCreateReq->comment);
@@ -393,7 +393,7 @@ static char* processAutoCreateTable(STaosxRsp* rsp) {
   }
   string = buildCreateCTableJson(pCreateReq, rsp->createTableNum);
 
-_exit:
+  _exit:
   for (int i = 0; i < rsp->createTableNum; i++) {
     tDecoderClear(&decoder[i]);
     taosMemoryFreeClear(pCreateReq[i].comment);
@@ -515,7 +515,7 @@ static char* processAlterTable(SMqMetaRsp* metaRsp) {
   }
   string = cJSON_PrintUnformatted(json);
 
-_exit:
+  _exit:
   cJSON_Delete(json);
   tDecoderClear(&decoder);
   return string;
@@ -548,11 +548,50 @@ static char* processDropSTable(SMqMetaRsp* metaRsp) {
 
   string = cJSON_PrintUnformatted(json);
 
-_exit:
+  _exit:
   cJSON_Delete(json);
   tDecoderClear(&decoder);
   return string;
 }
+static char* processDeleteTable(SMqMetaRsp* metaRsp){
+  SDeleteRes req = {0};
+  SDecoder   coder = {0};
+  int32_t    code = TSDB_CODE_SUCCESS;
+  cJSON*     json = NULL;
+  char*      string = NULL;
+
+  // decode and process req
+  void*   data = POINTER_SHIFT(metaRsp->metaRsp, sizeof(SMsgHead));
+  int32_t len = metaRsp->metaRspLen - sizeof(SMsgHead);
+
+  tDecoderInit(&coder, data, len);
+  if (tDecodeDeleteRes(&coder, &req) < 0) {
+    code = TSDB_CODE_INVALID_PARA;
+    goto _exit;
+  }
+
+  //  getTbName(req.tableFName);
+  char sql[256] = {0};
+  snprintf(sql, sizeof(sql), "delete from `%s` where `%s` >= %" PRId64 " and `%s` <= %" PRId64, req.tableFName,
+           req.tsColName, req.skey, req.tsColName, req.ekey);
+  uDebug("delete sql:%s\n", sql);
+
+  json = cJSON_CreateObject();
+  if (json == NULL) {
+    goto _exit;
+  }
+  cJSON* type = cJSON_CreateString("delete");
+  cJSON_AddItemToObject(json, "type", type);
+  cJSON* sqlJson = cJSON_CreateString(sql);
+  cJSON_AddItemToObject(json, "sql", sqlJson);
+
+  string = cJSON_PrintUnformatted(json);
+
+  _exit:
+  cJSON_Delete(json);
+  tDecoderClear(&coder);
+  return string;
+}
 
 static char* processDropTable(SMqMetaRsp* metaRsp) {
   SDecoder         decoder = {0};
@@ -590,7 +629,7 @@ static char* processDropTable(SMqMetaRsp* metaRsp) {
 
   string = cJSON_PrintUnformatted(json);
 
-_exit:
+  _exit:
   cJSON_Delete(json);
   tDecoderClear(&decoder);
   return string;
@@ -678,7 +717,7 @@ static int32_t taosCreateStb(TAOS* taos, void* meta, int32_t metaLen) {
   code = pRequest->code;
   taosMemoryFree(pCmdMsg.pMsg);
 
-end:
+  end:
   destroyRequest(pRequest);
   tFreeSMCreateStbReq(&pReq);
   tDecoderClear(&coder);
@@ -748,7 +787,7 @@ static int32_t taosDropStb(TAOS* taos, void* meta, int32_t metaLen) {
   code = pRequest->code;
   taosMemoryFree(pCmdMsg.pMsg);
 
-end:
+  end:
   destroyRequest(pRequest);
   tDecoderClear(&coder);
   return code;
@@ -809,9 +848,9 @@ static int32_t taosCreateTable(TAOS* taos, void* meta, int32_t metaLen) {
   taosHashSetFreeFp(pVgroupHashmap, destroyCreateTbReqBatch);
 
   SRequestConnInfo conn = {.pTrans = pTscObj->pAppInfo->pTransporter,
-                           .requestId = pRequest->requestId,
+      .requestId = pRequest->requestId,
-                           .requestObjRefId = pRequest->self,
+      .requestObjRefId = pRequest->self,
-                           .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
+      .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
 
   pRequest->tableList = taosArrayInit(req.nReqs, sizeof(SName));
   // loop to create table
@@ -891,7 +930,7 @@ static int32_t taosCreateTable(TAOS* taos, void* meta, int32_t metaLen) {
 
   code = pRequest->code;
 
-end:
+  end:
   for (int32_t iReq = 0; iReq < req.nReqs; iReq++) {
     pCreateReq = req.pReqs + iReq;
     taosMemoryFreeClear(pCreateReq->comment);
@@ -961,9 +1000,9 @@ static int32_t taosDropTable(TAOS* taos, void* meta, int32_t metaLen) {
   taosHashSetFreeFp(pVgroupHashmap, destroyDropTbReqBatch);
 
   SRequestConnInfo conn = {.pTrans = pTscObj->pAppInfo->pTransporter,
-                           .requestId = pRequest->requestId,
+      .requestId = pRequest->requestId,
-                           .requestObjRefId = pRequest->self,
+      .requestObjRefId = pRequest->self,
-                           .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
+      .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
   pRequest->tableList = taosArrayInit(req.nReqs, sizeof(SName));
   // loop to create table
   for (int32_t iReq = 0; iReq < req.nReqs; iReq++) {
@@ -1015,7 +1054,7 @@ static int32_t taosDropTable(TAOS* taos, void* meta, int32_t metaLen) {
   }
   code = pRequest->code;
 
-end:
+  end:
   taosHashCleanup(pVgroupHashmap);
   destroyRequest(pRequest);
   tDecoderClear(&coder);
@@ -1083,7 +1122,7 @@ static int32_t taosDeleteData(TAOS* taos, void* meta, int32_t metaLen) {
   }
   taos_free_result(res);
 
-end:
+  end:
   tDecoderClear(&coder);
   return code;
 }
@@ -1130,9 +1169,9 @@ static int32_t taosAlterTable(TAOS* taos, void* meta, int32_t metaLen) {
   }
 
   SRequestConnInfo conn = {.pTrans = pTscObj->pAppInfo->pTransporter,
-                           .requestId = pRequest->requestId,
+      .requestId = pRequest->requestId,
-                           .requestObjRefId = pRequest->self,
+      .requestObjRefId = pRequest->self,
-                           .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
+      .mgmtEps = getEpSet_s(&pTscObj->pAppInfo->mgmtEp)};
 
   SVgroupInfo pInfo = {0};
   SName       pName = {0};
@@ -1191,7 +1230,7 @@ static int32_t taosAlterTable(TAOS* taos, void* meta, int32_t metaLen) {
       code = handleAlterTbExecRes(pRes->res, pCatalog);
     }
   }
-end:
+  end:
   taosArrayDestroy(pArray);
   if (pVgData) taosMemoryFreeClear(pVgData->pData);
   taosMemoryFreeClear(pVgData);
@@ -1354,7 +1393,7 @@ int taos_write_raw_block(TAOS* taos, int rows, char* pData, const char* tbname)
   launchQueryImpl(pRequest, pQuery, true, NULL);
   code = pRequest->code;
 
-end:
+  end:
   taosMemoryFreeClear(pTableMeta);
   qDestroyQuery(pQuery);
   destroyRequest(pRequest);
@@ -1414,7 +1453,6 @@ static int32_t tmqWriteRawDataImpl(TAOS* taos, void* data, int32_t dataLen) {
     goto end;
   }
   pVgHash = taosHashInit(16, taosGetDefaultHashFunction(TSDB_DATA_TYPE_INT), true, HASH_NO_LOCK);
-
   uDebug("raw data block num:%d\n", rspObj.rsp.blockNum);
   while (++rspObj.resIter < rspObj.rsp.blockNum) {
     SRetrieveTableRsp* pRetrieve = (SRetrieveTableRsp*)taosArrayGetP(rspObj.rsp.blockData, rspObj.resIter);
@@ -1474,7 +1512,7 @@ static int32_t tmqWriteRawDataImpl(TAOS* taos, void* data, int32_t dataLen) {
   launchQueryImpl(pRequest, pQuery, true, NULL);
   code = pRequest->code;
 
-end:
+  end:
   tDeleteSMqDataRsp(&rspObj.rsp);
   tDecoderClear(&decoder);
   qDestroyQuery(pQuery);
@@ -1647,7 +1685,12 @@ char* tmq_get_json_meta(TAOS_RES* res) {
     return processAlterTable(&pMetaRspObj->metaRsp);
   } else if (pMetaRspObj->metaRsp.resMsgType == TDMT_VND_DROP_TABLE) {
     return processDropTable(&pMetaRspObj->metaRsp);
+  } else if (pMetaRspObj->metaRsp.resMsgType == TDMT_VND_DROP_TABLE) {
+    return processDropTable(&pMetaRspObj->metaRsp);
+  } else if (pMetaRspObj->metaRsp.resMsgType == TDMT_VND_DELETE) {
+    return processDeleteTable(&pMetaRspObj->metaRsp);
   }
+
   return NULL;
 }
 

source/client/src/clientTmq.c

@@ -1902,10 +1902,6 @@ tmq_res_t tmq_get_res_type(TAOS_RES* res) {
   if (TD_RES_TMQ(res)) {
     return TMQ_RES_DATA;
   } else if (TD_RES_TMQ_META(res)) {
-    SMqMetaRspObj* pMetaRspObj = (SMqMetaRspObj*)res;
-    if (pMetaRspObj->metaRsp.resMsgType == TDMT_VND_DELETE) {
-      return TMQ_RES_DATA;
-    }
     return TMQ_RES_TABLE_META;
   } else if (TD_RES_TMQ_METADATA(res)) {
     return TMQ_RES_METADATA;

source/common/src/tdatablock.c

@@ -1879,7 +1879,7 @@ void blockDebugShowDataBlocks(const SArray* dataBlocks, const char* flag) {
   char    pBuf[128] = {0};
   int32_t sz = taosArrayGetSize(dataBlocks);
   for (int32_t i = 0; i < sz; i++) {
-    SSDataBlock* pDataBlock = taosArrayGetP(dataBlocks, i);
+    SSDataBlock* pDataBlock = taosArrayGet(dataBlocks, i);
     size_t       numOfCols = taosArrayGetSize(pDataBlock->pDataBlock);
 
     int32_t rows = pDataBlock->info.rows;
@@ -1891,21 +1891,37 @@ void blockDebugShowDataBlocks(const SArray* dataBlocks, const char* flag) {
       for (int32_t k = 0; k < numOfCols; k++) {
         SColumnInfoData* pColInfoData = taosArrayGet(pDataBlock->pDataBlock, k);
         void*            var = POINTER_SHIFT(pColInfoData->pData, j * pColInfoData->info.bytes);
+        if (k == 0) {
+          printf("cols:%d |", (int32_t)numOfCols);
+        }
         if (colDataIsNull(pColInfoData, rows, j, NULL)) {
           printf(" %15s |", "NULL");
           continue;
         }
+
         switch (pColInfoData->info.type) {
           case TSDB_DATA_TYPE_TIMESTAMP:
             formatTimestamp(pBuf, *(uint64_t*)var, TSDB_TIME_PRECISION_MILLI);
             printf(" %25s |", pBuf);
             break;
           case TSDB_DATA_TYPE_BOOL:
-            printf(" %15d |", *(int32_t*)var);
+            printf(" %15" PRIi8 " |", *(int8_t*)var);
+            break;
+          case TSDB_DATA_TYPE_TINYINT:
+            printf(" %15" PRIi8 " |", *(int8_t*)var);
+            break;
+          case TSDB_DATA_TYPE_SMALLINT:
+            printf(" %15" PRIi16 " |", *(int16_t*)var);
             break;
           case TSDB_DATA_TYPE_INT:
             printf(" %15d |", *(int32_t*)var);
             break;
+          case TSDB_DATA_TYPE_UTINYINT:
+            printf(" %15" PRIu8 " |", *(uint8_t*)var);
+            break;
+          case TSDB_DATA_TYPE_USMALLINT:
+            printf(" %15" PRIu16 " |", *(uint16_t*)var);
+            break;
           case TSDB_DATA_TYPE_UINT:
             printf(" %15u |", *(uint32_t*)var);
             break;

source/common/src/tdataformat.c

@@ -2091,28 +2091,28 @@ _exit:
   return code;
 }
 
-int32_t tColDataAddValueByDataBlock(SColData *pColData, int8_t type, int32_t bytes,
+int32_t tColDataAddValueByDataBlock(SColData *pColData, int8_t type, int32_t bytes, int32_t nRows, char *lengthOrbitmap,
-                                    int32_t nRows, char* lengthOrbitmap, char *data) {
+                                    char *data) {
   int32_t code = 0;
 
   if (IS_VAR_DATA_TYPE(type)) {  // var-length data type
     for (int32_t i = 0; i < nRows; ++i) {
-      int32_t offset = *((int32_t*)lengthOrbitmap + i);
+      int32_t offset = *((int32_t *)lengthOrbitmap + i);
       if (offset == -1) {
         code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_NULL](pColData, NULL, 0);
         if (code) goto _exit;
       } else {
-        code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](
+        code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](pColData, (uint8_t *)varDataVal(data + offset),
-            pColData, (uint8_t *)varDataVal(data + offset), varDataLen(data + offset));
+                                                                      varDataLen(data + offset));
       }
     }
   } else {  // fixed-length data type
     bool allValue = true;
-    bool allNull  = true;
+    bool allNull = true;
     for (int32_t i = 0; i < nRows; ++i) {
-      if(!colDataIsNull_f(lengthOrbitmap, i)){
+      if (!colDataIsNull_f(lengthOrbitmap, i)) {
         allNull = false;
-      }else{
+      } else {
         allValue = false;
       }
     }
@@ -2120,8 +2120,7 @@ int32_t tColDataAddValueByDataBlock(SColData *pColData, int8_t type, int32_t byt
     if (allValue) {
       // optimize (todo)
       for (int32_t i = 0; i < nRows; ++i) {
-        code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](
+        code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](pColData, (uint8_t *)data + bytes * i, bytes);
-            pColData, (uint8_t *)data + bytes * i, bytes);
       }
     } else if (allNull) {
       // optimize (todo)
@@ -2135,14 +2134,13 @@ int32_t tColDataAddValueByDataBlock(SColData *pColData, int8_t type, int32_t byt
           code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_NULL](pColData, NULL, 0);
           if (code) goto _exit;
         } else {
-          code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](
+          code = tColDataAppendValueImpl[pColData->flag][CV_FLAG_VALUE](pColData, (uint8_t *)data + bytes * i, bytes);
-              pColData, (uint8_t *)data + bytes * i, bytes);
         }
       }
     }
   }
 
-  _exit:
+_exit:
   return code;
 }
 
@@ -2218,13 +2216,18 @@ static int32_t tColDataSwapValue(SColData *pColData, int32_t i, int32_t j) {
     if (nData1 > nData2) {
       memcpy(pData, pColData->pData + pColData->aOffset[i], nData1);
       memcpy(pColData->pData + pColData->aOffset[i], pColData->pData + pColData->aOffset[j], nData2);
-      memmove(pColData->pData + pColData->aOffset[i] + nData2, pColData->pData + pColData->aOffset[i] + nData1,
+      // memmove(pColData->pData + pColData->aOffset[i] + nData2, pColData->pData + pColData->aOffset[i] + nData1,
+      //         pColData->aOffset[j] - pColData->aOffset[i + 1]);
+      memmove(pColData->pData + pColData->aOffset[i] + nData2, pColData->pData + pColData->aOffset[i + 1],
               pColData->aOffset[j] - pColData->aOffset[i + 1]);
       memcpy(pColData->pData + pColData->aOffset[j] + nData2 - nData1, pData, nData1);
     } else {
       memcpy(pData, pColData->pData + pColData->aOffset[j], nData2);
       memcpy(pColData->pData + pColData->aOffset[j] + nData2 - nData1, pColData->pData + pColData->aOffset[i], nData1);
-      memmove(pColData->pData + pColData->aOffset[j] + nData2 - nData1, pColData->pData + pColData->aOffset[i] + nData1,
+      // memmove(pColData->pData + pColData->aOffset[j] + nData2 - nData1, pColData->pData + pColData->aOffset[i] +
+      // nData1,
+      //         pColData->aOffset[j] - pColData->aOffset[i + 1]);
+      memmove(pColData->pData + pColData->aOffset[i] + nData2, pColData->pData + pColData->aOffset[i + 1],
               pColData->aOffset[j] - pColData->aOffset[i + 1]);
       memcpy(pColData->pData + pColData->aOffset[i], pData, nData2);
     }
@@ -2244,6 +2247,7 @@ static int32_t tColDataSwapValue(SColData *pColData, int32_t i, int32_t j) {
 _exit:
   return code;
 }
+
 static void tColDataSwap(SColData *pColData, int32_t i, int32_t j) {
   ASSERT(i < j);
   ASSERT(j < pColData->nVal);
@@ -2278,9 +2282,174 @@ static void tColDataSwap(SColData *pColData, int32_t i, int32_t j) {
       break;
   }
 }
-static void tColDataSort(SColData *aColData, int32_t nColData) {
+
-  if (aColData[0].nVal == 0) return;
+static int32_t tColDataCopyRowCell(SColData *pFromColData, int32_t iFromRow, SColData *pToColData, int32_t iToRow) {
-  // TODO
+  int32_t code = TSDB_CODE_SUCCESS;
+
+  if (IS_VAR_DATA_TYPE(pToColData->type)) {
+    int32_t nData = (iFromRow < pFromColData->nVal - 1)
+                        ? pFromColData->aOffset[iFromRow + 1] - pFromColData->aOffset[iFromRow]
+                        : pFromColData->nData - pFromColData->aOffset[iFromRow];
+    if (iToRow == 0) {
+      pToColData->aOffset[iToRow] = 0;
+    }
+
+    if (iToRow < pToColData->nVal - 1) {
+      pToColData->aOffset[iToRow + 1] = pToColData->aOffset[iToRow] + nData;
+    }
+
+    memcpy(pToColData->pData + pToColData->aOffset[iToRow], pFromColData->pData + pFromColData->aOffset[iFromRow],
+           nData);
+  } else {
+    memcpy(&pToColData->pData[TYPE_BYTES[pToColData->type] * iToRow],
+           &pFromColData->pData[TYPE_BYTES[pToColData->type] * iFromRow], TYPE_BYTES[pToColData->type]);
+  }
+  return code;
+}
+
+static int32_t tColDataCopyRowSingleCol(SColData *pFromColData, int32_t iFromRow, SColData *pToColData,
+                                        int32_t iToRow) {
+  int32_t code = TSDB_CODE_SUCCESS;
+
+  switch (pFromColData->flag) {
+    case HAS_NONE:
+    case HAS_NULL:
+      break;
+    case (HAS_NULL | HAS_NONE): {
+      SET_BIT1(pToColData->pBitMap, iToRow, GET_BIT1(pFromColData->pBitMap, iFromRow));
+    } break;
+    case HAS_VALUE: {
+      tColDataCopyRowCell(pFromColData, iFromRow, pToColData, iToRow);
+    } break;
+    case (HAS_VALUE | HAS_NONE):
+    case (HAS_VALUE | HAS_NULL): {
+      SET_BIT1(pToColData->pBitMap, iToRow, GET_BIT1(pFromColData->pBitMap, iFromRow));
+      tColDataCopyRowCell(pFromColData, iFromRow, pToColData, iToRow);
+    } break;
+    case (HAS_VALUE | HAS_NULL | HAS_NONE): {
+      SET_BIT2(pToColData->pBitMap, iToRow, GET_BIT2(pFromColData->pBitMap, iFromRow));
+      tColDataCopyRowCell(pFromColData, iFromRow, pToColData, iToRow);
+    } break;
+    default:
+      return -1;
+  }
+
+  return code;
+}
+
+static int32_t tColDataCopyRow(SColData *aFromColData, int32_t iFromRow, SColData *aToColData, int32_t iToRow,
+                               int32_t nColData) {
+  int32_t code = TSDB_CODE_SUCCESS;
+
+  for (int32_t i = 0; i < nColData; i++) {
+    code = tColDataCopyRowSingleCol(&aFromColData[i], iFromRow, &aToColData[i], iToRow);
+    if (code != TSDB_CODE_SUCCESS) {
+      return code;
+    }
+  }
+
+  return code;
+}
+
+static int32_t tColDataCopyRowAppend(SColData *aFromColData, int32_t iFromRow, SColData *aToColData, int32_t nColData) {
+  int32_t code = TSDB_CODE_SUCCESS;
+
+  for (int32_t i = 0; i < nColData; i++) {
+    SColVal cv = {0};
+    tColDataGetValue(&aFromColData[i], iFromRow, &cv);
+    code = tColDataAppendValue(&aToColData[i], &cv);
+    if (code != TSDB_CODE_SUCCESS) {
+      return code;
+    }
+  }
+
+  return code;
+}
+
+static int32_t tColDataMergeSortMerge(SColData *aColData, int32_t start, int32_t mid, int32_t end, int32_t nColData) {
+  SColData *aDstColData = NULL;
+  TSKEY    *aKey = (TSKEY *)aColData[0].pData;
+
+  int32_t i = start, j = mid + 1, k = 0;
+
+  if (end > start) {
+    aDstColData = taosMemoryCalloc(1, sizeof(SColData) * nColData);
+    for (int c = 0; c < nColData; ++c) {
+      tColDataInit(&aDstColData[c], aColData[c].cid, aColData[c].type, aColData[c].smaOn);
+    }
+    if (aDstColData == NULL) {
+      return TSDB_CODE_OUT_OF_MEMORY;
+    }
+    /*
+    for (int32_t i = 0; i < nColData; i++) {
+      tColDataCopy(&aColData[i], &aDstColData[i], tColDataDefaultMalloc, NULL);
+    }
+    */
+  }
+
+  while (i <= mid && j <= end) {
+    if (aKey[i] <= aKey[j]) {
+      // tColDataCopyRow(aColData, i++, aDstColData, k++);
+      tColDataCopyRowAppend(aColData, i++, aDstColData, nColData);
+    } else {
+      // tColDataCopyRow(aColData, j++, aDstColData, k++);
+      tColDataCopyRowAppend(aColData, j++, aDstColData, nColData);
+    }
+  }
+
+  while (i <= mid) {
+    // tColDataCopyRow(aColData, i++, aDstColData, k++);
+    tColDataCopyRowAppend(aColData, i++, aDstColData, nColData);
+  }
+
+  while (j <= end) {
+    // tColDataCopyRow(aColData, j++, aDstColData, k++);
+    tColDataCopyRowAppend(aColData, j++, aDstColData, nColData);
+  }
+
+  for (i = start, k = 0; i <= end; ++i, ++k) {
+    tColDataCopyRow(aDstColData, k, aColData, i, nColData);
+  }
+
+  if (aDstColData) {
+    for (int32_t i = 0; i < nColData; i++) {
+      tColDataDestroy(&aDstColData[i]);
+    }
+    taosMemoryFree(aDstColData);
+  }
+
+  return TSDB_CODE_SUCCESS;
+}
+
+static int32_t tColDataMergeSort(SColData *aColData, int32_t start, int32_t end, int32_t nColData) {
+  int32_t ret = TSDB_CODE_SUCCESS;
+  int32_t mid;
+
+  if (start >= end) {
+    return TSDB_CODE_SUCCESS;
+  }
+
+  mid = (start + end) / 2;
+
+  ret = tColDataMergeSort(aColData, start, mid, nColData);
+  if (ret != TSDB_CODE_SUCCESS) {
+    return ret;
+  }
+
+  ret = tColDataMergeSort(aColData, mid + 1, end, nColData);
+  if (ret != TSDB_CODE_SUCCESS) {
+    return ret;
+  }
+
+  return tColDataMergeSortMerge(aColData, start, mid, end, nColData);
+}
+
+static int32_t tColDataSort(SColData *aColData, int32_t nColData) {
+  int32_t nVal = aColData[0].nVal;
+
+  if (nVal < 2) return TSDB_CODE_SUCCESS;
+
+  return tColDataMergeSort(aColData, 0, nVal - 1, nColData);
 }
 static void tColDataMergeImpl(SColData *pColData, int32_t iStart, int32_t iEnd /* not included */) {
   switch (pColData->flag) {
@@ -2530,6 +2699,15 @@ void tColDataSortMerge(SArray *colDataArr) {
     tColDataSort(aColData, nColData);
   }
 
+  if (doMerge != 1) {
+    for (int32_t iVal = 1; iVal < aColData[0].nVal; ++iVal) {
+      if (aKey[iVal] == aKey[iVal - 1]) {
+        doMerge = 1;
+        break;
+      }
+    }
+  }
+
   // merge -------
   if (doMerge) {
     tColDataMerge(aColData, nColData);

source/common/src/tglobal.c

@@ -47,7 +47,7 @@ int32_t tsNumOfMnodeQueryThreads = 4;
 int32_t tsNumOfMnodeFetchThreads = 1;
 int32_t tsNumOfMnodeReadThreads = 1;
 int32_t tsNumOfVnodeQueryThreads = 4;
-int32_t tsNumOfVnodeStreamThreads = 2;
+float   tsRatioOfVnodeStreamThreads = 1.0;
 int32_t tsNumOfVnodeFetchThreads = 4;
 int32_t tsNumOfVnodeRsmaThreads = 2;
 int32_t tsNumOfQnodeQueryThreads = 4;
@@ -392,9 +392,7 @@ static int32_t taosAddServerCfg(SConfig *pCfg) {
   tsNumOfVnodeQueryThreads = TMAX(tsNumOfVnodeQueryThreads, 4);
   if (cfgAddInt32(pCfg, "numOfVnodeQueryThreads", tsNumOfVnodeQueryThreads, 4, 1024, 0) != 0) return -1;
 
-  tsNumOfVnodeStreamThreads = tsNumOfCores / 4;
+  if (cfgAddFloat(pCfg, "ratioOfVnodeStreamThreads", tsRatioOfVnodeStreamThreads, 0.01, 100, 0) != 0) return -1;
-  tsNumOfVnodeStreamThreads = TMAX(tsNumOfVnodeStreamThreads, 4);
-  if (cfgAddInt32(pCfg, "numOfVnodeStreamThreads", tsNumOfVnodeStreamThreads, 4, 1024, 0) != 0) return -1;
 
   tsNumOfVnodeFetchThreads = tsNumOfCores / 4;
   tsNumOfVnodeFetchThreads = TMAX(tsNumOfVnodeFetchThreads, 4);
@@ -513,11 +511,9 @@ static int32_t taosUpdateServerCfg(SConfig *pCfg) {
     pItem->stype = stype;
   }
 
-  pItem = cfgGetItem(tsCfg, "numOfVnodeStreamThreads");
+  pItem = cfgGetItem(tsCfg, "ratioOfVnodeStreamThreads");
   if (pItem != NULL && pItem->stype == CFG_STYPE_DEFAULT) {
-    tsNumOfVnodeStreamThreads = numOfCores / 4;
+    pItem->fval = tsRatioOfVnodeStreamThreads;
-    tsNumOfVnodeStreamThreads = TMAX(tsNumOfVnodeStreamThreads, 4);
-    pItem->i32 = tsNumOfVnodeStreamThreads;
     pItem->stype = stype;
   }
 
@@ -710,7 +706,7 @@ static int32_t taosSetServerCfg(SConfig *pCfg) {
   tsNumOfCommitThreads = cfgGetItem(pCfg, "numOfCommitThreads")->i32;
   tsNumOfMnodeReadThreads = cfgGetItem(pCfg, "numOfMnodeReadThreads")->i32;
   tsNumOfVnodeQueryThreads = cfgGetItem(pCfg, "numOfVnodeQueryThreads")->i32;
-  tsNumOfVnodeStreamThreads = cfgGetItem(pCfg, "numOfVnodeStreamThreads")->i32;
+  tsRatioOfVnodeStreamThreads = cfgGetItem(pCfg, "ratioOfVnodeStreamThreads")->fval;
   tsNumOfVnodeFetchThreads = cfgGetItem(pCfg, "numOfVnodeFetchThreads")->i32;
   tsNumOfVnodeRsmaThreads = cfgGetItem(pCfg, "numOfVnodeRsmaThreads")->i32;
   tsNumOfQnodeQueryThreads = cfgGetItem(pCfg, "numOfQnodeQueryThreads")->i32;
@@ -1316,12 +1312,17 @@ void taosSetDebugFlag(int32_t *pFlagPtr, const char *flagName, int32_t flagVal,
   if (pItem != NULL && (rewrite || pItem->i32 == 0)) {
     pItem->i32 = flagVal;
   }
-  *pFlagPtr = flagVal;
+  if (pFlagPtr != NULL) {
+    *pFlagPtr = flagVal;
+  }
 }
 
 void taosSetAllDebugFlag(int32_t flag, bool rewrite) {
   if (flag <= 0) return;
 
+  taosSetDebugFlag(NULL, "debugFlag", flag, rewrite);
+  taosSetDebugFlag(NULL, "simDebugFlag", flag, rewrite);
+  taosSetDebugFlag(NULL, "tmrDebugFlag", flag, rewrite);
   taosSetDebugFlag(&uDebugFlag, "uDebugFlag", flag, rewrite);
   taosSetDebugFlag(&rpcDebugFlag, "rpcDebugFlag", flag, rewrite);
   taosSetDebugFlag(&jniDebugFlag, "jniDebugFlag", flag, rewrite);
@@ -1340,6 +1341,5 @@ void taosSetAllDebugFlag(int32_t flag, bool rewrite) {
   taosSetDebugFlag(&idxDebugFlag, "idxDebugFlag", flag, rewrite);
   taosSetDebugFlag(&tdbDebugFlag, "tdbDebugFlag", flag, rewrite);
   taosSetDebugFlag(&metaDebugFlag, "metaDebugFlag", flag, rewrite);
-  taosSetDebugFlag(&metaDebugFlag, "tmrDebugFlag", flag, rewrite);
   uInfo("all debug flag are set to %d", flag);
 }

source/common/src/tmsg.c

@@ -5424,6 +5424,7 @@ int32_t tSerializeSCMCreateStreamReq(void *buf, int32_t bufLen, const SCMCreateS
     if (tEncodeI32(&encoder, pField->bytes) < 0) return -1;
     if (tEncodeCStr(&encoder, pField->name) < 0) return -1;
   }
+  if (tEncodeI8(&encoder, pReq->createStb) < 0) return -1;
 
   tEndEncode(&encoder);
 
@@ -5484,6 +5485,7 @@ int32_t tDeserializeSCMCreateStreamReq(void *buf, int32_t bufLen, SCMCreateStrea
       }
     }
   }
+  if (tDecodeI8(&decoder, &pReq->createStb) < 0) return -1;
 
   tEndDecode(&decoder);
 
@@ -6936,6 +6938,8 @@ void tDestroySSubmitTbData(SSubmitTbData *pTbData, int32_t flag) {
 }
 
 void tDestroySSubmitReq2(SSubmitReq2 *pReq, int32_t flag) {
+  if (pReq->aSubmitTbData == NULL) return;
+
   int32_t        nSubmitTbData = TARRAY_SIZE(pReq->aSubmitTbData);
   SSubmitTbData *aSubmitTbData = (SSubmitTbData *)TARRAY_DATA(pReq->aSubmitTbData);
 

source/dnode/mgmt/exe/dmMain.c

@@ -268,6 +268,11 @@ int mainWindows(int argc, char **argv) {
 
   if (dmInit() != 0) {
     dError("failed to init dnode since %s", terrstr());
+
+    taosCleanupCfg();
+    taosCloseLog();
+    taosCleanupArgs();
+    taosConvDestroy();
     return -1;
   }
 

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

@@ -41,6 +41,8 @@ static void dmProcessStatusRsp(SDnodeMgmt *pMgmt, SRpcMsg *pRsp) {
              pMgmt->statusSeq);
       pMgmt->pData->dropped = 1;
       dmWriteEps(pMgmt->pData);
+      dInfo("dnode will exit since it is in the dropped state");
+      raise(SIGINT);
     }
   } else {
     SStatusRsp statusRsp = {0};

source/dnode/mgmt/mgmt_vnode/inc/vmInt.h

@@ -26,20 +26,20 @@ extern "C" {
 #endif
 
 typedef struct SVnodeMgmt {
-  SDnodeData    *pData;
+  SDnodeData      *pData;
-  SMsgCb         msgCb;
+  SMsgCb           msgCb;
-  const char    *path;
+  const char      *path;
-  const char    *name;
+  const char      *name;
-  SQWorkerPool   queryPool;
+  SQWorkerPool     queryPool;
-  SQWorkerPool   streamPool;
+  SAutoQWorkerPool streamPool;
-  SWWorkerPool   fetchPool;
+  SWWorkerPool     fetchPool;
-  SSingleWorker  mgmtWorker;
+  SSingleWorker    mgmtWorker;
-  SHashObj      *hash;
+  SHashObj        *hash;
-  TdThreadRwlock lock;
+  TdThreadRwlock   lock;
-  SVnodesStat    state;
+  SVnodesStat      state;
-  STfs          *pTfs;
+  STfs            *pTfs;
-  TdThread       thread;
+  TdThread         thread;
-  bool           stop;
+  bool             stop;
 } SVnodeMgmt;
 
 typedef struct {

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

@@ -140,6 +140,12 @@ static void vmProcessSyncQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOf
 
 static int32_t vmPutMsgToQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg, EQueueType qtype) {
   const STraceId *trace = &pMsg->info.traceId;
+  if (pMsg->contLen < sizeof(SMsgHead)) {
+    dGError("invalid rpc msg with no msg head at pCont. pMsg:%p, type:%s, contLen:%d", pMsg, TMSG_INFO(pMsg->msgType),
+            pMsg->contLen);
+    return -1;
+  }
+
   SMsgHead       *pHead = pMsg->pCont;
   int32_t         code = 0;
 
@@ -233,6 +239,14 @@ int32_t vmPutMsgToMgmtQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) {
 }
 
 int32_t vmPutRpcMsgToQueue(SVnodeMgmt *pMgmt, EQueueType qtype, SRpcMsg *pRpc) {
+  if (pRpc->contLen < sizeof(SMsgHead)) {
+    dError("invalid rpc msg with no msg head at pCont. pRpc:%p, type:%s, len:%d", pRpc, TMSG_INFO(pRpc->msgType),
+           pRpc->contLen);
+    rpcFreeCont(pRpc->pCont);
+    pRpc->pCont = NULL;
+    return -1;
+  }
+
   SRpcMsg *pMsg = taosAllocateQitem(sizeof(SRpcMsg), RPC_QITEM, pRpc->contLen);
   if (pMsg == NULL) {
     rpcFreeCont(pRpc->pCont);
@@ -304,7 +318,7 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
   (void)tMultiWorkerInit(&pVnode->pApplyW, &acfg);
 
   pVnode->pQueryQ = tQWorkerAllocQueue(&pMgmt->queryPool, pVnode, (FItem)vmProcessQueryQueue);
-  pVnode->pStreamQ = tQWorkerAllocQueue(&pMgmt->streamPool, pVnode, (FItem)vmProcessStreamQueue);
+  pVnode->pStreamQ = tAutoQWorkerAllocQueue(&pMgmt->streamPool, pVnode, (FItem)vmProcessStreamQueue);
   pVnode->pFetchQ = tWWorkerAllocQueue(&pMgmt->fetchPool, pVnode, (FItems)vmProcessFetchQueue);
 
   if (pVnode->pWriteW.queue == NULL || pVnode->pSyncW.queue == NULL || pVnode->pSyncCtrlW.queue == NULL ||
@@ -330,7 +344,7 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
 
 void vmFreeQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
   tQWorkerFreeQueue(&pMgmt->queryPool, pVnode->pQueryQ);
-  tQWorkerFreeQueue(&pMgmt->streamPool, pVnode->pStreamQ);
+  tAutoQWorkerFreeQueue(&pMgmt->streamPool, pVnode->pStreamQ);
   tWWorkerFreeQueue(&pMgmt->fetchPool, pVnode->pFetchQ);
   pVnode->pQueryQ = NULL;
   pVnode->pStreamQ = NULL;
@@ -345,11 +359,10 @@ int32_t vmStartWorker(SVnodeMgmt *pMgmt) {
   pQPool->max = tsNumOfVnodeQueryThreads;
   if (tQWorkerInit(pQPool) != 0) return -1;
 
-  SQWorkerPool *pStreamPool = &pMgmt->streamPool;
+  SAutoQWorkerPool *pStreamPool = &pMgmt->streamPool;
   pStreamPool->name = "vnode-stream";
-  pStreamPool->min = tsNumOfVnodeStreamThreads;
+  pStreamPool->ratio = tsRatioOfVnodeStreamThreads;
-  pStreamPool->max = tsNumOfVnodeStreamThreads;
+  if (tAutoQWorkerInit(pStreamPool) != 0) return -1;
-  if (tQWorkerInit(pStreamPool) != 0) return -1;
 
   SWWorkerPool *pFPool = &pMgmt->fetchPool;
   pFPool->name = "vnode-fetch";
@@ -371,7 +384,7 @@ int32_t vmStartWorker(SVnodeMgmt *pMgmt) {
 
 void vmStopWorker(SVnodeMgmt *pMgmt) {
   tQWorkerCleanup(&pMgmt->queryPool);
-  tQWorkerCleanup(&pMgmt->streamPool);
+  tAutoQWorkerCleanup(&pMgmt->streamPool);
   tWWorkerCleanup(&pMgmt->fetchPool);
   dDebug("vnode workers are closed");
 }

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

@@ -53,6 +53,15 @@ static bool dmFailFastFp(tmsg_t msgType) {
   return msgType == TDMT_SYNC_HEARTBEAT || msgType == TDMT_SYNC_APPEND_ENTRIES;
 }
 
+static void dmConvertErrCode(tmsg_t msgType) {
+  if (terrno != TSDB_CODE_APP_IS_STOPPING) {
+    return;
+  }
+  if ((msgType > TDMT_VND_MSG && msgType < TDMT_VND_MAX_MSG) ||
+      (msgType > TDMT_SCH_MSG && msgType < TDMT_SCH_MAX_MSG)) {
+    terrno = TSDB_CODE_VND_STOPPED;
+  }
+}
 static void dmProcessRpcMsg(SDnode *pDnode, SRpcMsg *pRpc, SEpSet *pEpSet) {
   SDnodeTrans  *pTrans = &pDnode->trans;
   int32_t       code = -1;
@@ -102,6 +111,11 @@ static void dmProcessRpcMsg(SDnode *pDnode, SRpcMsg *pRpc, SEpSet *pEpSet) {
     dGError("msg:%p, type:%s pCont is NULL", pRpc, TMSG_INFO(pRpc->msgType));
     terrno = TSDB_CODE_INVALID_MSG_LEN;
     goto _OVER;
+  } else if ((pRpc->code == TSDB_CODE_RPC_NETWORK_UNAVAIL || pRpc->code == TSDB_CODE_RPC_BROKEN_LINK) &&
+             (!IsReq(pRpc)) && (pRpc->pCont == NULL)) {
+    dGError("msg:%p, type:%s pCont is NULL, err: %s", pRpc, TMSG_INFO(pRpc->msgType), tstrerror(pRpc->code));
+    terrno = pRpc->code;
+    goto _OVER;
   }
 
   if (pHandle->defaultNtype == NODE_END) {
@@ -152,6 +166,7 @@ static void dmProcessRpcMsg(SDnode *pDnode, SRpcMsg *pRpc, SEpSet *pEpSet) {
 
 _OVER:
   if (code != 0) {
+    dmConvertErrCode(pRpc->msgType);
     if (terrno != 0) code = terrno;
     if (pMsg) {
       dGTrace("msg:%p, failed to process %s since %s", pMsg, TMSG_INFO(pMsg->msgType), terrstr());
@@ -233,8 +248,9 @@ static inline void dmReleaseHandle(SRpcHandleInfo *pHandle, int8_t type) { rpcRe
 
 static bool rpcRfp(int32_t code, tmsg_t msgType) {
   if (code == TSDB_CODE_RPC_NETWORK_UNAVAIL || code == TSDB_CODE_RPC_BROKEN_LINK || code == TSDB_CODE_MNODE_NOT_FOUND ||
-      code == TSDB_CODE_SYN_NOT_LEADER || code == TSDB_CODE_SYN_RESTORING || code == TSDB_CODE_VND_STOPPED ||
+      code == TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED || code == TSDB_CODE_SYN_NOT_LEADER ||
-      code == TSDB_CODE_APP_IS_STARTING || code == TSDB_CODE_APP_IS_STOPPING) {
+      code == TSDB_CODE_SYN_RESTORING || code == TSDB_CODE_VND_STOPPED || code == TSDB_CODE_APP_IS_STARTING ||
+      code == TSDB_CODE_APP_IS_STOPPING) {
     if (msgType == TDMT_SCH_QUERY || msgType == TDMT_SCH_MERGE_QUERY || msgType == TDMT_SCH_FETCH ||
         msgType == TDMT_SCH_MERGE_FETCH) {
       return false;

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

@@ -193,6 +193,7 @@ typedef struct {
   int64_t    lastAccessTime;
   int32_t    accessTimes;
   int32_t    numOfVnodes;
+  int32_t    numOfOtherNodes;
   int32_t    numOfSupportVnodes;
   float      numOfCores;
   int64_t    memTotal;

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

@@ -557,6 +557,8 @@ static int32_t mndCreateDb(SMnode *pMnode, SRpcMsg *pReq, SCreateDbReq *pCreate,
   mInfo("trans:%d, used to create db:%s", pTrans->id, pCreate->db);
 
   mndTransSetDbName(pTrans, dbObj.name, NULL);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
+
   mndTransSetOper(pTrans, MND_OPER_CREATE_DB);
   if (mndSetCreateDbRedoLogs(pMnode, pTrans, &dbObj, pVgroups) != 0) goto _OVER;
   if (mndSetCreateDbUndoLogs(pMnode, pTrans, &dbObj, pVgroups) != 0) goto _OVER;
@@ -776,7 +778,7 @@ static int32_t mndAlterDb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pOld, SDbObj *p
 
   int32_t code = -1;
   mndTransSetDbName(pTrans, pOld->name, NULL);
-  if (mndTrancCheckConflict(pMnode, pTrans) != 0) return -1;
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   if (mndSetAlterDbRedoLogs(pMnode, pTrans, pOld, pNew) != 0) goto _OVER;
   if (mndSetAlterDbCommitLogs(pMnode, pTrans, pOld, pNew) != 0) goto _OVER;
@@ -1038,7 +1040,7 @@ static int32_t mndDropDb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb) {
 
   mInfo("trans:%d, used to drop db:%s", pTrans->id, pDb->name);
   mndTransSetDbName(pTrans, pDb->name, NULL);
-
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
   if (mndCheckTopicExist(pMnode, pDb) < 0) goto _OVER;
 
   if (mndSetDropDbRedoLogs(pMnode, pTrans, pDb) != 0) goto _OVER;

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

@@ -217,8 +217,18 @@ SDnodeObj *mndAcquireDnode(SMnode *pMnode, int32_t dnodeId) {
   SSdb      *pSdb = pMnode->pSdb;
   SDnodeObj *pDnode = sdbAcquire(pSdb, SDB_DNODE, &dnodeId);
   if (pDnode == NULL) {
-    terrno = TSDB_CODE_MND_DNODE_NOT_EXIST;
+    if (terrno == TSDB_CODE_SDB_OBJ_NOT_THERE) {
+      terrno = TSDB_CODE_MND_DNODE_NOT_EXIST;
+    } else if (terrno == TSDB_CODE_SDB_OBJ_CREATING) {
+      terrno = TSDB_CODE_MND_DNODE_IN_CREATING;
+    } else if (terrno == TSDB_CODE_SDB_OBJ_DROPPING) {
+      terrno = TSDB_CODE_MND_DNODE_IN_DROPPING;
+    } else {
+      terrno = TSDB_CODE_APP_ERROR;
+      mFatal("dnode:%d, failed to acquire db since %s", dnodeId, terrstr());
+    }
   }
+
   return pDnode;
 }
 
@@ -254,6 +264,27 @@ static SDnodeObj *mndAcquireDnodeByEp(SMnode *pMnode, char *pEpStr) {
   return NULL;
 }
 
+static SDnodeObj *mndAcquireDnodeAllStatusByEp(SMnode *pMnode, char *pEpStr) {
+  SSdb *pSdb = pMnode->pSdb;
+
+  void *pIter = NULL;
+  while (1) {
+    SDnodeObj *pDnode = NULL;
+    ESdbStatus objStatus = 0;
+    pIter = sdbFetchAll(pSdb, SDB_DNODE, pIter, (void **)&pDnode, &objStatus, true);
+    if (pIter == NULL) break;
+
+    if (strncasecmp(pEpStr, pDnode->ep, TSDB_EP_LEN) == 0) {
+      sdbCancelFetch(pSdb, pIter);
+      return pDnode;
+    }
+
+    sdbRelease(pSdb, pDnode);
+  }
+
+  return NULL;
+}
+
 int32_t mndGetDnodeSize(SMnode *pMnode) {
   SSdb *pSdb = pMnode->pSdb;
   return sdbGetSize(pSdb, SDB_DNODE);
@@ -340,12 +371,22 @@ static int32_t mndProcessStatusReq(SRpcMsg *pReq) {
   } else {
     pDnode = mndAcquireDnode(pMnode, statusReq.dnodeId);
     if (pDnode == NULL) {
+      int32_t err = terrno;
       pDnode = mndAcquireDnodeByEp(pMnode, statusReq.dnodeEp);
       if (pDnode != NULL) {
         pDnode->offlineReason = DND_REASON_DNODE_ID_NOT_MATCH;
+        terrno = err;
+        goto _OVER;
+      }
+
+      mError("dnode:%d, %s not exist, code:0x%x", statusReq.dnodeId, statusReq.dnodeEp, err);
+      if (err == TSDB_CODE_MND_DNODE_NOT_EXIST) {
+        terrno = err;
+        goto _OVER;
+      } else {
+        pDnode = mndAcquireDnodeAllStatusByEp(pMnode, statusReq.dnodeEp);
+        if (pDnode == NULL) goto _OVER;
       }
-      mError("dnode:%d, %s not exist", statusReq.dnodeId, statusReq.dnodeEp);
-      goto _OVER;
     }
   }
 
@@ -356,8 +397,6 @@ static int32_t mndProcessStatusReq(SRpcMsg *pReq) {
   bool    reboot = (pDnode->rebootTime != statusReq.rebootTime);
   bool    needCheck = !online || dnodeChanged || reboot;
 
-  pDnode->accessTimes++;
-  pDnode->lastAccessTime = curMs;
   const STraceId *trace = &pReq->info.traceId;
   mGTrace("dnode:%d, status received, accessTimes:%d check:%d online:%d reboot:%d changed:%d statusSeq:%d", pDnode->id,
           pDnode->accessTimes, needCheck, online, reboot, dnodeChanged, statusReq.statusSeq);
@@ -493,6 +532,8 @@ static int32_t mndProcessStatusReq(SRpcMsg *pReq) {
     pReq->info.rsp = pHead;
   }
 
+  pDnode->accessTimes++;
+  pDnode->lastAccessTime = curMs;
   code = 0;
 
 _OVER:
@@ -517,6 +558,7 @@ static int32_t mndCreateDnode(SMnode *pMnode, SRpcMsg *pReq, SCreateDnodeReq *pC
   pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_GLOBAL, pReq, "create-dnode");
   if (pTrans == NULL) goto _OVER;
   mInfo("trans:%d, used to create dnode:%s", pTrans->id, dnodeObj.ep);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   pRaw = mndDnodeActionEncode(&dnodeObj);
   if (pRaw == NULL || mndTransAppendCommitlog(pTrans, pRaw) != 0) goto _OVER;
@@ -698,6 +740,7 @@ static int32_t mndDropDnode(SMnode *pMnode, SRpcMsg *pReq, SDnodeObj *pDnode, SM
   if (pTrans == NULL) goto _OVER;
   mndTransSetSerial(pTrans);
   mInfo("trans:%d, used to drop dnode:%d, force:%d", pTrans->id, pDnode->id, force);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   pRaw = mndDnodeActionEncode(pDnode);
   if (pRaw == NULL) goto _OVER;
@@ -762,11 +805,12 @@ static int32_t mndProcessDropDnodeReq(SRpcMsg *pReq) {
 
   pDnode = mndAcquireDnode(pMnode, dropReq.dnodeId);
   if (pDnode == NULL) {
-    char ep[TSDB_EP_LEN + 1] = {0};
+    int32_t err = terrno;
+    char    ep[TSDB_EP_LEN + 1] = {0};
     snprintf(ep, sizeof(ep), dropReq.fqdn, dropReq.port);
     pDnode = mndAcquireDnodeByEp(pMnode, ep);
     if (pDnode == NULL) {
-      terrno = TSDB_CODE_MND_DNODE_NOT_EXIST;
+      terrno = err;
       goto _OVER;
     }
   }
@@ -967,11 +1011,12 @@ static int32_t mndRetrieveDnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
   SSdb      *pSdb = pMnode->pSdb;
   int32_t    numOfRows = 0;
   int32_t    cols = 0;
+  ESdbStatus objStatus = 0;
   SDnodeObj *pDnode = NULL;
   int64_t    curMs = taosGetTimestampMs();
 
   while (numOfRows < rows) {
-    pShow->pIter = sdbFetch(pSdb, SDB_DNODE, pShow->pIter, (void **)&pDnode);
+    pShow->pIter = sdbFetchAll(pSdb, SDB_DNODE, pShow->pIter, (void **)&pDnode, &objStatus, true);
     if (pShow->pIter == NULL) break;
     bool online = mndIsDnodeOnline(pDnode, curMs);
 
@@ -993,8 +1038,20 @@ static int32_t mndRetrieveDnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, (const char *)&pDnode->numOfSupportVnodes, false);
 
+    const char *status = "ready";
+    if (objStatus == SDB_STATUS_CREATING) status = "creating";
+    if (objStatus == SDB_STATUS_DROPPING) status = "dropping";
+    if (!online) {
+      if (objStatus == SDB_STATUS_CREATING)
+        status = "creating*";
+      else if (objStatus == SDB_STATUS_DROPPING)
+        status = "dropping*";
+      else
+        status = "offline";
+    }
+
     char b1[9] = {0};
-    STR_TO_VARSTR(b1, online ? "ready" : "offline");
+    STR_TO_VARSTR(b1, status);
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, numOfRows, b1, false);
 

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

@@ -390,6 +390,7 @@ static int32_t mndCreateMnode(SMnode *pMnode, SRpcMsg *pReq, SDnodeObj *pDnode,
   if (pTrans == NULL) goto _OVER;
   mndTransSetSerial(pTrans);
   mInfo("trans:%d, used to create mnode:%d", pTrans->id, pCreate->dnodeId);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   if (mndSetCreateMnodeRedoActions(pMnode, pTrans, pDnode, &mnodeObj) != 0) goto _OVER;
   if (mndSetCreateMnodeRedoLogs(pMnode, pTrans, &mnodeObj) != 0) goto _OVER;
@@ -526,6 +527,7 @@ static int32_t mndDropMnode(SMnode *pMnode, SRpcMsg *pReq, SMnodeObj *pObj) {
   if (pTrans == NULL) goto _OVER;
   mndTransSetSerial(pTrans);
   mInfo("trans:%d, used to drop mnode:%d", pTrans->id, pObj->id);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   if (mndSetDropMnodeInfoToTrans(pMnode, pTrans, pObj, false) != 0) goto _OVER;
   if (mndTransPrepare(pMnode, pTrans) != 0) goto _OVER;
@@ -633,6 +635,7 @@ static int32_t mndRetrieveMnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
     const char *status = "ready";
     if (objStatus == SDB_STATUS_CREATING) status = "creating";
     if (objStatus == SDB_STATUS_DROPPING) status = "dropping";
+    if (!mndIsDnodeOnline(pObj->pDnode, curMs)) status = "offline";
     char b3[9 + VARSTR_HEADER_SIZE] = {0};
     STR_WITH_MAXSIZE_TO_VARSTR(b3, status, pShow->pMeta->pSchemas[cols].bytes);
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);

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

@@ -595,6 +595,8 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_DB, pReq, "create-sma");
   if (pTrans == NULL) goto _OVER;
   mndTransSetDbName(pTrans, pDb->name, NULL);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
+
   mndTransSetSerial(pTrans);
   mInfo("trans:%d, used to create sma:%s stream:%s", pTrans->id, pCreate->name, streamObj.name);
 
@@ -809,6 +811,8 @@ static int32_t mndDropSma(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SSmaObj *p
 
   mInfo("trans:%d, used to drop sma:%s", pTrans->id, pSma->name);
   mndTransSetDbName(pTrans, pDb->name, NULL);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
+
   mndTransSetSerial(pTrans);
 
   char streamName[TSDB_TABLE_FNAME_LEN] = {0};

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

@@ -823,6 +823,7 @@ _OVER:
 
 int32_t mndAddStbToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SStbObj *pStb) {
   mndTransSetDbName(pTrans, pDb->name, pStb->name);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) return -1;
   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;
@@ -1856,6 +1857,7 @@ static int32_t mndAlterStbImp(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SStbOb
 
   mInfo("trans:%d, used to alter stb:%s", pTrans->id, pStb->name);
   mndTransSetDbName(pTrans, pDb->name, pStb->name);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   if (needRsp) {
     void   *pCont = NULL;
@@ -2055,6 +2057,7 @@ static int32_t mndDropStb(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SStbObj *p
 
   mInfo("trans:%d, used to drop stb:%s", pTrans->id, pStb->name);
   mndTransSetDbName(pTrans, pDb->name, pStb->name);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
 
   if (mndSetDropStbRedoLogs(pMnode, pTrans, pStb) != 0) goto _OVER;
   if (mndSetDropStbCommitLogs(pMnode, pTrans, pStb) != 0) goto _OVER;

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

@@ -631,11 +631,13 @@ static int32_t mndProcessCreateStreamReq(SRpcMsg *pReq) {
     mError("stream:%s, failed to create since %s", createStreamReq.name, terrstr());
     goto _OVER;
   }
-  mndTransSetDbName(pTrans, createStreamReq.sourceDB, streamObj.targetDb);
   mInfo("trans:%d, used to create stream:%s", pTrans->id, createStreamReq.name);
 
+  mndTransSetDbName(pTrans, createStreamReq.sourceDB, streamObj.targetDb);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) goto _OVER;
+
   // create stb for stream
-  if (mndCreateStbForStream(pMnode, pTrans, &streamObj, pReq->info.conn.user) < 0) {
+  if (createStreamReq.createStb == STREAM_CREATE_STABLE_TRUE && 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;
@@ -781,6 +783,12 @@ static int32_t mndProcessStreamDoCheckpoint(SRpcMsg *pReq) {
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_DB_INSIDE, pReq, "stream-checkpoint");
   if (pTrans == NULL) return -1;
   mndTransSetDbName(pTrans, pStream->sourceDb, pStream->targetDb);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) {
+    mndReleaseStream(pMnode, pStream);
+    mndTransDrop(pTrans);
+    return -1;
+  }
+
   taosRLockLatch(&pStream->lock);
   // 1. redo action: broadcast checkpoint source msg for all source vg
   int32_t totLevel = taosArrayGetSize(pStream->tasks);
@@ -871,11 +879,11 @@ static int32_t mndProcessDropStreamReq(SRpcMsg *pReq) {
   }
 
   if (mndCheckDbPrivilegeByName(pMnode, pReq->info.conn.user, MND_OPER_WRITE_DB, pStream->targetDb) != 0) {
+    sdbRelease(pMnode->pSdb, pStream);
     return -1;
   }
 
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_DB_INSIDE, pReq, "drop-stream");
-  mndTransSetDbName(pTrans, pStream->sourceDb, pStream->targetDb);
   if (pTrans == NULL) {
     mError("stream:%s, failed to drop since %s", dropReq.name, terrstr());
     sdbRelease(pMnode->pSdb, pStream);
@@ -883,6 +891,13 @@ static int32_t mndProcessDropStreamReq(SRpcMsg *pReq) {
   }
   mInfo("trans:%d, used to drop stream:%s", pTrans->id, dropReq.name);
 
+  mndTransSetDbName(pTrans, pStream->sourceDb, pStream->targetDb);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) {
+    sdbRelease(pMnode->pSdb, pStream);
+    mndTransDrop(pTrans);
+    return -1;
+  }
+
   // drop all tasks
   if (mndDropStreamTasks(pMnode, pTrans, pStream) < 0) {
     mError("stream:%s, failed to drop task since %s", dropReq.name, terrstr());

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

@@ -442,7 +442,12 @@ 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, "tmq-reb");
   if (pTrans == NULL) return -1;
+
   mndTransSetDbName(pTrans, pOutput->pSub->dbName, NULL);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) {
+    mndTransDrop(pTrans);
+    return -1;
+  }
 
   // make txn:
   // 1. redo action: action to all vg

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

@@ -142,10 +142,9 @@ int32_t mndSyncGetSnapshot(const SSyncFSM *pFsm, SSnapshot *pSnapshot, void *pRe
   return 0;
 }
 
-int32_t mndSyncGetSnapshotInfo(const SSyncFSM *pFsm, SSnapshot *pSnapshot) {
+static void mndSyncGetSnapshotInfo(const SSyncFSM *pFsm, SSnapshot *pSnapshot) {
   SMnode *pMnode = pFsm->data;
   sdbGetCommitInfo(pMnode->pSdb, &pSnapshot->lastApplyIndex, &pSnapshot->lastApplyTerm, &pSnapshot->lastConfigIndex);
-  return 0;
 }
 
 void mndRestoreFinish(const SSyncFSM *pFsm) {
@@ -170,10 +169,10 @@ int32_t mndSnapshotStartRead(const SSyncFSM *pFsm, void *pParam, void **ppReader
   return sdbStartRead(pMnode->pSdb, (SSdbIter **)ppReader, NULL, NULL, NULL);
 }
 
-int32_t mndSnapshotStopRead(const SSyncFSM *pFsm, void *pReader) {
+static void mndSnapshotStopRead(const SSyncFSM *pFsm, void *pReader) {
   mInfo("stop to read snapshot from sdb");
   SMnode *pMnode = pFsm->data;
-  return sdbStopRead(pMnode->pSdb, pReader);
+  sdbStopRead(pMnode->pSdb, pReader);
 }
 
 int32_t mndSnapshotDoRead(const SSyncFSM *pFsm, void *pReader, void **ppBuf, int32_t *len) {

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

@@ -701,13 +701,19 @@ static int32_t mndProcessDropTopicReq(SRpcMsg *pReq) {
   }
 
   STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_DB_INSIDE, pReq, "drop-topic");
-  mndTransSetDbName(pTrans, pTopic->db, NULL);
   if (pTrans == NULL) {
     mError("topic:%s, failed to drop since %s", pTopic->name, terrstr());
     mndReleaseTopic(pMnode, pTopic);
     return -1;
   }
 
+  mndTransSetDbName(pTrans, pTopic->db, NULL);
+  if (mndTrancCheckConflict(pMnode, pTrans) != 0) {
+    mndReleaseTopic(pMnode, pTopic);
+    mndTransDrop(pTrans);
+    return -1;
+  }
+
   mInfo("trans:%d, used to drop topic:%s", pTrans->id, pTopic->name);
 
   // TODO check if rebalancing

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

@@ -957,9 +957,13 @@ static void mndTransSendRpcRsp(SMnode *pMnode, STrans *pTrans) {
   for (int32_t i = 0; i < size; ++i) {
     SRpcHandleInfo *pInfo = taosArrayGet(pTrans->pRpcArray, i);
     if (pInfo->handle != NULL) {
-      if (code == TSDB_CODE_RPC_NETWORK_UNAVAIL) {
+      if (code == TSDB_CODE_RPC_NETWORK_UNAVAIL || code == TSDB_CODE_RPC_SOMENODE_NOT_CONNECTED) {
         code = TSDB_CODE_MND_TRANS_NETWORK_UNAVAILL;
       }
+      if (code == TSDB_CODE_SYN_TIMEOUT) {
+        code = TSDB_CODE_MND_TRNAS_SYNC_TIMEOUT;
+      }
+
       if (i != 0 && code == 0) {
         code = TSDB_CODE_MNODE_NOT_FOUND;
       }

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

@@ -425,6 +425,7 @@ void *mndBuildDropVnodeReq(SMnode *pMnode, SDnodeObj *pDnode, SDbObj *pDb, SVgOb
 static bool mndResetDnodesArrayFp(SMnode *pMnode, void *pObj, void *p1, void *p2, void *p3) {
   SDnodeObj *pDnode = pObj;
   pDnode->numOfVnodes = 0;
+  pDnode->numOfOtherNodes = 0;
   return true;
 }
 
@@ -447,7 +448,7 @@ static bool mndBuildDnodesArrayFp(SMnode *pMnode, void *pObj, void *p1, void *p2
         pDnode->numOfVnodes, pDnode->numOfSupportVnodes, isMnode, online, pDnode->memAvail, pDnode->memUsed);
 
   if (isMnode) {
-    pDnode->numOfVnodes++;
+    pDnode->numOfOtherNodes++;
   }
 
   if (online && pDnode->numOfSupportVnodes > 0) {
@@ -468,14 +469,25 @@ SArray *mndBuildDnodesArray(SMnode *pMnode, int32_t exceptDnodeId) {
 
   sdbTraverse(pSdb, SDB_DNODE, mndResetDnodesArrayFp, NULL, NULL, NULL);
   sdbTraverse(pSdb, SDB_DNODE, mndBuildDnodesArrayFp, pArray, &exceptDnodeId, NULL);
+
+  mDebug("build %d dnodes array", (int32_t)taosArrayGetSize(pArray));
+  for (int32_t i = 0; i < (int32_t)taosArrayGetSize(pArray); ++i) {
+    SDnodeObj *pDnode = taosArrayGet(pArray, i);
+    mDebug("dnode:%d, vnodes:%d others:%d", pDnode->id, pDnode->numOfVnodes, pDnode->numOfOtherNodes);
+  }
   return pArray;
 }
 
 static int32_t mndCompareDnodeId(int32_t *dnode1Id, int32_t *dnode2Id) { return *dnode1Id >= *dnode2Id ? 1 : 0; }
 
+static float mndGetDnodeScore(SDnodeObj *pDnode, int32_t additionDnodes, float ratio) {
+  float totalDnodes = pDnode->numOfVnodes + (float)pDnode->numOfOtherNodes * ratio + additionDnodes;
+  return totalDnodes / pDnode->numOfSupportVnodes;
+}
+
 static int32_t mndCompareDnodeVnodes(SDnodeObj *pDnode1, SDnodeObj *pDnode2) {
-  float d1Score = (float)pDnode1->numOfVnodes / pDnode1->numOfSupportVnodes;
+  float d1Score = mndGetDnodeScore(pDnode1, 0, 0.9);
-  float d2Score = (float)pDnode2->numOfVnodes / pDnode2->numOfSupportVnodes;
+  float d2Score = mndGetDnodeScore(pDnode2, 0, 0.9);
   return d1Score >= d2Score ? 1 : 0;
 }
 
@@ -494,7 +506,12 @@ static int32_t mndGetAvailableDnode(SMnode *pMnode, SDbObj *pDb, SVgObj *pVgroup
   int32_t allocedVnodes = 0;
   void   *pIter = NULL;
 
+  mDebug("start to sort %d dnodes", (int32_t)taosArrayGetSize(pArray));
   taosArraySort(pArray, (__compar_fn_t)mndCompareDnodeVnodes);
+  for (int32_t i = 0; i < (int32_t)taosArrayGetSize(pArray); ++i) {
+    SDnodeObj *pDnode = taosArrayGet(pArray, i);
+    mDebug("dnode:%d, score:%f", pDnode->id, mndGetDnodeScore(pDnode, 0, 0.9));
+  }
 
   int32_t size = taosArrayGetSize(pArray);
   if (size < pVgroup->replica) {
@@ -875,7 +892,7 @@ static int32_t mndAddVnodeToVgroup(SMnode *pMnode, STrans *pTrans, SVgObj *pVgro
   taosArraySort(pArray, (__compar_fn_t)mndCompareDnodeVnodes);
   for (int32_t i = 0; i < taosArrayGetSize(pArray); ++i) {
     SDnodeObj *pDnode = taosArrayGet(pArray, i);
-    mInfo("dnode:%d, equivalent vnodes:%d", pDnode->id, pDnode->numOfVnodes);
+    mInfo("dnode:%d, equivalent vnodes:%d others:%d", pDnode->id, pDnode->numOfVnodes, pDnode->numOfOtherNodes);
   }
 
   SVnodeGid *pVgid = &pVgroup->vnodeGid[pVgroup->replica];
@@ -935,7 +952,7 @@ static int32_t mndRemoveVnodeFromVgroup(SMnode *pMnode, STrans *pTrans, SVgObj *
   taosArraySort(pArray, (__compar_fn_t)mndCompareDnodeVnodes);
   for (int32_t i = 0; i < taosArrayGetSize(pArray); ++i) {
     SDnodeObj *pDnode = taosArrayGet(pArray, i);
-    mInfo("dnode:%d, equivalent vnodes:%d", pDnode->id, pDnode->numOfVnodes);
+    mInfo("dnode:%d, equivalent vnodes:%d others:%d", pDnode->id, pDnode->numOfVnodes, pDnode->numOfOtherNodes);
   }
 
   int32_t code = -1;
@@ -1970,16 +1987,16 @@ static int32_t mndBalanceVgroup(SMnode *pMnode, SRpcMsg *pReq, SArray *pArray) {
     taosArraySort(pArray, (__compar_fn_t)mndCompareDnodeVnodes);
     for (int32_t i = 0; i < taosArrayGetSize(pArray); ++i) {
       SDnodeObj *pDnode = taosArrayGet(pArray, i);
-      mInfo("dnode:%d, equivalent vnodes:%d support:%d, score:%f", pDnode->id, pDnode->numOfVnodes,
+      mInfo("dnode:%d, equivalent vnodes:%d others:%d support:%d, score:%f", pDnode->id, pDnode->numOfVnodes,
-            pDnode->numOfSupportVnodes, (float)pDnode->numOfVnodes / pDnode->numOfSupportVnodes);
+            pDnode->numOfSupportVnodes, pDnode->numOfOtherNodes, mndGetDnodeScore(pDnode, 0, 1));
     }
 
     SDnodeObj *pSrc = taosArrayGet(pArray, taosArrayGetSize(pArray) - 1);
     SDnodeObj *pDst = taosArrayGet(pArray, 0);
 
-    float srcScore = (float)(pSrc->numOfVnodes - 1) / pSrc->numOfSupportVnodes;
+    float srcScore = mndGetDnodeScore(pSrc, -1, 1);
-    float dstScore = (float)(pDst->numOfVnodes + 1) / pDst->numOfSupportVnodes;
+    float dstScore = mndGetDnodeScore(pDst, 1, 1);
-    mInfo("trans:%d, after balance, src dnode:%d score:%f, dst dnode:%d score:%f", pTrans->id, pSrc->id, srcScore,
+    mInfo("trans:%d, after balance, src dnode:%d score:%f, dst dnode:%d score:%f", pTrans->id, pSrc->id, dstScore,
           pDst->id, dstScore);
 
     if (srcScore > dstScore - 0.000001) {

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

@@ -392,7 +392,7 @@ void    *sdbGetRowObj(SSdbRow *pRow);
 void     sdbFreeRow(SSdb *pSdb, SSdbRow *pRow, bool callFunc);
 
 int32_t sdbStartRead(SSdb *pSdb, SSdbIter **ppIter, int64_t *index, int64_t *term, int64_t *config);
-int32_t sdbStopRead(SSdb *pSdb, SSdbIter *pIter);
+void    sdbStopRead(SSdb *pSdb, SSdbIter *pIter);
 int32_t sdbDoRead(SSdb *pSdb, SSdbIter *pIter, void **ppBuf, int32_t *len);
 
 int32_t sdbStartWrite(SSdb *pSdb, SSdbIter **ppIter);

source/dnode/mnode/sdb/src/sdbFile.c

@@ -585,10 +585,7 @@ int32_t sdbStartRead(SSdb *pSdb, SSdbIter **ppIter, int64_t *index, int64_t *ter
   return 0;
 }
 
-int32_t sdbStopRead(SSdb *pSdb, SSdbIter *pIter) {
+void sdbStopRead(SSdb *pSdb, SSdbIter *pIter) { sdbCloseIter(pIter); }
-  sdbCloseIter(pIter);
-  return 0;
-}
 
 int32_t sdbDoRead(SSdb *pSdb, SSdbIter *pIter, void **ppBuf, int32_t *len) {
   int32_t maxlen = 4096;

source/dnode/vnode/inc/vnode.h

@@ -185,7 +185,7 @@ void        *tsdbGetIvtIdx(SMeta *pMeta);
 uint64_t     getReaderMaxVersion(STsdbReader *pReader);
 
 int32_t tsdbCacherowsReaderOpen(void *pVnode, int32_t type, void *pTableIdList, int32_t numOfTables, int32_t numOfCols,
-                                uint64_t suid, void **pReader);
+                                uint64_t suid, void **pReader, const char* idstr);
 int32_t tsdbRetrieveCacheRows(void *pReader, SSDataBlock *pResBlock, const int32_t *slotIds, SArray *pTableUids);
 void   *tsdbCacherowsReaderClose(void *pReader);
 int32_t tsdbGetTableSchema(SVnode *pVnode, int64_t uid, STSchema **pSchema, int64_t *suid);
@@ -275,7 +275,7 @@ int32_t smaGetTSmaDays(SVnodeCfg *pCfg, void *pCont, uint32_t contLen, int32_t *
 
 // SVSnapReader
 int32_t vnodeSnapReaderOpen(SVnode *pVnode, int64_t sver, int64_t ever, SVSnapReader **ppReader);
-int32_t vnodeSnapReaderClose(SVSnapReader *pReader);
+void    vnodeSnapReaderClose(SVSnapReader *pReader);
 int32_t vnodeSnapRead(SVSnapReader *pReader, uint8_t **ppData, uint32_t *nData);
 // SVSnapWriter
 int32_t vnodeSnapWriterOpen(SVnode *pVnode, int64_t sver, int64_t ever, SVSnapWriter **ppWriter);

source/dnode/vnode/src/inc/tq.h

@@ -121,6 +121,7 @@ typedef struct {
 struct STQ {
   SVnode* pVnode;
   char*   path;
+  int64_t walLogLastVer;
 
   SRWLatch pushLock;
 

source/dnode/vnode/src/inc/tsdb.h

@@ -726,21 +726,21 @@ void              *destroyLastBlockLoadInfo(SSttBlockLoadInfo *pLoadInfo);
 
 // tsdbCache ==============================================================================================
 typedef struct SCacheRowsReader {
-  SVnode   *pVnode;
+  SVnode            *pVnode;
-  STSchema *pSchema;
+  STSchema          *pSchema;
-  uint64_t  uid;
+  uint64_t           uid;
-  uint64_t  suid;
+  uint64_t           suid;
-  char    **transferBuf;  // todo remove it soon
+  char             **transferBuf;  // todo remove it soon
-  int32_t   numOfCols;
+  int32_t            numOfCols;
-  int32_t   type;
+  int32_t            type;
-  int32_t   tableIndex;  // currently returned result tables
+  int32_t            tableIndex;  // currently returned result tables
-
   STableKeyInfo     *pTableList;  // table id list
   int32_t            numOfTables;
   SSttBlockLoadInfo *pLoadInfo;
   STsdbReadSnap     *pReadSnap;
   SDataFReader      *pDataFReader;
   SDataFReader      *pDataFReaderLast;
+  const char        *idstr;
 } SCacheRowsReader;
 
 typedef struct {

source/dnode/vnode/src/inc/vnodeInt.h

@@ -90,7 +90,9 @@ typedef struct SCommitInfo        SCommitInfo;
 #define VND_INFO_FNAME "vnode.json"
 
 // vnd.h
-void* vnodeBufPoolMalloc(SVBufPool* pPool, int32_t size);
+
+void* vnodeBufPoolMalloc(SVBufPool* pPool, int size);
+void* vnodeBufPoolMallocAligned(SVBufPool* pPool, int size);
 void  vnodeBufPoolFree(SVBufPool* pPool, void* p);
 void  vnodeBufPoolRef(SVBufPool* pPool);
 void  vnodeBufPoolUnRef(SVBufPool* pPool);
@@ -199,6 +201,7 @@ int32_t tqProcessTaskRecover1Req(STQ* pTq, SRpcMsg* pMsg);
 int32_t tqProcessTaskRecover2Req(STQ* pTq, int64_t version, char* msg, int32_t msgLen);
 int32_t tqProcessTaskRecoverFinishReq(STQ* pTq, SRpcMsg* pMsg);
 int32_t tqProcessTaskRecoverFinishRsp(STQ* pTq, SRpcMsg* pMsg);
+int32_t tqCheckLogInWal(STQ* pTq, int64_t version);
 
 int32_t tqBlockToSubmit(SVnode* pVnode, const SArray* pBlocks, const STSchema* pSchema,
                         SSchemaWrapper* pTagSchemaWrapper, bool createTb, int64_t suid, const char* stbFullName,

source/dnode/vnode/src/meta/metaCache.c

@@ -54,9 +54,9 @@ struct SMetaCache {
 
   // query cache
   struct STagFilterResCache {
+    TdThreadMutex lock;
     SHashObj*  pTableEntry;
     SLRUCache* pUidResCache;
-    uint64_t   keyBuf[3];
   } sTagFilterResCache;
 };
 
@@ -140,6 +140,8 @@ int32_t metaCacheOpen(SMeta* pMeta) {
   }
 
   taosHashSetFreeFp(pCache->sTagFilterResCache.pTableEntry, freeCacheEntryFp);
+  taosThreadMutexInit(&pCache->sTagFilterResCache.lock, NULL);
+
   pMeta->pCache = pCache;
   return code;
 
@@ -159,6 +161,8 @@ void metaCacheClose(SMeta* pMeta) {
 
     taosHashCleanup(pMeta->pCache->sTagFilterResCache.pTableEntry);
     taosLRUCacheCleanup(pMeta->pCache->sTagFilterResCache.pUidResCache);
+    taosThreadMutexDestroy(&pMeta->pCache->sTagFilterResCache.lock);
+
     taosMemoryFree(pMeta->pCache);
     pMeta->pCache = NULL;
   }
@@ -206,7 +210,7 @@ _exit:
 int32_t metaCacheUpsert(SMeta* pMeta, SMetaInfo* pInfo) {
   int32_t code = 0;
 
-  // ASSERT(metaIsWLocked(pMeta));
+  // meta is wlocked for calling this func.
 
   // search
   SMetaCache*       pCache = pMeta->pCache;
@@ -217,7 +221,10 @@ int32_t metaCacheUpsert(SMeta* pMeta, SMetaInfo* pInfo) {
   }
 
   if (*ppEntry) {  // update
-    ASSERT(pInfo->suid == (*ppEntry)->info.suid);
+    if (pInfo->suid != (*ppEntry)->info.suid) {
+      metaError("meta/cache: suid should be same as the one in cache.");
+      return TSDB_CODE_FAILED;
+    }
     if (pInfo->version > (*ppEntry)->info.version) {
       (*ppEntry)->info.version = pInfo->version;
       (*ppEntry)->info.skmVer = pInfo->skmVer;
@@ -336,7 +343,7 @@ _exit:
 int32_t metaStatsCacheUpsert(SMeta* pMeta, SMetaStbStats* pInfo) {
   int32_t code = 0;
 
-  // ASSERT(metaIsWLocked(pMeta));
+  // meta is wlocked for calling this func.
 
   // search
   SMetaCache*          pCache = pMeta->pCache;
@@ -422,63 +429,82 @@ int32_t metaStatsCacheGet(SMeta* pMeta, int64_t uid, SMetaStbStats* pInfo) {
 
 int32_t metaGetCachedTableUidList(SMeta* pMeta, tb_uid_t suid, const uint8_t* pKey, int32_t keyLen, SArray* pList1,
                                   bool* acquireRes) {
-  uint64_t* pBuf = pMeta->pCache->sTagFilterResCache.keyBuf;
-
   // generate the composed key for LRU cache
-  SLRUCache* pCache = pMeta->pCache->sTagFilterResCache.pUidResCache;
+  SLRUCache*     pCache = pMeta->pCache->sTagFilterResCache.pUidResCache;
+  SHashObj*      pTableMap = pMeta->pCache->sTagFilterResCache.pTableEntry;
+  TdThreadMutex* pLock = &pMeta->pCache->sTagFilterResCache.lock;
 
-  pBuf[0] = suid;
+  uint64_t buf[3] = {0};
-  memcpy(&pBuf[1], pKey, keyLen);
+  uint32_t times = 0;
+
+  *acquireRes = 0;
+  buf[0] = suid;
+  memcpy(&buf[1], pKey, keyLen);
+
+  taosThreadMutexLock(pLock);
 
   int32_t    len = keyLen + sizeof(uint64_t);
-  LRUHandle* pHandle = taosLRUCacheLookup(pCache, pBuf, len);
+  LRUHandle* pHandle = taosLRUCacheLookup(pCache, buf, len);
   if (pHandle == NULL) {
-    *acquireRes = 0;
+    taosThreadMutexUnlock(pLock);
     return TSDB_CODE_SUCCESS;
-  } else {  // do some book mark work after acquiring the filter result from cache
+  }
-    STagFilterResEntry** pEntry = taosHashGet(pMeta->pCache->sTagFilterResCache.pTableEntry, &suid, sizeof(uint64_t));
-    ASSERT(pEntry != NULL);
-    *acquireRes = 1;
 
-    const char* p = taosLRUCacheValue(pMeta->pCache->sTagFilterResCache.pUidResCache, pHandle);
+  // do some book mark work after acquiring the filter result from cache
-    int32_t     size = *(int32_t*)p;
+  STagFilterResEntry** pEntry = taosHashGet(pTableMap, &suid, sizeof(uint64_t));
-    taosArrayAddBatch(pList1, p + sizeof(int32_t), size);
+  if (NULL == pEntry) {
+    metaError("meta/cache: pEntry should not be NULL.");
+    return TSDB_CODE_FAILED;
+  }
 
-    (*pEntry)->qTimes += 1;
+  *acquireRes = 1;
-    taosLRUCacheRelease(pCache, pHandle, false);
 
-    // check if scanning all items are necessary or not
+  const char* p = taosLRUCacheValue(pCache, pHandle);
-    if ((*pEntry)->qTimes >= 5000 && TD_DLIST_NELES(&(*pEntry)->list) > 10) {
+  int32_t     size = *(int32_t*)p;
-      SArray* pList = taosArrayInit(64, POINTER_BYTES);
 
-      SListIter iter = {0};
+  // set the result into the buffer
-      tdListInitIter(&(*pEntry)->list, &iter, TD_LIST_FORWARD);
+  taosArrayAddBatch(pList1, p + sizeof(int32_t), size);
 
-      SListNode* pNode = NULL;
+  times = atomic_add_fetch_32(&(*pEntry)->qTimes, 1);
-      while ((pNode = tdListNext(&iter)) != NULL) {
+  taosLRUCacheRelease(pCache, pHandle, false);
-        memcpy(&pBuf[1], pNode->data, keyLen);
 
-        // check whether it is existed in LRU cache, and remove it from linked list if not.
+  // unlock meta
-        LRUHandle* pRes = taosLRUCacheLookup(pCache, pBuf, len);
+  taosThreadMutexUnlock(pLock);
-        if (pRes == NULL) {  // remove the item in the linked list
+
-          taosArrayPush(pList, &pNode);
+  // check if scanning all items are necessary or not
-        } else {
+  if (times >= 5000 && TD_DLIST_NELES(&(*pEntry)->list) > 10) {
-          taosLRUCacheRelease(pCache, pRes, false);
+    taosThreadMutexLock(pLock);
-        }
+
+    SArray* pInvalidRes = taosArrayInit(64, POINTER_BYTES);
+
+    SListIter iter = {0};
+    tdListInitIter(&(*pEntry)->list, &iter, TD_LIST_FORWARD);
+
+    SListNode* pNode = NULL;
+    while ((pNode = tdListNext(&iter)) != NULL) {
+      memcpy(&buf[1], pNode->data, keyLen);
+
+      // check whether it is existed in LRU cache, and remove it from linked list if not.
+      LRUHandle* pRes = taosLRUCacheLookup(pCache, buf, len);
+      if (pRes == NULL) {  // remove the item in the linked list
+        taosArrayPush(pInvalidRes, &pNode);
+      } else {
+        taosLRUCacheRelease(pCache, pRes, false);
       }
-
-      // remove the keys, of which query uid lists have been replaced already.
-      size_t s = taosArrayGetSize(pList);
-      for (int32_t i = 0; i < s; ++i) {
-        SListNode** p1 = taosArrayGet(pList, i);
-        tdListPopNode(&(*pEntry)->list, *p1);
-        taosMemoryFree(*p1);
-      }
-
-      (*pEntry)->qTimes = 0;  // reset the query times
-
-      taosArrayDestroy(pList);
     }
+
+    // remove the keys, of which query uid lists have been replaced already.
+    size_t s = taosArrayGetSize(pInvalidRes);
+    for (int32_t i = 0; i < s; ++i) {
+      SListNode** p1 = taosArrayGet(pInvalidRes, i);
+      tdListPopNode(&(*pEntry)->list, *p1);
+      taosMemoryFree(*p1);
+    }
+
+    atomic_store_32(&(*pEntry)->qTimes, 0);  // reset the query times
+    taosArrayDestroy(pInvalidRes);
+
+    taosThreadMutexUnlock(pLock);
   }
 
   return TSDB_CODE_SUCCESS;
@@ -510,8 +536,11 @@ int32_t metaUidFilterCachePut(SMeta* pMeta, uint64_t suid, const void* pKey, int
     return TSDB_CODE_SUCCESS;
   }
 
-  SLRUCache* pCache = pMeta->pCache->sTagFilterResCache.pUidResCache;
+  SLRUCache*     pCache = pMeta->pCache->sTagFilterResCache.pUidResCache;
-  SHashObj*  pTableEntry = pMeta->pCache->sTagFilterResCache.pTableEntry;
+  SHashObj*      pTableEntry = pMeta->pCache->sTagFilterResCache.pTableEntry;
+  TdThreadMutex* pLock = &pMeta->pCache->sTagFilterResCache.lock;
+
+  taosThreadMutexLock(pLock);
 
   STagFilterResEntry** pEntry = taosHashGet(pTableEntry, &suid, sizeof(uint64_t));
   if (pEntry == NULL) {
@@ -524,15 +553,21 @@ int32_t metaUidFilterCachePut(SMeta* pMeta, uint64_t suid, const void* pKey, int
     tdListAppend(&(*pEntry)->list, pKey);
   }
 
-  uint64_t* pBuf = pMeta->pCache->sTagFilterResCache.keyBuf;
+  uint64_t buf[3] = {0};
-  pBuf[0] = suid;
+  buf[0] = suid;
 
-  memcpy(&pBuf[1], pKey, keyLen);
+  memcpy(&buf[1], pKey, keyLen);
-  ASSERT(sizeof(uint64_t) + keyLen == 24);
+  if (sizeof(uint64_t) + keyLen != 24) {
+    metaError("meta/cache: incorrect keyLen:%" PRId32 " length.", keyLen);
+    return TSDB_CODE_FAILED;
+  }
 
   // add to cache.
-  taosLRUCacheInsert(pCache, pBuf, sizeof(uint64_t) + keyLen, pPayload, payloadLen, freePayload, NULL,
+  taosLRUCacheInsert(pCache, buf, sizeof(uint64_t) + keyLen, pPayload, payloadLen, freePayload, NULL,
                      TAOS_LRU_PRIORITY_LOW);
+
+  taosThreadMutexUnlock(pLock);
+
   metaDebug("vgId:%d, suid:%" PRIu64 " list cache added into cache, total:%d, tables:%d", TD_VID(pMeta->pVnode), suid,
             (int32_t)taosLRUCacheGetUsage(pCache), taosHashGetSize(pTableEntry));
 
@@ -541,15 +576,19 @@ int32_t metaUidFilterCachePut(SMeta* pMeta, uint64_t suid, const void* pKey, int
 
 // remove the lru cache that are expired due to the tags value update, or creating, or dropping, of child tables
 int32_t metaUidCacheClear(SMeta* pMeta, uint64_t suid) {
-  STagFilterResEntry** pEntry = taosHashGet(pMeta->pCache->sTagFilterResCache.pTableEntry, &suid, sizeof(uint64_t));
-  if (pEntry == NULL || listNEles(&(*pEntry)->list) == 0) {
-    return TSDB_CODE_SUCCESS;
-  }
-
   int32_t  keyLen = sizeof(uint64_t) * 3;
   uint64_t p[3] = {0};
   p[0] = suid;
 
+  TdThreadMutex* pLock = &pMeta->pCache->sTagFilterResCache.lock;
+
+  taosThreadMutexLock(pLock);
+  STagFilterResEntry** pEntry = taosHashGet(pMeta->pCache->sTagFilterResCache.pTableEntry, &suid, sizeof(uint64_t));
+  if (pEntry == NULL || listNEles(&(*pEntry)->list) == 0) {
+    taosThreadMutexUnlock(pLock);
+    return TSDB_CODE_SUCCESS;
+  }
+
   SListIter iter = {0};
   tdListInitIter(&(*pEntry)->list, &iter, TD_LIST_FORWARD);
 
@@ -562,5 +601,6 @@ int32_t metaUidCacheClear(SMeta* pMeta, uint64_t suid) {
   (*pEntry)->qTimes = 0;
   tdListEmpty(&(*pEntry)->list);
 
+  taosThreadMutexUnlock(pLock);
   return TSDB_CODE_SUCCESS;
 }

source/dnode/vnode/src/meta/metaCommit.c

@@ -15,8 +15,10 @@
 
 #include "meta.h"
 
-static FORCE_INLINE void *metaMalloc(void *pPool, size_t size) { return vnodeBufPoolMalloc((SVBufPool *)pPool, size); }
+static FORCE_INLINE void *metaMalloc(void *pPool, size_t size) {
-static FORCE_INLINE void  metaFree(void *pPool, void *p) { vnodeBufPoolFree((SVBufPool *)pPool, p); }
+  return vnodeBufPoolMallocAligned((SVBufPool *)pPool, size);
+}
+static FORCE_INLINE void metaFree(void *pPool, void *p) { vnodeBufPoolFree((SVBufPool *)pPool, p); }
 
 // begin a meta txn
 int metaBegin(SMeta *pMeta, int8_t heap) {

source/dnode/vnode/src/meta/metaEntry.c

@@ -51,7 +51,9 @@ int metaEncodeEntry(SEncoder *pCoder, const SMetaEntry *pME) {
   } else if (pME->type == TSDB_TSMA_TABLE) {
     if (tEncodeTSma(pCoder, pME->smaEntry.tsma) < 0) return -1;
   } else {
-    ASSERT(0);
+    metaError("meta/entry: invalide table type: %" PRId8 " encode failed.", pME->type);
+
+    return -1;
   }
 
   tEndEncode(pCoder);
@@ -99,7 +101,9 @@ int metaDecodeEntry(SDecoder *pCoder, SMetaEntry *pME) {
     }
     if (tDecodeTSma(pCoder, pME->smaEntry.tsma, true) < 0) return -1;
   } else {
-    ASSERT(0);
+    metaError("meta/entry: invalide table type: %" PRId8 " decode failed.", pME->type);
+
+    return -1;
   }
 
   tEndDecode(pCoder);

source/dnode/vnode/src/meta/metaOpen.c

@@ -358,7 +358,10 @@ static int tagIdxKeyCmpr(const void *pKey1, int kLen1, const void *pKey2, int kL
     return -1;
   }
 
-  ASSERT(pTagIdxKey1->type == pTagIdxKey2->type);
+  if (pTagIdxKey1->type != pTagIdxKey2->type) {
+    metaError("meta/open: incorrect tag idx type.");
+    return TSDB_CODE_FAILED;
+  }
 
   // check NULL, NULL is always the smallest
   if (pTagIdxKey1->isNull && !pTagIdxKey2->isNull) {

source/dnode/vnode/src/meta/metaQuery.c

@@ -19,7 +19,7 @@ void metaReaderInit(SMetaReader *pReader, SMeta *pMeta, int32_t flags) {
   memset(pReader, 0, sizeof(*pReader));
   pReader->flags = flags;
   pReader->pMeta = pMeta;
-  if (!(flags & META_READER_NOLOCK)) {
+  if (pReader->pMeta && !(flags & META_READER_NOLOCK)) {
     metaRLock(pMeta);
   }
 }
@@ -152,7 +152,7 @@ bool metaIsTableExist(SMeta *pMeta, tb_uid_t uid) {
 }
 
 int metaGetTableEntryByUid(SMetaReader *pReader, tb_uid_t uid) {
-  SMeta *pMeta = pReader->pMeta;
+  SMeta  *pMeta = pReader->pMeta;
   int64_t version1;
 
   // query uid.idx
@@ -239,7 +239,6 @@ int metaGetTableSzNameByUid(void *meta, uint64_t uid, char *tbName) {
   return 0;
 }
 
-
 int metaGetTableUidByName(void *meta, char *tbName, uint64_t *uid) {
   int         code = 0;
   SMetaReader mr = {0};
@@ -653,7 +652,11 @@ int32_t metaGetTbTSchemaEx(SMeta *pMeta, tb_uid_t suid, tb_uid_t uid, int32_t sv
         goto _exit;
       }
 
-      ASSERT(c);
+      if (c == 0) {
+        metaError("meta/query: incorrect c: %" PRId32 ".", c);
+        code = TSDB_CODE_FAILED;
+        goto _exit;
+      }
 
       if (c < 0) {
         tdbTbcMoveToPrev(pSkmDbC);
@@ -677,7 +680,11 @@ int32_t metaGetTbTSchemaEx(SMeta *pMeta, tb_uid_t suid, tb_uid_t uid, int32_t sv
     }
   }
 
-  ASSERT(sver > 0);
+  if (sver <= 0) {
+    metaError("meta/query: incorrect sver: %" PRId32 ".", sver);
+    code = TSDB_CODE_FAILED;
+    goto _exit;
+  }
 
   skmDbKey.uid = suid ? suid : uid;
   skmDbKey.sver = sver;
@@ -737,9 +744,7 @@ int64_t metaGetTimeSeriesNum(SMeta *pMeta) {
   return pMeta->pVnode->config.vndStats.numOfTimeSeries + pMeta->pVnode->config.vndStats.numOfNTimeSeries;
 }
 
-int64_t metaGetNtbNum(SMeta *pMeta) {
+int64_t metaGetNtbNum(SMeta *pMeta) { return pMeta->pVnode->config.vndStats.numOfNTables; }
-  return pMeta->pVnode->config.vndStats.numOfNTables;
-}
 
 typedef struct {
   SMeta   *pMeta;

source/dnode/vnode/src/meta/metaSnapshot.c

@@ -100,7 +100,10 @@ int32_t metaSnapRead(SMetaSnapReader* pReader, uint8_t** ppData) {
     break;
   }
 
-  ASSERT(pData && nData);
+  if (!pData || !nData) {
+    metaError("meta/snap: invalide nData: %" PRId32 " meta snap read failed.", nData);
+    goto _exit;
+  }
 
   *ppData = taosMemoryMalloc(sizeof(SSnapDataHdr) + nData);
   if (*ppData == NULL) {
@@ -113,8 +116,8 @@ int32_t metaSnapRead(SMetaSnapReader* pReader, uint8_t** ppData) {
   pHdr->size = nData;
   memcpy(pHdr->data, pData, nData);
 
-  metaInfo("vgId:%d, vnode snapshot meta read data, version:%" PRId64 " uid:%" PRId64 " nData:%d",
+  metaDebug("vgId:%d, vnode snapshot meta read data, version:%" PRId64 " uid:%" PRId64 " blockLen:%d",
-           TD_VID(pReader->pMeta->pVnode), key.version, key.uid, nData);
+            TD_VID(pReader->pMeta->pVnode), key.version, key.uid, nData);
 
 _exit:
   return code;
@@ -356,7 +359,11 @@ int32_t buildSnapContext(SMeta* pMeta, int64_t snapVersion, int64_t suid, int8_t
   for (int i = 0; i < taosArrayGetSize(ctx->idList); i++) {
     int64_t* uid = taosArrayGet(ctx->idList, i);
     SIdInfo* idData = (SIdInfo*)taosHashGet(ctx->idVersion, uid, sizeof(int64_t));
-    ASSERT(idData);
+    if (!idData) {
+      metaError("meta/snap: null idData");
+      return TSDB_CODE_FAILED;
+    }
+
     idData->index = i;
     metaDebug("tmqsnap init idVersion uid:%" PRIi64 " version:%" PRIi64 " index:%d", *uid, idData->version,
               idData->index);
@@ -473,7 +480,10 @@ int32_t getMetafromSnapShot(SSnapContext* ctx, void** pBuf, int32_t* contLen, in
     int64_t* uidTmp = taosArrayGet(ctx->idList, ctx->index);
     ctx->index++;
     SIdInfo* idInfo = (SIdInfo*)taosHashGet(ctx->idVersion, uidTmp, sizeof(tb_uid_t));
-    ASSERT(idInfo);
+    if (!idInfo) {
+      metaError("meta/snap: null idInfo");
+      return TSDB_CODE_FAILED;
+    }
 
     *uid = *uidTmp;
     ret = MoveToPosition(ctx, idInfo->version, *uidTmp);
@@ -507,7 +517,11 @@ int32_t getMetafromSnapShot(SSnapContext* ctx, void** pBuf, int32_t* contLen, in
              (ctx->subType == TOPIC_SUB_TYPE__TABLE && me.type == TSDB_CHILD_TABLE && me.ctbEntry.suid == ctx->suid)) {
     STableInfoForChildTable* data =
         (STableInfoForChildTable*)taosHashGet(ctx->suidInfo, &me.ctbEntry.suid, sizeof(tb_uid_t));
-    ASSERT(data);
+    if (!data) {
+      metaError("meta/snap: null data");
+      return TSDB_CODE_FAILED;
+    }
+
     SVCreateTbReq req = {0};
 
     req.type = TSDB_CHILD_TABLE;
@@ -528,7 +542,8 @@ int32_t getMetafromSnapShot(SSnapContext* ctx, void** pBuf, int32_t* contLen, in
     } else {
       SArray* pTagVals = NULL;
       if (tTagToValArray((const STag*)p, &pTagVals) != 0) {
-        ASSERT(0);
+        metaError("meta/snap: tag to val array failed.");
+        return TSDB_CODE_FAILED;
       }
       int16_t nCols = taosArrayGetSize(pTagVals);
       for (int j = 0; j < nCols; ++j) {
@@ -572,7 +587,8 @@ int32_t getMetafromSnapShot(SSnapContext* ctx, void** pBuf, int32_t* contLen, in
     ret = buildNormalChildTableInfo(&req, pBuf, contLen);
     *type = TDMT_VND_CREATE_TABLE;
   } else {
-    ASSERT(0);
+    metaError("meta/snap: invalid topic sub type: %" PRId8 " get meta from snap failed.", ctx->subType);
+    ret = -1;
   }
   tDecoderClear(&dc);
 
@@ -593,7 +609,10 @@ SMetaTableInfo getUidfromSnapShot(SSnapContext* ctx) {
     int64_t* uidTmp = taosArrayGet(ctx->idList, ctx->index);
     ctx->index++;
     SIdInfo* idInfo = (SIdInfo*)taosHashGet(ctx->idVersion, uidTmp, sizeof(tb_uid_t));
-    ASSERT(idInfo);
+    if (!idInfo) {
+      metaError("meta/snap: null idInfo");
+      return result;
+    }
 
     int32_t ret = MoveToPosition(ctx, idInfo->version, *uidTmp);
     if (ret != 0) {

source/dnode/vnode/src/meta/metaTable.c

@@ -46,7 +46,7 @@ static void metaGetEntryInfo(const SMetaEntry *pEntry, SMetaInfo *pInfo) {
     pInfo->suid = 0;
     pInfo->skmVer = pEntry->ntbEntry.schemaRow.version;
   } else {
-    ASSERT(0);
+    metaError("meta/table: invalide table type: %" PRId8 " get entry info failed.", pEntry->type);
   }
 }
 
@@ -342,10 +342,18 @@ int metaAlterSTable(SMeta *pMeta, int64_t version, SVCreateStbReq *pReq) {
 
   tdbTbcOpen(pMeta->pTbDb, &pTbDbc, NULL);
   ret = tdbTbcMoveTo(pTbDbc, &((STbDbKey){.uid = pReq->suid, .version = oversion}), sizeof(STbDbKey), &c);
-  ASSERT(ret == 0 && c == 0);
+  if (!(ret == 0 && c == 0)) {
+    metaError("meta/table: invalide ret: %" PRId32 " or c: %" PRId32 "alter stb failed.", ret, c);
+    return -1;
+  }
 
   ret = tdbTbcGet(pTbDbc, NULL, NULL, &pData, &nData);
-  ASSERT(ret == 0);
+  if (ret < 0) {
+    tdbTbcClose(pTbDbc);
+
+    terrno = TSDB_CODE_TDB_STB_NOT_EXIST;
+    return -1;
+  }
 
   oStbEntry.pBuf = taosMemoryMalloc(nData);
   memcpy(oStbEntry.pBuf, pData, nData);
@@ -558,7 +566,8 @@ static void metaBuildTtlIdxKey(STtlIdxKey *ttlKey, const SMetaEntry *pME) {
     ctime = pME->ntbEntry.ctime;
     ttlDays = pME->ntbEntry.ttlDays;
   } else {
-    ASSERT(0);
+    metaError("meta/table: invalide table type: %" PRId8 " build ttl idx key failed.", pME->type);
+    return;
   }
 
   if (ttlDays <= 0) return;
@@ -710,8 +719,8 @@ int metaUpdateCtimeIdx(SMeta *pMeta, const SMetaEntry *pME) {
   if (metaBuildCtimeIdxKey(&ctimeKey, pME) < 0) {
     return 0;
   }
-  metaDebug("vgId:%d, start to save ctime:%" PRId64 " uid:%" PRId64 " ct:%" PRId64, TD_VID(pMeta->pVnode), pME->version,
+  metaTrace("vgId:%d, start to save version:%" PRId64 " uid:%" PRId64 " ctime:%" PRId64, TD_VID(pMeta->pVnode),
-            pME->uid, ctimeKey.ctime);
+            pME->version, pME->uid, ctimeKey.ctime);
 
   return tdbTbInsert(pMeta->pCtimeIdx, &ctimeKey, sizeof(ctimeKey), NULL, 0, pMeta->txn);
 }
@@ -773,7 +782,10 @@ static int metaAlterTableColumn(SMeta *pMeta, int64_t version, SVAlterTbReq *pAl
 
   tdbTbcOpen(pMeta->pUidIdx, &pUidIdxc, NULL);
   tdbTbcMoveTo(pUidIdxc, &uid, sizeof(uid), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " alt tb column failed.", c);
+    return -1;
+  }
 
   tdbTbcGet(pUidIdxc, NULL, NULL, &pData, &nData);
   oversion = ((SUidIdxVal *)pData)[0].version;
@@ -783,7 +795,11 @@ static int metaAlterTableColumn(SMeta *pMeta, int64_t version, SVAlterTbReq *pAl
 
   tdbTbcOpen(pMeta->pTbDb, &pTbDbc, NULL);
   tdbTbcMoveTo(pTbDbc, &((STbDbKey){.uid = uid, .version = oversion}), sizeof(STbDbKey), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " alt tb column failed.", c);
+    return -1;
+  }
+
   tdbTbcGet(pTbDbc, NULL, NULL, &pData, &nData);
 
   // get table entry
@@ -792,7 +808,11 @@ static int metaAlterTableColumn(SMeta *pMeta, int64_t version, SVAlterTbReq *pAl
   memcpy(entry.pBuf, pData, nData);
   tDecoderInit(&dc, entry.pBuf, nData);
   ret = metaDecodeEntry(&dc, &entry);
-  ASSERT(ret == 0);
+  if (ret != 0) {
+    tDecoderClear(&dc);
+    metaError("meta/table: invalide ret: %" PRId32 " alt tb column failed.", ret);
+    return -1;
+  }
 
   if (entry.type != TSDB_NORMAL_TABLE) {
     terrno = TSDB_CODE_VND_INVALID_TABLE_ACTION;
@@ -812,7 +832,11 @@ static int metaAlterTableColumn(SMeta *pMeta, int64_t version, SVAlterTbReq *pAl
     if (iCol >= pSchema->nCols) break;
     pColumn = &pSchema->pSchema[iCol];
 
-    ASSERT(pAlterTbReq->colName);
+    if (NULL == pAlterTbReq->colName) {
+      metaError("meta/table: null pAlterTbReq->colName");
+      return -1;
+    }
+
     if (strcmp(pColumn->name, pAlterTbReq->colName) == 0) break;
     iCol++;
   }
@@ -964,7 +988,10 @@ static int metaUpdateTableTagVal(SMeta *pMeta, int64_t version, SVAlterTbReq *pA
 
   tdbTbcOpen(pMeta->pUidIdx, &pUidIdxc, NULL);
   tdbTbcMoveTo(pUidIdxc, &uid, sizeof(uid), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " update tb tag val failed.", c);
+    return -1;
+  }
 
   tdbTbcGet(pUidIdxc, NULL, NULL, &pData, &nData);
   oversion = ((SUidIdxVal *)pData)[0].version;
@@ -977,7 +1004,11 @@ static int metaUpdateTableTagVal(SMeta *pMeta, int64_t version, SVAlterTbReq *pA
   /* get ctbEntry */
   tdbTbcOpen(pMeta->pTbDb, &pTbDbc, NULL);
   tdbTbcMoveTo(pTbDbc, &((STbDbKey){.uid = uid, .version = oversion}), sizeof(STbDbKey), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " update tb tag val failed.", c);
+    return -1;
+  }
+
   tdbTbcGet(pTbDbc, NULL, NULL, &pData, &nData);
 
   ctbEntry.pBuf = taosMemoryMalloc(nData);
@@ -1075,7 +1106,11 @@ static int metaUpdateTableTagVal(SMeta *pMeta, int64_t version, SVAlterTbReq *pA
     metaUpdateTagIdx(pMeta, &ctbEntry);
   }
 
-  ASSERT(ctbEntry.ctbEntry.pTags);
+  if (NULL == ctbEntry.ctbEntry.pTags) {
+    metaError("meta/table: null tags, update tag val failed.");
+    goto _err;
+  }
+
   SCtbIdxKey ctbIdxKey = {.suid = ctbEntry.ctbEntry.suid, .uid = uid};
   tdbTbUpsert(pMeta->pCtbIdx, &ctbIdxKey, sizeof(ctbIdxKey), ctbEntry.ctbEntry.pTags,
               ((STag *)(ctbEntry.ctbEntry.pTags))->len, pMeta->txn);
@@ -1130,7 +1165,10 @@ static int metaUpdateTableOptions(SMeta *pMeta, int64_t version, SVAlterTbReq *p
 
   tdbTbcOpen(pMeta->pUidIdx, &pUidIdxc, NULL);
   tdbTbcMoveTo(pUidIdxc, &uid, sizeof(uid), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " update tb options failed.", c);
+    return -1;
+  }
 
   tdbTbcGet(pUidIdxc, NULL, NULL, &pData, &nData);
   oversion = ((SUidIdxVal *)pData)[0].version;
@@ -1140,7 +1178,11 @@ static int metaUpdateTableOptions(SMeta *pMeta, int64_t version, SVAlterTbReq *p
 
   tdbTbcOpen(pMeta->pTbDb, &pTbDbc, NULL);
   tdbTbcMoveTo(pTbDbc, &((STbDbKey){.uid = uid, .version = oversion}), sizeof(STbDbKey), &c);
-  ASSERT(c == 0);
+  if (c != 0) {
+    metaError("meta/table: invalide c: %" PRId32 " update tb options failed.", c);
+    return -1;
+  }
+
   tdbTbcGet(pTbDbc, NULL, NULL, &pData, &nData);
 
   // get table entry
@@ -1149,7 +1191,11 @@ static int metaUpdateTableOptions(SMeta *pMeta, int64_t version, SVAlterTbReq *p
   memcpy(entry.pBuf, pData, nData);
   tDecoderInit(&dc, entry.pBuf, nData);
   ret = metaDecodeEntry(&dc, &entry);
-  ASSERT(ret == 0);
+  if (ret != 0) {
+    tDecoderClear(&dc);
+    metaError("meta/table: invalide ret: %" PRId32 " alt tb options failed.", ret);
+    return -1;
+  }
 
   entry.version = version;
   metaWLock(pMeta);
@@ -1408,7 +1454,8 @@ static int metaSaveToSkmDb(SMeta *pMeta, const SMetaEntry *pME) {
   } else if (pME->type == TSDB_NORMAL_TABLE) {
     pSW = &pME->ntbEntry.schemaRow;
   } else {
-    ASSERT(0);
+    metaError("meta/table: invalide table type: %" PRId8 " save skm db failed.", pME->type);
+    return TSDB_CODE_FAILED;
   }
 
   skmDbKey.uid = pME->uid;

source/dnode/vnode/src/sma/smaTimeRange.c

@@ -241,7 +241,7 @@ static int32_t tdProcessTSmaInsertImpl(SSma *pSma, int64_t indexUid, const char
   SRpcMsg submitReqMsg = {
       .msgType = TDMT_VND_SUBMIT,
       .pCont = pSubmitReq,
-      .contLen = ntohl(contLen),
+      .contLen = contLen,
   };
 
   if (tmsgPutToQueue(&pSma->pVnode->msgCb, WRITE_QUEUE, &submitReqMsg) < 0) {

source/dnode/vnode/src/tq/tq.c

@@ -80,6 +80,7 @@ STQ* tqOpen(const char* path, SVnode* pVnode) {
   }
   pTq->path = strdup(path);
   pTq->pVnode = pVnode;
+  pTq->walLogLastVer = pVnode->pWal->vers.lastVer;
 
   pTq->pHandle = taosHashInit(64, MurmurHash3_32, true, HASH_ENTRY_LOCK);
   taosHashSetFreeFp(pTq->pHandle, destroySTqHandle);
@@ -1013,6 +1014,7 @@ int32_t tqProcessStreamTaskCheckReq(STQ* pTq, SRpcMsg* pMsg) {
     tqError("unable to encode rsp %d", __LINE__);
     return -1;
   }
+
   void* buf = rpcMallocCont(sizeof(SMsgHead) + len);
   ((SMsgHead*)buf)->vgId = htonl(req.upstreamNodeId);
 
@@ -1542,3 +1544,5 @@ FAIL:
   taosFreeQitem(pMsg);
   return -1;
 }
+
+int32_t tqCheckLogInWal(STQ* pTq, int64_t version) { return version <= pTq->walLogLastVer; }

source/dnode/vnode/src/tq/tqSink.c

@@ -426,7 +426,7 @@ int32_t tqBlockToSubmit(SVnode* pVnode, const SArray* pBlocks, const STSchema* p
       goto _end;
     }
 
-    for (int32_t j = 0; j < rows; j++) {
+    for (int32_t j = 0; j < rows; ++j) {
       taosArrayClear(pVals);
       for (int32_t k = 0; k < pTSchema->numOfCols; k++) {
         const STColumn*  pCol = &pTSchema->columns[k];
@@ -469,7 +469,7 @@ int32_t tqBlockToSubmit(SVnode* pVnode, const SArray* pBlocks, const STSchema* p
     if (NULL == pBuf) {
       goto _end;
     }
-    ((SMsgHead*)pBuf)->vgId = htonl(TD_VID(pVnode));
+    ((SMsgHead*)pBuf)->vgId = TD_VID(pVnode);
     ((SMsgHead*)pBuf)->contLen = htonl(len);
     tEncoderInit(&encoder, POINTER_SHIFT(pBuf, sizeof(SMsgHead)), len - sizeof(SMsgHead));
     if (tEncodeSSubmitReq2(&encoder, pReq) < 0) {
@@ -756,24 +756,14 @@ void tqSinkToTablePipeline2(SStreamTask* pTask, void* vnode, int64_t ver, void*
 
   tqDebug("vgId:%d, task %d write into table, block num: %d", TD_VID(pVnode), pTask->taskId, blockSz);
 
-  void*        pBuf = NULL;
+  void*   pBuf = NULL;
-  SSubmitReq2* pReq = NULL;
+  SArray* tagArray = NULL;
-  SArray*      tagArray = NULL;
+  SArray* pVals = NULL;
-  SArray*      pVals = NULL;
 
   if (!(tagArray = taosArrayInit(1, sizeof(STagVal)))) {
     goto _end;
   }
 
-  if (!(pReq = taosMemoryCalloc(1, sizeof(SSubmitReq2)))) {
-    terrno = TSDB_CODE_OUT_OF_MEMORY;
-    goto _end;
-  }
-
-  if (!(pReq->aSubmitTbData = taosArrayInit(1, sizeof(SSubmitTbData)))) {
-    goto _end;
-  }
-
   for (int32_t i = 0; i < blockSz; i++) {
     SSDataBlock* pDataBlock = taosArrayGet(pBlocks, i);
     if (pDataBlock->info.type == STREAM_DELETE_RESULT) {
@@ -948,13 +938,17 @@ void tqSinkToTablePipeline2(SStreamTask* pTask, void* vnode, int64_t ver, void*
         taosArrayPush(tbData.aRowP, &pRow);
       }
 
-      taosArrayClear(pReq->aSubmitTbData);
+      SSubmitReq2 submitReq = {0};
-      taosArrayPush(pReq->aSubmitTbData, &tbData);
+      if (!(submitReq.aSubmitTbData = taosArrayInit(1, sizeof(SSubmitTbData)))) {
+        goto _end;
+      }
+
+      taosArrayPush(submitReq.aSubmitTbData, &tbData);
 
       // encode
       int32_t len;
       int32_t code;
-      tEncodeSize(tEncodeSSubmitReq2, pReq, len, code);
+      tEncodeSize(tEncodeSSubmitReq2, &submitReq, len, code);
       SEncoder encoder;
       len += sizeof(SMsgHead);
       pBuf = rpcMallocCont(len);
@@ -964,7 +958,7 @@ void tqSinkToTablePipeline2(SStreamTask* pTask, void* vnode, int64_t ver, void*
       ((SMsgHead*)pBuf)->vgId = TD_VID(pVnode);
       ((SMsgHead*)pBuf)->contLen = htonl(len);
       tEncoderInit(&encoder, POINTER_SHIFT(pBuf, sizeof(SMsgHead)), len - sizeof(SMsgHead));
-      if (tEncodeSSubmitReq2(&encoder, pReq) < 0) {
+      if (tEncodeSSubmitReq2(&encoder, &submitReq) < 0) {
         terrno = TSDB_CODE_OUT_OF_MEMORY;
         tqError("failed to encode submit req since %s", terrstr());
         tEncoderClear(&encoder);
@@ -972,6 +966,7 @@ void tqSinkToTablePipeline2(SStreamTask* pTask, void* vnode, int64_t ver, void*
         continue;
       }
       tEncoderClear(&encoder);
+      tDestroySSubmitReq2(&submitReq, TSDB_MSG_FLG_ENCODE);
 
       SRpcMsg msg = {
           .msgType = TDMT_VND_SUBMIT,
@@ -988,8 +983,6 @@ _end:
   taosArrayDestroy(tagArray);
   taosArrayDestroy(pVals);
   // TODO: change
-  tDestroySSubmitReq2(pReq, TSDB_MSG_FLG_ENCODE);
-  taosMemoryFree(pReq);
 }
 
 #if 0

source/dnode/vnode/src/tsdb/tsdbCache.c

@@ -1473,29 +1473,6 @@ int32_t tsdbCacheGetLastrowH(SLRUCache *pCache, tb_uid_t uid, SCacheRowsReader *
 
   return code;
 }
-// int32_t tsdbCacheLastArray2Row(SArray *pLastArray, STSRow **ppRow, STSchema *pTSchema) {
-//   int32_t code = 0;
-//   int16_t nCol = taosArrayGetSize(pLastArray);
-//   SArray *pColArray = taosArrayInit(nCol, sizeof(SColVal));
-
-//   for (int16_t iCol = 0; iCol < nCol; ++iCol) {
-//     SLastCol *tTsVal = (SLastCol *)taosArrayGet(pLastArray, iCol);
-//     SColVal  *tColVal = &tTsVal->colVal;
-//     taosArrayPush(pColArray, tColVal);
-//   }
-
-//   code = tdSTSRowNew(pColArray, pTSchema, ppRow);
-//   if (code) goto _err;
-
-//   taosArrayDestroy(pColArray);
-
-//   return code;
-
-// _err:
-//   taosArrayDestroy(pColArray);
-
-//   return code;
-// }
 
 int32_t tsdbCacheGetLastH(SLRUCache *pCache, tb_uid_t uid, SCacheRowsReader *pr, LRUHandle **handle) {
   int32_t code = 0;

source/dnode/vnode/src/tsdb/tsdbCacheRead.c

@@ -20,9 +20,8 @@
 
 #define HASTYPE(_type, _t) (((_type) & (_t)) == (_t))
 
-static void saveOneRow(SArray* pRow, SSDataBlock* pBlock, SCacheRowsReader* pReader, const int32_t* slotIds,
+static int32_t saveOneRow(SArray* pRow, SSDataBlock* pBlock, SCacheRowsReader* pReader, const int32_t* slotIds,
-                       void** pRes) {
+                          void** pRes, const char* idStr) {
-  ASSERT(pReader->numOfCols <= taosArrayGetSize(pBlock->pDataBlock));
   int32_t numOfRows = pBlock->info.rows;
 
   if (HASTYPE(pReader->type, CACHESCAN_RETRIEVE_LAST)) {
@@ -65,9 +64,7 @@ static void saveOneRow(SArray* pRow, SSDataBlock* pBlock, SCacheRowsReader* pRea
     }
 
     pBlock->info.rows += allNullRow ? 0 : 1;
-  } else {
+  } else if (HASTYPE(pReader->type, CACHESCAN_RETRIEVE_LAST_ROW)) {
-    ASSERT(HASTYPE(pReader->type, CACHESCAN_RETRIEVE_LAST_ROW));
-
     for (int32_t i = 0; i < pReader->numOfCols; ++i) {
       SColumnInfoData* pColInfoData = taosArrayGet(pBlock->pDataBlock, i);
 
@@ -94,11 +91,16 @@ static void saveOneRow(SArray* pRow, SSDataBlock* pBlock, SCacheRowsReader* pRea
     }
 
     pBlock->info.rows += 1;
+  } else {
+    tsdbError("invalid retrieve type:%d, %s", pReader->type, idStr);
+    return TSDB_CODE_INVALID_PARA;
   }
+
+  return TSDB_CODE_SUCCESS;
 }
 
 int32_t tsdbCacherowsReaderOpen(void* pVnode, int32_t type, void* pTableIdList, int32_t numOfTables, int32_t numOfCols,
-                                uint64_t suid, void** pReader) {
+                                uint64_t suid, void** pReader, const char* idstr) {
   *pReader = NULL;
   SCacheRowsReader* p = taosMemoryCalloc(1, sizeof(SCacheRowsReader));
   if (p == NULL) {
@@ -142,6 +144,8 @@ int32_t tsdbCacherowsReaderOpen(void* pVnode, int32_t type, void* pTableIdList,
     return TSDB_CODE_OUT_OF_MEMORY;
   }
 
+  p->idstr = taosMemoryStrDup(idstr);
+
   *pReader = p;
   return TSDB_CODE_SUCCESS;
 }
@@ -160,6 +164,7 @@ void* tsdbCacherowsReaderClose(void* pReader) {
 
   destroyLastBlockLoadInfo(p->pLoadInfo);
 
+  taosMemoryFree((void*) p->idstr);
   taosMemoryFree(pReader);
   return NULL;
 }
@@ -312,7 +317,7 @@ int32_t tsdbRetrieveCacheRows(void* pReader, SSDataBlock* pResBlock, const int32
     }
 
     if (hasRes) {
-      saveOneRow(pLastCols, pResBlock, pr, slotIds, pRes);
+      saveOneRow(pLastCols, pResBlock, pr, slotIds, pRes, pr->idstr);
     }
 
   } else if (HASTYPE(pr->type, CACHESCAN_RETRIEVE_TYPE_ALL)) {
@@ -331,7 +336,7 @@ int32_t tsdbRetrieveCacheRows(void* pReader, SSDataBlock* pResBlock, const int32
         continue;
       }
 
-      saveOneRow(pRow, pResBlock, pr, slotIds, pRes);
+      saveOneRow(pRow, pResBlock, pr, slotIds, pRes, pr->idstr);
       // TODO reset the pRes
 
       taosArrayPush(pTableUidList, &pKeyInfo->uid);

source/dnode/vnode/src/tsdb/tsdbMergeTree.c

@@ -32,13 +32,13 @@ struct SLDataIter {
 };
 
 SSttBlockLoadInfo *tCreateLastBlockLoadInfo(STSchema *pSchema, int16_t *colList, int32_t numOfCols) {
-  SSttBlockLoadInfo *pLoadInfo = taosMemoryCalloc(TSDB_DEFAULT_STT_FILE, sizeof(SSttBlockLoadInfo));
+  SSttBlockLoadInfo *pLoadInfo = taosMemoryCalloc(TSDB_MAX_STT_TRIGGER, sizeof(SSttBlockLoadInfo));
   if (pLoadInfo == NULL) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
     return NULL;
   }
 
-  for (int32_t i = 0; i < TSDB_DEFAULT_STT_FILE; ++i) {
+  for (int32_t i = 0; i < TSDB_MAX_STT_TRIGGER; ++i) {
     pLoadInfo[i].blockIndex[0] = -1;
     pLoadInfo[i].blockIndex[1] = -1;
     pLoadInfo[i].currentLoadBlockIndex = 1;
@@ -63,7 +63,7 @@ SSttBlockLoadInfo *tCreateLastBlockLoadInfo(STSchema *pSchema, int16_t *colList,
 }
 
 void resetLastBlockLoadInfo(SSttBlockLoadInfo *pLoadInfo) {
-  for (int32_t i = 0; i < TSDB_DEFAULT_STT_FILE; ++i) {
+  for (int32_t i = 0; i < TSDB_MAX_STT_TRIGGER; ++i) {
     pLoadInfo[i].currentLoadBlockIndex = 1;
     pLoadInfo[i].blockIndex[0] = -1;
     pLoadInfo[i].blockIndex[1] = -1;
@@ -77,14 +77,14 @@ void resetLastBlockLoadInfo(SSttBlockLoadInfo *pLoadInfo) {
 }
 
 void getLastBlockLoadInfo(SSttBlockLoadInfo *pLoadInfo, int64_t *blocks, double *el) {
-  for (int32_t i = 0; i < TSDB_DEFAULT_STT_FILE; ++i) {
+  for (int32_t i = 0; i < TSDB_MAX_STT_TRIGGER; ++i) {
     *el += pLoadInfo[i].elapsedTime;
     *blocks += pLoadInfo[i].loadBlocks;
   }
 }
 
 void *destroyLastBlockLoadInfo(SSttBlockLoadInfo *pLoadInfo) {
-  for (int32_t i = 0; i < TSDB_DEFAULT_STT_FILE; ++i) {
+  for (int32_t i = 0; i < TSDB_MAX_STT_TRIGGER; ++i) {
     pLoadInfo[i].currentLoadBlockIndex = 1;
     pLoadInfo[i].blockIndex[0] = -1;
     pLoadInfo[i].blockIndex[1] = -1;
@@ -158,9 +158,9 @@ static SBlockData *loadLastBlock(SLDataIter *pIter, const char *idStr) {
             pBlock, el, idStr);
 
   pInfo->blockIndex[pInfo->currentLoadBlockIndex] = pIter->iSttBlk;
-  tsdbDebug("last block index list:%d, %d, %s", pInfo->blockIndex[0], pInfo->blockIndex[1], idStr);
-
   pIter->iRow = (pIter->backward) ? pInfo->blockData[pInfo->currentLoadBlockIndex].nRow : -1;
+
+  tsdbDebug("last block index list:%d, %d, rowIndex:%d %s", pInfo->blockIndex[0], pInfo->blockIndex[1], pIter->iRow, idStr);
   return &pInfo->blockData[pInfo->currentLoadBlockIndex];
 
 _exit:
@@ -419,6 +419,7 @@ static void findNextValidRow(SLDataIter *pIter, const char *idStr) {
       pBlockData->aUid != NULL) {
     i = binarySearchForStartRowIndex((uint64_t *)pBlockData->aUid, pBlockData->nRow, pIter->uid, pIter->backward);
     if (i == -1) {
+      tsdbDebug("failed to find the data in pBlockData, uid:%"PRIu64" , %s", pIter->uid, idStr);
       pIter->iRow = -1;
       return;
     }
@@ -500,7 +501,12 @@ bool tLDataIterNextRow(SLDataIter *pIter, const char *idStr) {
 
     if (iBlockL != pIter->iSttBlk) {
       pBlockData = loadLastBlock(pIter, idStr);
-      pIter->iRow += step;
+      if (pBlockData == NULL) {
+        goto _exit;
+      }
+
+      // set start row index
+      pIter->iRow = pIter->backward? pBlockData->nRow-1:0;
     }
   }
 
@@ -561,7 +567,6 @@ int32_t tMergeTreeOpen(SMergeTree *pMTree, int8_t backward, SDataFReader *pFRead
 
   pMTree->pLoadInfo = pBlockLoadInfo;
   pMTree->destroyLoadInfo = destroyLoadInfo;
-  ASSERT(pMTree->pLoadInfo != NULL);
 
   for (int32_t i = 0; i < pFReader->pSet->nSttF; ++i) {  // open all last file
     struct SLDataIter *pIter = NULL;

source/dnode/vnode/src/tsdb/tsdbRead.c

@@ -244,7 +244,7 @@ static int32_t setColumnIdSlotList(SBlockLoadSuppInfo* pSupInfo, SColumnInfo* pC
   return TSDB_CODE_SUCCESS;
 }
 
-static void updateBlockSMAInfo(STSchema* pSchema, SBlockLoadSuppInfo* pSupInfo) {
+static int32_t updateBlockSMAInfo(STSchema* pSchema, SBlockLoadSuppInfo* pSupInfo) {
   int32_t i = 0, j = 0;
 
   while (i < pSchema->numOfCols && j < pSupInfo->numOfCols) {
@@ -252,7 +252,7 @@ static void updateBlockSMAInfo(STSchema* pSchema, SBlockLoadSuppInfo* pSupInfo)
     if (pTCol->colId == pSupInfo->colId[j]) {
       if (!IS_BSMA_ON(pTCol)) {
         pSupInfo->smaValid = false;
-        return;
+        return TSDB_CODE_SUCCESS;
       }
 
       i += 1;
@@ -261,9 +261,11 @@ static void updateBlockSMAInfo(STSchema* pSchema, SBlockLoadSuppInfo* pSupInfo)
       // do nothing
       i += 1;
     } else {
-      ASSERT(0);
+      return TSDB_CODE_INVALID_PARA;
     }
   }
+
+  return TSDB_CODE_SUCCESS;
 }
 
 static int32_t initBlockScanInfoBuf(SBlockInfoBuf* pBuf, int32_t numOfTables) {
@@ -398,7 +400,6 @@ static void destroyAllBlockScanInfo(SHashObj* pTableMap) {
 }
 
 static bool isEmptyQueryTimeWindow(STimeWindow* pWindow) {
-  ASSERT(pWindow != NULL);
   return pWindow->skey > pWindow->ekey;
 }
 
@@ -582,7 +583,7 @@ static int32_t tsdbReaderCreate(SVnode* pVnode, SQueryTableDataCond* pCond, STsd
   }
 
   if (VND_IS_TSMA(pVnode)) {
-    tsdbDebug("vgId:%d, tsma is selected to query", TD_VID(pVnode));
+    tsdbDebug("vgId:%d, tsma is selected to query, %s", TD_VID(pVnode), idstr);
   }
 
   initReaderStatus(&pReader->status);
@@ -597,7 +598,6 @@ static int32_t tsdbReaderCreate(SVnode* pVnode, SQueryTableDataCond* pCond, STsd
   pReader->type = pCond->type;
   pReader->window = updateQueryTimeWindow(pReader->pTsdb, &pCond->twindows);
   pReader->blockInfoBuf.numPerBucket = 1000;  // 1000 tables per bucket
-  ASSERT(pCond->numOfCols > 0);
 
   if (pReader->pResBlock == NULL) {
     pReader->freeBlock = true;
@@ -608,6 +608,12 @@ static int32_t tsdbReaderCreate(SVnode* pVnode, SQueryTableDataCond* pCond, STsd
     }
   }
 
+  if (pCond->numOfCols <= 0) {
+    tsdbError("vgId:%d, invalid column number %d in query cond, %s", TD_VID(pVnode), pCond->numOfCols, idstr);
+    code = TSDB_CODE_INVALID_PARA;
+    goto _end;
+  }
+
   // todo refactor.
   limitOutputBufferSize(pCond, &pReader->capacity);
 
@@ -796,8 +802,9 @@ static void doCopyColVal(SColumnInfoData* pColInfoData, int32_t rowIndex, int32_
 }
 
 static SFileDataBlockInfo* getCurrentBlockInfo(SDataBlockIter* pBlockIter) {
-  if (taosArrayGetSize(pBlockIter->blockList) == 0) {
+  size_t num = taosArrayGetSize(pBlockIter->blockList);
-    ASSERT(pBlockIter->numOfBlocks == taosArrayGetSize(pBlockIter->blockList));
+  if (num == 0) {
+    ASSERT(pBlockIter->numOfBlocks == num);
     return NULL;
   }
 
@@ -807,73 +814,6 @@ static SFileDataBlockInfo* getCurrentBlockInfo(SDataBlockIter* pBlockIter) {
 
 static SDataBlk* getCurrentBlock(SDataBlockIter* pBlockIter) { return &pBlockIter->block; }
 
-int32_t binarySearchForTs(char* pValue, int num, TSKEY key, int order) {
-  int32_t midPos = -1;
-  int32_t numOfRows;
-
-  ASSERT(order == TSDB_ORDER_ASC || order == TSDB_ORDER_DESC);
-
-  TSKEY*  keyList = (TSKEY*)pValue;
-  int32_t firstPos = 0;
-  int32_t lastPos = num - 1;
-
-  if (order == TSDB_ORDER_DESC) {
-    // find the first position which is smaller than the key
-    while (1) {
-      if (key >= keyList[firstPos]) return firstPos;
-      if (key == keyList[lastPos]) return lastPos;
-
-      if (key < keyList[lastPos]) {
-        lastPos += 1;
-        if (lastPos >= num) {
-          return -1;
-        } else {
-          return lastPos;
-        }
-      }
-
-      numOfRows = lastPos - firstPos + 1;
-      midPos = (numOfRows >> 1) + firstPos;
-
-      if (key < keyList[midPos]) {
-        firstPos = midPos + 1;
-      } else if (key > keyList[midPos]) {
-        lastPos = midPos - 1;
-      } else {
-        break;
-      }
-    }
-
-  } else {
-    // find the first position which is bigger than the key
-    while (1) {
-      if (key <= keyList[firstPos]) return firstPos;
-      if (key == keyList[lastPos]) return lastPos;
-
-      if (key > keyList[lastPos]) {
-        lastPos = lastPos + 1;
-        if (lastPos >= num)
-          return -1;
-        else
-          return lastPos;
-      }
-
-      numOfRows = lastPos - firstPos + 1;
-      midPos = (numOfRows >> 1u) + firstPos;
-
-      if (key < keyList[midPos]) {
-        lastPos = midPos - 1;
-      } else if (key > keyList[midPos]) {
-        firstPos = midPos + 1;
-      } else {
-        break;
-      }
-    }
-  }
-
-  return midPos;
-}
-
 static int doBinarySearchKey(TSKEY* keyList, int num, int pos, TSKEY key, int order) {
   // start end position
   int s, e;
@@ -974,8 +914,8 @@ static void copyNumericCols(const SColData* pData, SFileBlockDumpInfo* pDumpInfo
 
   int32_t step = asc ? 1 : -1;
 
-  // make sure it is aligned to 8bit
+  // make sure it is aligned to 8bit, the allocated memory address is aligned to 256bit
-  ASSERT((((uint64_t)pColData->pData) & (0x8 - 1)) == 0);
+//  ASSERT((((uint64_t)pColData->pData) & (0x8 - 1)) == 0);
 
   // 1. copy data in a batch model
   memcpy(pColData->pData, p, dumpedRows * tDataTypes[pData->type].bytes);
@@ -1073,11 +1013,20 @@ static int32_t copyBlockDataToSDataBlock(STsdbReader* pReader, STableBlockScanIn
       // pDumpInfo->rowIndex = 0;
     } else if (!asc && pReader->window.ekey >= pBlock->maxKey.ts) {
       // pDumpInfo->rowIndex = pBlock->nRow - 1;
-    } else {
+    } else {  // find the appropriate the start position in current block, and set it to be the current rowIndex
       int32_t pos = asc ? pBlock->nRow - 1 : 0;
       int32_t order = asc ? TSDB_ORDER_DESC : TSDB_ORDER_ASC;
       int64_t key = asc ? pReader->window.skey : pReader->window.ekey;
       pDumpInfo->rowIndex = doBinarySearchKey(pBlockData->aTSKEY, pBlock->nRow, pos, key, order);
+
+      if (pDumpInfo->rowIndex < 0) {
+        tsdbError(
+            "%p failed to locate the start position in current block, global index:%d, table index:%d, brange:%" PRId64
+            "-%" PRId64 ", minVer:%" PRId64 ", maxVer:%" PRId64 " %s",
+            pReader, pBlockIter->index, pBlockInfo->tbBlockIdx, pBlock->minKey.ts, pBlock->maxKey.ts, pBlock->minVer,
+            pBlock->maxVer, pReader->idStr);
+        return TSDB_CODE_INVALID_PARA;
+      }
     }
   }
 
@@ -1186,7 +1135,6 @@ static int32_t doLoadFileBlockData(STsdbReader* pReader, SDataBlockIter* pBlockI
 
   SFileDataBlockInfo* pBlockInfo = getCurrentBlockInfo(pBlockIter);
   SFileBlockDumpInfo* pDumpInfo = &pReader->status.fBlockDumpInfo;
-  ASSERT(pBlockInfo != NULL);
 
   SDataBlk* pBlock = getCurrentBlock(pBlockIter);
   code = tsdbReadDataBlock(pReader->pFileReader, pBlock, pBlockData);
@@ -1224,8 +1172,6 @@ static void cleanupBlockOrderSupporter(SBlockOrderSupporter* pSup) {
 }
 
 static int32_t initBlockOrderSupporter(SBlockOrderSupporter* pSup, int32_t numOfTables) {
-  ASSERT(numOfTables >= 1);
-
   pSup->numOfBlocksPerTable = taosMemoryCalloc(1, sizeof(int32_t) * numOfTables);
   pSup->indexPerTable = taosMemoryCalloc(1, sizeof(int32_t) * numOfTables);
   pSup->pDataBlockInfo = taosMemoryCalloc(1, POINTER_BYTES * numOfTables);
@@ -1332,7 +1278,10 @@ static int32_t initBlockIterator(STsdbReader* pReader, SDataBlockIter* pBlockIte
     sup.numOfTables += 1;
   }
 
-  ASSERT(numOfBlocks == cnt);
+  if (numOfBlocks != cnt && sup.numOfTables != numOfTables) {
+    cleanupBlockOrderSupporter(&sup);
+    return TSDB_CODE_INVALID_PARA;
+  }
 
   // since there is only one table qualified, blocks are not sorted
   if (sup.numOfTables == 1) {
@@ -1354,10 +1303,9 @@ static int32_t initBlockIterator(STsdbReader* pReader, SDataBlockIter* pBlockIte
   tsdbDebug("%p create data blocks info struct completed, %d blocks in %d tables %s", pReader, cnt, sup.numOfTables,
             pReader->idStr);
 
-  ASSERT(cnt <= numOfBlocks && sup.numOfTables <= numOfTables);
-
   SMultiwayMergeTreeInfo* pTree = NULL;
-  uint8_t                 ret = tMergeTreeCreate(&pTree, sup.numOfTables, &sup, fileDataBlockOrderCompar);
+
+  uint8_t ret = tMergeTreeCreate(&pTree, sup.numOfTables, &sup, fileDataBlockOrderCompar);
   if (ret != TSDB_CODE_SUCCESS) {
     cleanupBlockOrderSupporter(&sup);
     return TSDB_CODE_OUT_OF_MEMORY;
@@ -1435,8 +1383,6 @@ static bool getNeighborBlockOfSameTable(SFileDataBlockInfo* pBlockInfo, STableBl
 }
 
 static int32_t findFileBlockInfoIndex(SDataBlockIter* pBlockIter, SFileDataBlockInfo* pFBlockInfo) {
-  ASSERT(pBlockIter != NULL && pFBlockInfo != NULL);
-
   int32_t step = ASCENDING_TRAVERSE(pBlockIter->order) ? 1 : -1;
   int32_t index = pBlockIter->index;
 
@@ -1449,7 +1395,6 @@ static int32_t findFileBlockInfoIndex(SDataBlockIter* pBlockIter, SFileDataBlock
     index += step;
   }
 
-  ASSERT(0);
   return -1;
 }
 
@@ -1928,7 +1873,6 @@ static int32_t doMergeFileBlockAndLastBlock(SLastBlockReader* pLastBlockReader,
     }
 
     doMergeRowsInLastBlock(pLastBlockReader, pBlockScanInfo, tsLastBlock, &merge, &pReader->verRange);
-    ASSERT(mergeBlockData);
 
     // merge with block data if ts == key
     if (tsLastBlock == pBlockData->aTSKEY[pDumpInfo->rowIndex]) {
@@ -1994,7 +1938,6 @@ static int32_t mergeFileBlockAndLastBlock(STsdbReader* pReader, SLastBlockReader
         tsdbRowMergerClear(&merge);
         return code;
       } else {
-        ASSERT(0);
         return TSDB_CODE_SUCCESS;
       }
     } else {  // desc order
@@ -2015,7 +1958,6 @@ static int32_t doMergeMultiLevelRows(STsdbReader* pReader, STableBlockScanInfo*
 
   TSDBROW* pRow = getValidMemRow(&pBlockScanInfo->iter, pDelList, pReader);
   TSDBROW* piRow = getValidMemRow(&pBlockScanInfo->iiter, pDelList, pReader);
-  ASSERT(pRow != NULL && piRow != NULL);
 
   int64_t tsLast = INT64_MIN;
   if (hasDataInLastBlock(pLastBlockReader)) {
@@ -2239,7 +2181,6 @@ static int32_t initMemDataIterator(STableBlockScanInfo* pBlockScanInfo, STsdbRea
   if (pReader->pReadSnap->pMem != NULL) {
     d = tsdbGetTbDataFromMemTable(pReader->pReadSnap->pMem, pReader->suid, pBlockScanInfo->uid);
     if (d != NULL) {
-      ASSERT(pBlockScanInfo->iter.iter == NULL);
       code = tsdbTbDataIterCreate(d, &startKey, backward, &pBlockScanInfo->iter.iter);
       if (code == TSDB_CODE_SUCCESS) {
         pBlockScanInfo->iter.hasVal = (tsdbTbDataIterGet(pBlockScanInfo->iter.iter) != NULL);
@@ -2353,10 +2294,9 @@ static int64_t getCurrentKeyInLastBlock(SLastBlockReader* pLastBlockReader) {
 static bool hasDataInLastBlock(SLastBlockReader* pLastBlockReader) { return pLastBlockReader->mergeTree.pIter != NULL; }
 
 bool hasDataInFileBlock(const SBlockData* pBlockData, const SFileBlockDumpInfo* pDumpInfo) {
-  if (pBlockData->nRow > 0) {
+  if ((pBlockData->nRow > 0) && (pBlockData->nRow != pDumpInfo->totalRows)) {
-    ASSERT(pBlockData->nRow == pDumpInfo->totalRows);
+    return false; // this is an invalid result.
   }
-
   return pBlockData->nRow > 0 && (!pDumpInfo->allDumped);
 }
 
@@ -2423,6 +2363,46 @@ static int32_t buildComposedDataBlockImpl(STsdbReader* pReader, STableBlockScanI
   }
 }
 
+static int32_t loadNeighborIfOverlap(SFileDataBlockInfo* pBlockInfo, STableBlockScanInfo* pBlockScanInfo,
+                                     STsdbReader* pReader, bool* loadNeighbor) {
+  int32_t     code = TSDB_CODE_SUCCESS;
+  int32_t     step = ASCENDING_TRAVERSE(pReader->order) ? 1 : -1;
+  int32_t     nextIndex = -1;
+  SBlockIndex nxtBIndex = {0};
+
+  *loadNeighbor = false;
+  SDataBlk* pBlock = getCurrentBlock(&pReader->status.blockIter);
+
+  bool hasNeighbor = getNeighborBlockOfSameTable(pBlockInfo, pBlockScanInfo, &nextIndex, pReader->order, &nxtBIndex);
+  if (!hasNeighbor) {  // do nothing
+    return code;
+  }
+
+  if (overlapWithNeighborBlock(pBlock, &nxtBIndex, pReader->order)) {  // load next block
+    SReaderStatus*  pStatus = &pReader->status;
+    SDataBlockIter* pBlockIter = &pStatus->blockIter;
+
+    // 1. find the next neighbor block in the scan block list
+    SFileDataBlockInfo fb = {.uid = pBlockInfo->uid, .tbBlockIdx = nextIndex};
+    int32_t            neighborIndex = findFileBlockInfoIndex(pBlockIter, &fb);
+
+    // 2. remove it from the scan block list
+    setFileBlockActiveInBlockIter(pBlockIter, neighborIndex, step);
+
+    // 3. load the neighbor block, and set it to be the currently accessed file data block
+    code = doLoadFileBlockData(pReader, pBlockIter, &pStatus->fileBlockData, pBlockInfo->uid);
+    if (code != TSDB_CODE_SUCCESS) {
+      return code;
+    }
+
+    // 4. check the data values
+    initBlockDumpInfo(pReader, pBlockIter);
+    *loadNeighbor = true;
+  }
+
+  return code;
+}
+
 static int32_t buildComposedDataBlock(STsdbReader* pReader) {
   int32_t code = TSDB_CODE_SUCCESS;
 
@@ -2481,39 +2461,13 @@ static int32_t buildComposedDataBlock(STsdbReader* pReader) {
 
         SDataBlk* pBlock = getCurrentBlock(&pReader->status.blockIter);
         if (pDumpInfo->rowIndex >= pBlock->nRow || pDumpInfo->rowIndex < 0) {
-          SBlockIndex bIndex = {0};
+          pBlockInfo = getCurrentBlockInfo(&pReader->status.blockIter);  // NOTE: get the new block info
-          int32_t     nextIndex = -1;
-          bool        hasNeighbor = false;
-          if (pBlockInfo != NULL) {
-            hasNeighbor = getNeighborBlockOfSameTable(pBlockInfo, pBlockScanInfo, &nextIndex, pReader->order, &bIndex);
-          }
 
-          if (!hasNeighbor) {  // do nothing
+          // continue check for the next file block if the last ts in the current block
-            setBlockAllDumped(pDumpInfo, pBlock->maxKey.ts, pReader->order);
+          // is overlapped with the next neighbor block
-            break;
+          bool loadNeighbor = false;
-          }
+          code = loadNeighborIfOverlap(pBlockInfo, pBlockScanInfo, pReader, &loadNeighbor);
-
+          if ((!loadNeighbor) || (code != 0)) {
-          if (overlapWithNeighborBlock(pBlock, &bIndex, pReader->order)) {  // load next block
-            SReaderStatus*  pStatus = &pReader->status;
-            SDataBlockIter* pBlockIter = &pStatus->blockIter;
-
-            // 1. find the next neighbor block in the scan block list
-            SFileDataBlockInfo fb = {.uid = pBlockInfo->uid, .tbBlockIdx = nextIndex};
-            int32_t            neighborIndex = findFileBlockInfoIndex(pBlockIter, &fb);
-
-            // 2. remove it from the scan block list
-            setFileBlockActiveInBlockIter(pBlockIter, neighborIndex, step);
-
-            // 3. load the neighbor block, and set it to be the currently accessed file data block
-            code = doLoadFileBlockData(pReader, pBlockIter, &pStatus->fileBlockData, pBlockInfo->uid);
-            if (code != TSDB_CODE_SUCCESS) {
-              setBlockAllDumped(pDumpInfo, pBlock->maxKey.ts, pReader->order);
-              break;
-            }
-
-            // 4. check the data values
-            initBlockDumpInfo(pReader, pBlockIter);
-          } else {
             setBlockAllDumped(pDumpInfo, pBlock->maxKey.ts, pReader->order);
             break;
           }
@@ -2573,7 +2527,6 @@ int32_t initDelSkylineIterator(STableBlockScanInfo* pBlockScanInfo, STsdbReader*
 
   int32_t code = 0;
   SArray* pDelData = taosArrayInit(4, sizeof(SDelData));
-  ASSERT(pReader->pReadSnap != NULL);
 
   SDelFile* pDelFile = pReader->pReadSnap->fs.pDelFile;
   if (pDelFile && taosArrayGetSize(pReader->pDelIdx) > 0) {
@@ -2779,7 +2732,10 @@ static bool moveToNextTable(SUidOrderCheckInfo* pOrderedCheckInfo, SReaderStatus
 
   uint64_t uid = pOrderedCheckInfo->tableUidList[pOrderedCheckInfo->currentIndex];
   pStatus->pTableIter = taosHashGet(pStatus->pTableMap, &uid, sizeof(uid));
-  ASSERT(pStatus->pTableIter != NULL);
+  if (pStatus->pTableIter == NULL) {
+    return false;
+  }
+
   return true;
 }
 
@@ -2854,7 +2810,6 @@ static int32_t doBuildDataBlock(STsdbReader* pReader) {
   TSDBKEY keyInBuf = getCurrentKeyInBuf(pScanInfo, pReader);
 
   if (pBlockInfo == NULL) {  // build data block from last data file
-    ASSERT(pBlockIter->numOfBlocks == 0);
     code = buildComposedDataBlock(pReader);
   } else if (fileBlockShouldLoad(pReader, pBlockInfo, pBlock, pScanInfo, keyInBuf, pLastBlockReader)) {
     code = doLoadFileBlockData(pReader, pBlockIter, &pStatus->fileBlockData, pScanInfo->uid);
@@ -3124,10 +3079,10 @@ SVersionRange getQueryVerRange(SVnode* pVnode, SQueryTableDataCond* pCond, int8_
 }
 
 bool hasBeenDropped(const SArray* pDelList, int32_t* index, TSDBKEY* pKey, int32_t order, SVersionRange* pVerRange) {
-  ASSERT(pKey != NULL);
   if (pDelList == NULL) {
     return false;
   }
+
   size_t  num = taosArrayGetSize(pDelList);
   bool    asc = ASCENDING_TRAVERSE(order);
   int32_t step = asc ? 1 : -1;
@@ -3330,35 +3285,10 @@ static int32_t checkForNeighborFileBlock(STsdbReader* pReader, STableBlockScanIn
   *state = CHECK_FILEBLOCK_QUIT;
   int32_t step = ASCENDING_TRAVERSE(pReader->order) ? 1 : -1;
 
-  int32_t     nextIndex = -1;
+  bool loadNeighbor = true;
-  SBlockIndex bIndex = {0};
+  int32_t code = loadNeighborIfOverlap(pFBlock, pScanInfo, pReader, &loadNeighbor);
-
-  bool hasNeighbor = getNeighborBlockOfSameTable(pFBlock, pScanInfo, &nextIndex, pReader->order, &bIndex);
-  if (!hasNeighbor) {  // do nothing
-    return 0;
-  }
-
-  bool overlap = overlapWithNeighborBlock(pBlock, &bIndex, pReader->order);
-  if (overlap) {  // load next block
-    SReaderStatus*  pStatus = &pReader->status;
-    SDataBlockIter* pBlockIter = &pStatus->blockIter;
-
-    // 1. find the next neighbor block in the scan block list
-    SFileDataBlockInfo fb = {.uid = pFBlock->uid, .tbBlockIdx = nextIndex};
-    int32_t            neighborIndex = findFileBlockInfoIndex(pBlockIter, &fb);
-
-    // 2. remove it from the scan block list
-    setFileBlockActiveInBlockIter(pBlockIter, neighborIndex, step);
-
-    // 3. load the neighbor block, and set it to be the currently accessed file data block
-    int32_t code = doLoadFileBlockData(pReader, pBlockIter, &pStatus->fileBlockData, pFBlock->uid);
-    if (code != TSDB_CODE_SUCCESS) {
-      return code;
-    }
-
-    // 4. check the data values
-    initBlockDumpInfo(pReader, pBlockIter);
 
+  if (loadNeighbor && (code == TSDB_CODE_SUCCESS)) {
     pDumpInfo->rowIndex =
         doMergeRowsInFileBlockImpl(pBlockData, pDumpInfo->rowIndex, key, pMerger, &pReader->verRange, step);
     if (pDumpInfo->rowIndex >= pDumpInfo->totalRows) {
@@ -3366,7 +3296,7 @@ static int32_t checkForNeighborFileBlock(STsdbReader* pReader, STableBlockScanIn
     }
   }
 
-  return TSDB_CODE_SUCCESS;
+  return code;
 }
 
 int32_t doMergeRowsInFileBlocks(SBlockData* pBlockData, STableBlockScanInfo* pScanInfo, STsdbReader* pReader,
@@ -3552,6 +3482,7 @@ int32_t doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* p
   }
 
   int32_t code = tsdbRowMergerGetRow(&merge, pTSRow);
+  tsdbRowMergerClear(&merge);
   return code;
 }
 
@@ -3744,13 +3675,11 @@ int32_t buildDataBlockFromBufImpl(STableBlockScanInfo* pBlockScanInfo, int64_t e
     }
   } while (1);
 
-  ASSERT(pBlock->info.rows <= capacity);
   return TSDB_CODE_SUCCESS;
 }
 
 // TODO refactor: with createDataBlockScanInfo
 int32_t tsdbSetTableList(STsdbReader* pReader, const void* pTableList, int32_t num) {
-  ASSERT(pReader != NULL);
   int32_t size = taosHashGetSize(pReader->status.pTableMap);
 
   STableBlockScanInfo** p = NULL;
@@ -3883,15 +3812,16 @@ int32_t tsdbReaderOpen(SVnode* pVnode, SQueryTableDataCond* pCond, void* pTableL
   }
 
   if (pReader->pSchema != NULL) {
-    updateBlockSMAInfo(pReader->pSchema, &pReader->suppInfo);
+    code = updateBlockSMAInfo(pReader->pSchema, &pReader->suppInfo);
+    if (code != TSDB_CODE_SUCCESS) {
+      goto _err;
+    }
   }
 
   STsdbReader* p = (pReader->innerReader[0] != NULL) ? pReader->innerReader[0] : pReader;
   pReader->status.pTableMap = createDataBlockScanInfo(p, &pReader->blockInfoBuf, pTableList, numOfTables);
   if (pReader->status.pTableMap == NULL) {
-    tsdbReaderClose(p);
     *ppReader = NULL;
-
     code = TSDB_CODE_OUT_OF_MEMORY;
     goto _err;
   }
@@ -4114,7 +4044,6 @@ bool tsdbNextDataBlock(STsdbReader* pReader) {
 }
 
 static void setBlockInfo(const STsdbReader* pReader, int32_t* rows, uint64_t* uid, STimeWindow* pWindow) {
-  ASSERT(pReader != NULL);
   *rows = pReader->pResBlock->info.rows;
   *uid = pReader->pResBlock->info.id.uid;
   *pWindow = pReader->pResBlock->info.window;
@@ -4158,25 +4087,27 @@ static void doFillNullColSMA(SBlockLoadSuppInfo* pSup, int32_t numOfRows, int32_
 }
 
 int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock, bool* allHave) {
+  SColumnDataAgg*** pBlockSMA = &pDataBlock->pBlockAgg;
+
   int32_t code = 0;
-  SColumnDataAgg ***pBlockSMA = &pDataBlock->pBlockAgg;
   *allHave = false;
+  *pBlockSMA = NULL;
 
   if (pReader->type == TIMEWINDOW_RANGE_EXTERNAL) {
-    *pBlockSMA = NULL;
     return TSDB_CODE_SUCCESS;
   }
 
   // there is no statistics data for composed block
   if (pReader->status.composedDataBlock || (!pReader->suppInfo.smaValid)) {
-    *pBlockSMA = NULL;
     return TSDB_CODE_SUCCESS;
   }
 
   SFileDataBlockInfo* pFBlock = getCurrentBlockInfo(&pReader->status.blockIter);
   SBlockLoadSuppInfo* pSup = &pReader->suppInfo;
 
-  ASSERT(pReader->pResBlock->info.id.uid == pFBlock->uid);
+  if (pReader->pResBlock->info.id.uid != pFBlock->uid) {
+    return TSDB_CODE_SUCCESS;
+  }
 
   SDataBlk* pBlock = getCurrentBlock(&pReader->status.blockIter);
   if (tDataBlkHasSma(pBlock)) {
@@ -4204,13 +4135,10 @@ int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock,
   // update the number of NULL data rows
   size_t numOfCols = pSup->numOfCols;
 
-  int32_t i = 0, j = 0;
-  size_t  size = taosArrayGetSize(pSup->pColAgg);
-
   // ensure capacity
-  if(pDataBlock->pDataBlock) {
+  if (pDataBlock->pDataBlock) {
-     size_t colsNum = taosArrayGetSize(pDataBlock->pDataBlock);
+    size_t colsNum = taosArrayGetSize(pDataBlock->pDataBlock);
-     taosArrayEnsureCap(pSup->pColAgg, colsNum);
+    taosArrayEnsureCap(pSup->pColAgg, colsNum);
   }
 
   SSDataBlock* pResBlock = pReader->pResBlock;
@@ -4221,8 +4149,9 @@ int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock,
 
   // do fill all null column value SMA info
   doFillNullColSMA(pSup, pBlock->nRow, numOfCols, pTsAgg);
+  size_t size = taosArrayGetSize(pSup->pColAgg);
 
-  i = 0, j = 0;
+  int32_t i = 0, j = 0;
   while (j < numOfCols && i < size) {
     SColumnDataAgg* pAgg = taosArrayGet(pSup->pColAgg, i);
     if (pAgg->colId == pSup->colId[j]) {
@@ -4232,15 +4161,8 @@ int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock,
     } else if (pAgg->colId < pSup->colId[j]) {
       i += 1;
     } else if (pSup->colId[j] < pAgg->colId) {
-      if (pSup->colId[j] == PRIMARYKEY_TIMESTAMP_COL_ID) {
+      // ASSERT(pSup->colId[j] == PRIMARYKEY_TIMESTAMP_COL_ID);
-        pResBlock->pBlockAgg[pSup->slotId[j]] = &pSup->tsColAgg;
+      pResBlock->pBlockAgg[pSup->slotId[j]] = &pSup->tsColAgg;
-      } else {
-        // all date in this block are null
-        SColumnDataAgg nullColAgg = {.colId = pSup->colId[j], .numOfNull = pBlock->nRow};
-        taosArrayPush(pSup->pColAgg, &nullColAgg);
-
-        pResBlock->pBlockAgg[pSup->slotId[j]] = taosArrayGetLast(pSup->pColAgg);
-      }
       j += 1;
     }
   }
@@ -4349,6 +4271,7 @@ int32_t tsdbGetFileBlocksDistInfo(STsdbReader* pReader, STableBlockDistInfo* pTa
   int32_t code = TSDB_CODE_SUCCESS;
   pTableBlockInfo->totalSize = 0;
   pTableBlockInfo->totalRows = 0;
+  pTableBlockInfo->numOfVgroups = 1;
 
   // find the start data block in file
   SReaderStatus* pStatus = &pReader->status;
@@ -4393,6 +4316,8 @@ int32_t tsdbGetFileBlocksDistInfo(STsdbReader* pReader, STableBlockDistInfo* pTa
         pTableBlockInfo->numOfSmallBlocks += 1;
       }
 
+      pTableBlockInfo->totalSize += pBlock->aSubBlock[0].szBlock;
+
       int32_t bucketIndex = getBucketIndex(pTableBlockInfo->defMinRows, bucketRange, numOfRows);
       pTableBlockInfo->blockRowsHisto[bucketIndex]++;
 
@@ -4470,9 +4395,12 @@ int32_t tsdbGetTableSchema(SVnode* pVnode, int64_t uid, STSchema** pSchema, int6
       return terrno;
     }
     sversion = mr.me.stbEntry.schemaRow.version;
-  } else {
+  } else if (mr.me.type == TSDB_NORMAL_TABLE) {
-    ASSERT(mr.me.type == TSDB_NORMAL_TABLE);
     sversion = mr.me.ntbEntry.schemaRow.version;
+  } else {
+    terrno = TSDB_CODE_INVALID_PARA;
+    metaReaderClear(&mr);
+    return terrno;
   }
 
   metaReaderClear(&mr);

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

@@ -123,7 +123,48 @@ void vnodeBufPoolReset(SVBufPool *pPool) {
   pPool->ptr = pPool->node.data;
 }
 
-void *vnodeBufPoolMalloc(SVBufPool *pPool, int32_t size) {
+
+void *vnodeBufPoolMallocAligned(SVBufPool *pPool, int size) {
+  SVBufPoolNode *pNode;
+  void          *p = NULL;
+  uint8_t       *ptr = NULL;
+  int            paddingLen = 0;
+  ASSERT(pPool != NULL);
+
+  if (pPool->lock) taosThreadSpinLock(pPool->lock);
+
+  ptr = pPool->ptr;
+  paddingLen = (((long)ptr + 7) & ~7) - (long)ptr;
+
+  if (pPool->node.size >= pPool->ptr - pPool->node.data + size + paddingLen) {
+    // allocate from the anchor node
+    p = pPool->ptr + paddingLen;
+    size += paddingLen;
+    pPool->ptr = pPool->ptr + size;
+    pPool->size += size;
+  } else {
+    // allocate a new node
+    pNode = taosMemoryMalloc(sizeof(*pNode) + size);
+    if (pNode == NULL) {
+      terrno = TSDB_CODE_OUT_OF_MEMORY;
+      if (pPool->lock) taosThreadSpinUnlock(pPool->lock);
+      return NULL;
+    }
+
+    p = pNode->data;
+    pNode->size = size;
+    pNode->prev = pPool->pTail;
+    pNode->pnext = &pPool->pTail;
+    pPool->pTail->pnext = &pNode->prev;
+    pPool->pTail = pNode;
+
+    pPool->size = pPool->size + sizeof(*pNode) + size;
+  }
+  if (pPool->lock) taosThreadSpinUnlock(pPool->lock);
+  return p;
+}
+
+void *vnodeBufPoolMalloc(SVBufPool *pPool, int size) {
   SVBufPoolNode *pNode;
   void          *p = NULL;
   ASSERT(pPool != NULL);

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

@@ -184,18 +184,51 @@ _err:
   return -1;
 }
 
-static int32_t vnodePrepareCommit(SVnode *pVnode) {
+static int32_t vnodePrepareCommit(SVnode *pVnode, SCommitInfo *pInfo) {
   int32_t code = 0;
+  int32_t lino = 0;
+  char    dir[TSDB_FILENAME_LEN] = {0};
+
   tsem_wait(&pVnode->canCommit);
 
+  pVnode->state.commitTerm = pVnode->state.applyTerm;
+
+  pInfo->info.config = pVnode->config;
+  pInfo->info.state.committed = pVnode->state.applied;
+  pInfo->info.state.commitTerm = pVnode->state.applyTerm;
+  pInfo->info.state.commitID = pVnode->state.commitID;
+  pInfo->pVnode = pVnode;
+  pInfo->txn = metaGetTxn(pVnode->pMeta);
+
+  // save info
+  if (pVnode->pTfs) {
+    snprintf(dir, TSDB_FILENAME_LEN, "%s%s%s", tfsGetPrimaryPath(pVnode->pTfs), TD_DIRSEP, pVnode->path);
+  } else {
+    snprintf(dir, TSDB_FILENAME_LEN, "%s", pVnode->path);
+  }
+  if (vnodeSaveInfo(dir, &pInfo->info) < 0) {
+    code = terrno;
+    TSDB_CHECK_CODE(code, lino, _exit);
+  }
+
   tsdbPrepareCommit(pVnode->pTsdb);
+
   metaPrepareAsyncCommit(pVnode->pMeta);
+
   code = smaPrepareAsyncCommit(pVnode->pSma);
   if (code) goto _exit;
 
-_exit:
   vnodeBufPoolUnRef(pVnode->inUse);
   pVnode->inUse = NULL;
+
+_exit:
+  if (code) {
+    vError("vgId:%d, %s failed at line %d since %s, commit id:%" PRId64, TD_VID(pVnode), __func__, lino,
+           tstrerror(code), pVnode->state.commitID);
+  } else {
+    vDebug("vgId:%d, %s done", TD_VID(pVnode), __func__);
+  }
+
   return code;
 }
 
@@ -217,33 +250,32 @@ _exit:
 int vnodeAsyncCommit(SVnode *pVnode) {
   int32_t code = 0;
 
-  // prepare to commit
-  code = vnodePrepareCommit(pVnode);
-  if (code) goto _exit;
-
-  // schedule the task
-  pVnode->state.commitTerm = pVnode->state.applyTerm;
-
   SCommitInfo *pInfo = (SCommitInfo *)taosMemoryCalloc(1, sizeof(*pInfo));
   if (NULL == pInfo) {
     code = TSDB_CODE_OUT_OF_MEMORY;
     goto _exit;
   }
-  pInfo->info.config = pVnode->config;
+
-  pInfo->info.state.committed = pVnode->state.applied;
+  // prepare to commit
-  pInfo->info.state.commitTerm = pVnode->state.applyTerm;
+  code = vnodePrepareCommit(pVnode, pInfo);
-  pInfo->info.state.commitID = pVnode->state.commitID;
+  if (TSDB_CODE_SUCCESS != code) {
-  pInfo->pVnode = pVnode;
+    goto _exit;
-  pInfo->txn = metaGetTxn(pVnode->pMeta);
+  }
+
+  // schedule the task
   code = vnodeScheduleTask(vnodeCommitTask, pInfo);
 
 _exit:
   if (code) {
+    if (NULL != pInfo) {
+      taosMemoryFree(pInfo);
+    }
     tsem_post(&pVnode->canCommit);
     vError("vgId:%d, %s failed since %s, commit id:%" PRId64, TD_VID(pVnode), __func__, tstrerror(code),
            pVnode->state.commitID);
   } else {
-    vDebug("vgId:%d, %s done", TD_VID(pVnode), __func__);
+    vInfo("vgId:%d, vnode async commit done, commitId:%" PRId64 " term:%" PRId64 " applied:%" PRId64, TD_VID(pVnode),
+          pVnode->state.commitID, pVnode->state.applyTerm, pVnode->state.applied);
   }
   return code;
 }
@@ -262,7 +294,7 @@ static int vnodeCommitImpl(SCommitInfo *pInfo) {
   char    dir[TSDB_FILENAME_LEN] = {0};
   SVnode *pVnode = pInfo->pVnode;
 
-  vInfo("vgId:%d, start to commit, commit ID:%" PRId64 " version:%" PRId64 " term: %" PRId64, TD_VID(pVnode),
+  vInfo("vgId:%d, start to commit, commitId:%" PRId64 " version:%" PRId64 " term: %" PRId64, TD_VID(pVnode),
         pVnode->state.commitID, pVnode->state.applied, pVnode->state.applyTerm);
 
   // persist wal before starting
@@ -271,16 +303,11 @@ static int vnodeCommitImpl(SCommitInfo *pInfo) {
     return -1;
   }
 
-  // save info
   if (pVnode->pTfs) {
     snprintf(dir, TSDB_FILENAME_LEN, "%s%s%s", tfsGetPrimaryPath(pVnode->pTfs), TD_DIRSEP, pVnode->path);
   } else {
     snprintf(dir, TSDB_FILENAME_LEN, "%s", pVnode->path);
   }
-  if (vnodeSaveInfo(dir, &pInfo->info) < 0) {
-    code = terrno;
-    TSDB_CHECK_CODE(code, lino, _exit);
-  }
 
   // walBeginSnapshot(pVnode->pWal, pVnode->state.applied);
   syncBeginSnapshot(pVnode->sync, pVnode->state.applied);

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

@@ -67,9 +67,8 @@ _err:
   return code;
 }
 
-int32_t vnodeSnapReaderClose(SVSnapReader *pReader) {
+void vnodeSnapReaderClose(SVSnapReader *pReader) {
-  int32_t code = 0;
+  vInfo("vgId:%d, close vnode snapshot reader", TD_VID(pReader->pVnode));
-
   if (pReader->pRsmaReader) {
     rsmaSnapReaderClose(&pReader->pRsmaReader);
   }
@@ -82,9 +81,7 @@ int32_t vnodeSnapReaderClose(SVSnapReader *pReader) {
     metaSnapReaderClose(&pReader->pMetaReader);
   }
 
-  vInfo("vgId:%d, vnode snapshot reader closed", TD_VID(pReader->pVnode));
   taosMemoryFree(pReader);
-  return code;
 }
 
 int32_t vnodeSnapRead(SVSnapReader *pReader, uint8_t **ppData, uint32_t *nData) {
@@ -260,8 +257,8 @@ _exit:
     pReader->index++;
     *nData = sizeof(SSnapDataHdr) + pHdr->size;
     pHdr->index = pReader->index;
-    vInfo("vgId:%d, vnode snapshot read data,index:%" PRId64 " type:%d nData:%d ", TD_VID(pReader->pVnode),
+    vDebug("vgId:%d, vnode snapshot read data, index:%" PRId64 " type:%d blockLen:%d ", TD_VID(pReader->pVnode),
-          pReader->index, pHdr->type, *nData);
+           pReader->index, pHdr->type, *nData);
   } else {
     vInfo("vgId:%d, vnode snapshot read data end, index:%" PRId64, TD_VID(pReader->pVnode), pReader->index);
   }
@@ -426,8 +423,8 @@ int32_t vnodeSnapWrite(SVSnapWriter *pWriter, uint8_t *pData, uint32_t nData) {
   ASSERT(pHdr->index == pWriter->index + 1);
   pWriter->index = pHdr->index;
 
-  vInfo("vgId:%d, vnode snapshot write data, index:%" PRId64 " type:%d nData:%d", TD_VID(pVnode), pHdr->index,
+  vDebug("vgId:%d, vnode snapshot write data, index:%" PRId64 " type:%d blockLen:%d", TD_VID(pVnode), pHdr->index,
-        pHdr->type, nData);
+         pHdr->type, nData);
 
   switch (pHdr->type) {
     case SNAP_DATA_CFG: {

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

@@ -217,6 +217,10 @@ int32_t vnodeProcessWriteMsg(SVnode *pVnode, SRpcMsg *pMsg, int64_t version, SRp
 
   if (!syncUtilUserCommit(pMsg->msgType)) goto _exit;
 
+  if (pMsg->msgType == TDMT_VND_STREAM_RECOVER_BLOCKING_STAGE || pMsg->msgType == TDMT_STREAM_TASK_CHECK_RSP) {
+    if (tqCheckLogInWal(pVnode->pTq, version)) return 0;
+  }
+
   // skip header
   pReq = POINTER_SHIFT(pMsg->pCont, sizeof(SMsgHead));
   len = pMsg->contLen - sizeof(SMsgHead);
@@ -1352,7 +1356,7 @@ static int32_t vnodeProcessBatchDeleteReq(SVnode *pVnode, int64_t version, void
     SSingleDeleteReq *pOneReq = taosArrayGet(deleteReq.deleteReqs, i);
     char             *name = pOneReq->tbname;
     if (metaGetTableEntryByName(&mr, name) < 0) {
-      vDebug("stream delete msg, skip vgId:%d since no table: %s", pVnode->config.vgId, name);
+      vDebug("vgId:%d, stream delete msg, skip since no table: %s", pVnode->config.vgId, name);
       continue;
     }
 

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

@@ -380,9 +380,8 @@ static int32_t vnodeSyncSendMsg(const SEpSet *pEpSet, SRpcMsg *pMsg) {
   return code;
 }
 
-static int32_t vnodeSyncGetSnapshot(const SSyncFSM *pFsm, SSnapshot *pSnapshot) {
+static void vnodeSyncGetSnapshotInfo(const SSyncFSM *pFsm, SSnapshot *pSnapshot) {
   vnodeGetSnapshot(pFsm->data, pSnapshot);
-  return 0;
 }
 
 static int32_t vnodeSyncApplyMsg(const SSyncFSM *pFsm, SRpcMsg *pMsg, const SFsmCbMeta *pMeta) {
@@ -392,9 +391,9 @@ static int32_t vnodeSyncApplyMsg(const SSyncFSM *pFsm, SRpcMsg *pMsg, const SFsm
 
   const STraceId *trace = &pMsg->info.traceId;
   vGTrace("vgId:%d, commit-cb is excuted, fsm:%p, index:%" PRId64 ", term:%" PRIu64 ", msg-index:%" PRId64
-          ", weak:%d, code:%d, state:%d %s, type:%s",
+          ", weak:%d, code:%d, state:%d %s, type:%s code:0x%x",
           pVnode->config.vgId, pFsm, pMeta->index, pMeta->term, pMsg->info.conn.applyIndex, pMeta->isWeak, pMeta->code,
-          pMeta->state, syncStr(pMeta->state), TMSG_INFO(pMsg->msgType));
+          pMeta->state, syncStr(pMeta->state), TMSG_INFO(pMsg->msgType), pMsg->code);
 
   return tmsgPutToQueue(&pVnode->msgCb, APPLY_QUEUE, pMsg);
 }
@@ -424,10 +423,9 @@ static int32_t vnodeSnapshotStartRead(const SSyncFSM *pFsm, void *pParam, void *
   return code;
 }
 
-static int32_t vnodeSnapshotStopRead(const SSyncFSM *pFsm, void *pReader) {
+static void vnodeSnapshotStopRead(const SSyncFSM *pFsm, void *pReader) {
   SVnode *pVnode = pFsm->data;
-  int32_t code = vnodeSnapReaderClose(pReader);
+  vnodeSnapReaderClose(pReader);
-  return code;
 }
 
 static int32_t vnodeSnapshotDoRead(const SSyncFSM *pFsm, void *pReader, void **ppBuf, int32_t *len) {
@@ -467,9 +465,9 @@ static int32_t vnodeSnapshotStopWrite(const SSyncFSM *pFsm, void *pWriter, bool
 
 static int32_t vnodeSnapshotDoWrite(const SSyncFSM *pFsm, void *pWriter, void *pBuf, int32_t len) {
   SVnode *pVnode = pFsm->data;
-  vDebug("vgId:%d, continue write vnode snapshot, len:%d", pVnode->config.vgId, len);
+  vDebug("vgId:%d, continue write vnode snapshot, blockLen:%d", pVnode->config.vgId, len);
   int32_t code = vnodeSnapWrite(pWriter, pBuf, len);
-  vDebug("vgId:%d, continue write vnode snapshot finished, len:%d", pVnode->config.vgId, len);
+  vDebug("vgId:%d, continue write vnode snapshot finished, blockLen:%d", pVnode->config.vgId, len);
   return code;
 }
 
@@ -539,7 +537,7 @@ static SSyncFSM *vnodeSyncMakeFsm(SVnode *pVnode) {
   pFsm->FpCommitCb = vnodeSyncCommitMsg;
   pFsm->FpPreCommitCb = vnodeSyncPreCommitMsg;
   pFsm->FpRollBackCb = vnodeSyncRollBackMsg;
-  pFsm->FpGetSnapshotInfo = vnodeSyncGetSnapshot;
+  pFsm->FpGetSnapshotInfo = vnodeSyncGetSnapshotInfo;
   pFsm->FpRestoreFinishCb = vnodeRestoreFinish;
   pFsm->FpLeaderTransferCb = NULL;
   pFsm->FpApplyQueueEmptyCb = vnodeApplyQueueEmpty;

source/libs/catalog/inc/catalogInt.h

@@ -485,6 +485,9 @@ typedef struct SCtgOperation {
   ctgOpFunc func;
 } SCtgOperation;
 
+#define CTG_AUTH_READ(_t) ((_t) == AUTH_TYPE_READ || (_t) == AUTH_TYPE_READ_OR_WRITE)
+#define CTG_AUTH_WRITE(_t) ((_t) == AUTH_TYPE_WRITE || (_t) == AUTH_TYPE_READ_OR_WRITE)
+
 #define CTG_QUEUE_INC() atomic_add_fetch_64(&gCtgMgmt.queue.qRemainNum, 1)
 #define CTG_QUEUE_DEC() atomic_sub_fetch_64(&gCtgMgmt.queue.qRemainNum, 1)
 
@@ -579,34 +582,34 @@ typedef struct SCtgOperation {
 #define CTG_LOCK(type, _lock)                                                                        \
   do {                                                                                               \
     if (CTG_READ == (type)) {                                                                        \
-      assert(atomic_load_32((_lock)) >= 0);                                                          \
+      ASSERTS(atomic_load_32((_lock)) >= 0, "invalid lock value before read lock");                  \
       CTG_LOCK_DEBUG("CTG RLOCK%p:%d, %s:%d B", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
       taosRLockLatch(_lock);                                                                         \
       CTG_LOCK_DEBUG("CTG RLOCK%p:%d, %s:%d E", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
-      assert(atomic_load_32((_lock)) > 0);                                                           \
+      ASSERTS(atomic_load_32((_lock)) > 0, "invalid lock value after read lock");                    \
     } else {                                                                                         \
-      assert(atomic_load_32((_lock)) >= 0);                                                          \
+      ASSERTS(atomic_load_32((_lock)) >= 0, "invalid lock value before write lock");                 \
       CTG_LOCK_DEBUG("CTG WLOCK%p:%d, %s:%d B", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
       taosWLockLatch(_lock);                                                                         \
       CTG_LOCK_DEBUG("CTG WLOCK%p:%d, %s:%d E", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
-      assert(atomic_load_32((_lock)) == TD_RWLATCH_WRITE_FLAG_COPY);                                 \
+      ASSERTS(atomic_load_32((_lock)) == TD_RWLATCH_WRITE_FLAG_COPY, "invalid lock value after write lock");    \
     }                                                                                                \
   } while (0)
 
 #define CTG_UNLOCK(type, _lock)                                                                       \
   do {                                                                                                \
     if (CTG_READ == (type)) {                                                                         \
-      assert(atomic_load_32((_lock)) > 0);                                                            \
+      ASSERTS(atomic_load_32((_lock)) > 0, "invalid lock value before read unlock");                  \
       CTG_LOCK_DEBUG("CTG RULOCK%p:%d, %s:%d B", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
       taosRUnLockLatch(_lock);                                                                        \
       CTG_LOCK_DEBUG("CTG RULOCK%p:%d, %s:%d E", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
-      assert(atomic_load_32((_lock)) >= 0);                                                           \
+      ASSERTS(atomic_load_32((_lock)) >= 0, "invalid lock value after read unlock");                  \
     } else {                                                                                          \
-      assert(atomic_load_32((_lock)) == TD_RWLATCH_WRITE_FLAG_COPY);                                  \
+      ASSERTS(atomic_load_32((_lock)) == TD_RWLATCH_WRITE_FLAG_COPY, "invalid lock value before write unlock");   \
       CTG_LOCK_DEBUG("CTG WULOCK%p:%d, %s:%d B", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
       taosWUnLockLatch(_lock);                                                                        \
       CTG_LOCK_DEBUG("CTG WULOCK%p:%d, %s:%d E", (_lock), atomic_load_32(_lock), __FILE__, __LINE__); \
-      assert(atomic_load_32((_lock)) >= 0);                                                           \
+      ASSERTS(atomic_load_32((_lock)) >= 0, "invalid lock value after write unlock");                 \
     }                                                                                                 \
   } while (0)
 

source/libs/catalog/src/catalog.c

@@ -352,9 +352,9 @@ int32_t ctgChkAuth(SCatalog* pCtg, SRequestConnInfo* pConn, const char* user, co
     goto _return;
   }
 
-  if (type == AUTH_TYPE_READ && authRsp.readDbs && taosHashGet(authRsp.readDbs, dbFName, strlen(dbFName))) {
+  if (CTG_AUTH_READ(type) && authRsp.readDbs && taosHashGet(authRsp.readDbs, dbFName, strlen(dbFName))) {
     *pass = true;
-  } else if (type == AUTH_TYPE_WRITE && authRsp.writeDbs && taosHashGet(authRsp.writeDbs, dbFName, strlen(dbFName))) {
+  } else if (CTG_AUTH_WRITE(type) && authRsp.writeDbs && taosHashGet(authRsp.writeDbs, dbFName, strlen(dbFName))) {
     *pass = true;
   }
 

source/libs/catalog/src/ctgAsync.c

@@ -471,17 +471,20 @@ int32_t ctgHandleForceUpdate(SCatalog* pCtg, int32_t taskNum, SCtgJob* pJob, con
 }
 
 int32_t ctgInitTask(SCtgJob* pJob, CTG_TASK_TYPE type, void* param, int32_t* taskId) {
+  int32_t code = 0;
   int32_t tid = atomic_fetch_add_32(&pJob->taskIdx, 1);
 
   CTG_LOCK(CTG_WRITE, &pJob->taskLock);
-  CTG_ERR_RET((*gCtgAsyncFps[type].initFp)(pJob, tid, param));
+  CTG_ERR_JRET((*gCtgAsyncFps[type].initFp)(pJob, tid, param));
-  CTG_UNLOCK(CTG_WRITE, &pJob->taskLock);
 
   if (taskId) {
     *taskId = tid;
   }
 
-  return TSDB_CODE_SUCCESS;
+_return:
+  CTG_UNLOCK(CTG_WRITE, &pJob->taskLock);
+  
+  return code;
 }
 
 int32_t ctgInitJob(SCatalog* pCtg, SRequestConnInfo* pConn, SCtgJob** job, const SCatalogReq* pReq, catalogCallback fp,
@@ -1549,10 +1552,10 @@ int32_t ctgHandleGetUserRsp(SCtgTaskReq* tReq, int32_t reqType, const SDataBuf*
     goto _return;
   }
 
-  if (ctx->user.type == AUTH_TYPE_READ && pOut->readDbs &&
+  if (CTG_AUTH_READ(ctx->user.type) && pOut->readDbs &&
       taosHashGet(pOut->readDbs, ctx->user.dbFName, strlen(ctx->user.dbFName))) {
     pass = true;
-  } else if (ctx->user.type == AUTH_TYPE_WRITE && pOut->writeDbs &&
+  } else if (CTG_AUTH_WRITE(ctx->user.type) && pOut->writeDbs &&
              taosHashGet(pOut->writeDbs, ctx->user.dbFName, strlen(ctx->user.dbFName))) {
     pass = true;
   }

source/libs/catalog/src/ctgCache.c

@@ -718,11 +718,11 @@ int32_t ctgChkAuthFromCache(SCatalog *pCtg, char *user, char *dbFName, AUTH_TYPE
     return TSDB_CODE_SUCCESS;
   }
 
-  if (pUser->readDbs && taosHashGet(pUser->readDbs, dbFName, strlen(dbFName)) && type == AUTH_TYPE_READ) {
+  if (pUser->readDbs && taosHashGet(pUser->readDbs, dbFName, strlen(dbFName)) && CTG_AUTH_READ(type)) {
     *pass = true;
   }
 
-  if (pUser->writeDbs && taosHashGet(pUser->writeDbs, dbFName, strlen(dbFName)) && type == AUTH_TYPE_WRITE) {
+  if (pUser->writeDbs && taosHashGet(pUser->writeDbs, dbFName, strlen(dbFName)) && CTG_AUTH_WRITE(type)) {
     *pass = true;
   }
 
@@ -2500,6 +2500,7 @@ int32_t ctgGetTbMetasFromCache(SCatalog *pCtg, SRequestConnInfo *pConn, SCtgTbMe
 
     CTG_LOCK(CTG_READ, &pCache->metaLock);
     if (NULL == pCache->pMeta) {
+      CTG_UNLOCK(CTG_READ, &pCache->metaLock);
       ctgDebug("tb %s meta not in cache, dbFName:%s", pName->tname, dbFName);
       ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
       taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);

source/libs/catalog/src/ctgDbg.c

@@ -22,7 +22,6 @@ extern SCatalogMgmt gCtgMgmt;
 SCtgDebug           gCTGDebug = {0};
 
 void ctgdUserCallback(SMetaData *pResult, void *param, int32_t code) {
-  ASSERT(*(int32_t *)param == 1);
   taosMemoryFree(param);
 
   qDebug("async call result: %s", tstrerror(code));

source/libs/catalog/src/ctgRemote.c

@@ -40,7 +40,9 @@ int32_t ctgHandleBatchRsp(SCtgJob* pJob, SCtgTaskCallbackParam* cbParam, SDataBu
     msgNum = taosArrayGetSize(batchRsp.pRsps);
   }
   
-  ASSERT(taskNum == msgNum || 0 == msgNum);
+  if (ASSERTS(taskNum == msgNum || 0 == msgNum, "taskNum %d mis-match msgNum %d", taskNum, msgNum)) {
+    msgNum = 0;
+  }
 
   ctgDebug("QID:0x%" PRIx64 " ctg got batch %d rsp %s", pJob->queryId, cbParam->batchId,
            TMSG_INFO(cbParam->reqType + 1));
@@ -58,11 +60,19 @@ int32_t ctgHandleBatchRsp(SCtgJob* pJob, SCtgTaskCallbackParam* cbParam, SDataBu
     if (msgNum > 0) {
       pRsp = taosArrayGet(batchRsp.pRsps, i);
 
-      taskMsg.msgType = pRsp->reqType;
+      if (ASSERTS(pRsp->msgIdx == *msgIdx, "rsp msgIdx %d mis-match msgIdx %d", pRsp->msgIdx, *msgIdx)) {
-      taskMsg.pData = pRsp->msg;
+        pRsp = &rsp;
-      taskMsg.len = pRsp->msgLen;
+        pRsp->msgIdx = *msgIdx;
-
+        pRsp->reqType = -1;
-      ASSERT(pRsp->msgIdx == *msgIdx);
+        pRsp->rspCode = 0;
+        taskMsg.msgType = -1;
+        taskMsg.pData = NULL;
+        taskMsg.len = 0;
+      } else {
+        taskMsg.msgType = pRsp->reqType;
+        taskMsg.pData = pRsp->msg;
+        taskMsg.len = pRsp->msgLen;
+      }
     } else {
       pRsp = &rsp;
       pRsp->msgIdx = *msgIdx;

source/libs/command/src/command.c

@@ -21,6 +21,7 @@
 #include "tdatablock.h"
 #include "tglobal.h"
 #include "tgrant.h"
+#include "taosdef.h"
 
 extern SConfig* tsCfg;
 
@@ -40,7 +41,6 @@ static int32_t buildRetrieveTableRsp(SSDataBlock* pBlock, int32_t numOfCols, SRe
   (*pRsp)->numOfCols = htonl(numOfCols);
 
   int32_t len = blockEncode(pBlock, (*pRsp)->data, numOfCols);
-  ASSERT(len == rspSize - sizeof(SRetrieveTableRsp));
 
   return TSDB_CODE_SUCCESS;
 }
@@ -154,6 +154,23 @@ static int32_t buildCreateDBResultDataBlock(SSDataBlock** pOutput) {
   return code;
 }
 
+static int32_t buildAliveResultDataBlock(SSDataBlock** pOutput) {
+  SSDataBlock* pBlock = createDataBlock();
+  if (NULL == pBlock) {
+    return TSDB_CODE_OUT_OF_MEMORY;
+  }
+
+  SColumnInfoData infoData = createColumnInfoData(TSDB_DATA_TYPE_INT, sizeof(int32_t), 1);
+  int32_t         code = blockDataAppendColInfo(pBlock, &infoData);
+
+  if (TSDB_CODE_SUCCESS == code) {
+    *pOutput = pBlock;
+  } else {
+    blockDataDestroy(pBlock);
+  }
+  return code;
+}
+
 int64_t getValOfDiffPrecision(int8_t unit, int64_t val) {
   int64_t v = 0;
   switch (unit) {
@@ -281,6 +298,108 @@ static void setCreateDBResultIntoDataBlock(SSDataBlock* pBlock, char* dbFName, S
   colDataAppend(pCol2, 0, buf2, false);
 }
 
+#define CHECK_LEADER(n) (row[n] && (fields[n].type == TSDB_DATA_TYPE_VARCHAR && strncasecmp(row[n], "leader", varDataLen((char *)row[n] - VARSTR_HEADER_SIZE)) == 0))
+// on this row, if have leader return true else return false
+bool existLeaderRole(TAOS_ROW row, TAOS_FIELD* fields, int nFields) {
+  // vgroup_id | db_name | tables | v1_dnode | v1_status | v2_dnode | v2_status | v3_dnode | v3_status | v4_dnode |
+  // v4_status |  cacheload  | tsma |
+  if (nFields != 13) {
+    return false;
+  }
+
+  // check have leader on cloumn v*_status on 4 6 8 10
+  if (CHECK_LEADER(4) || CHECK_LEADER(6) || CHECK_LEADER(8) || CHECK_LEADER(10)) {
+    return true;
+  }
+
+  return false;
+}
+
+// get db alive status, return 1 is alive else return 0
+int32_t getAliveStatusFromApi(int64_t* pConnId, char* dbName, int32_t* pStatus) {
+  char    sql[128 + TSDB_DB_NAME_LEN] = "select * from information_schema.ins_vgroups";
+  int32_t code;
+
+  // filter with db name
+  if (dbName && dbName[0] != 0) {
+    char str[64 + TSDB_DB_NAME_LEN] = "";
+    // test db name exist
+    sprintf(str, "show create database %s ;", dbName);
+    TAOS_RES* dbRes = taos_query(pConnId, str);
+    code = taos_errno(dbRes);
+    if (code != TSDB_CODE_SUCCESS) {
+      taos_free_result(dbRes);
+      return code;
+    }
+    taos_free_result(dbRes);
+
+    sprintf(str, " where db_name='%s' ;", dbName);
+    strcat(sql, str);
+  }
+
+  TAOS_RES* res = taos_query(pConnId, sql);
+  code = taos_errno(res);
+  if (code != TSDB_CODE_SUCCESS) {
+    taos_free_result(res);
+    return code;
+  }
+
+  TAOS_ROW    row = NULL;
+  TAOS_FIELD* fields = taos_fetch_fields(res);
+  int32_t     nFields = taos_num_fields(res);
+  int32_t     nAvailble = 0;
+  int32_t     nUnAvailble = 0;
+
+  while ((row = taos_fetch_row(res)) != NULL) {
+    if (existLeaderRole(row, fields, nFields)) {
+      nAvailble++;
+    } else {
+      nUnAvailble++;
+    }
+  }
+  taos_free_result(res);
+
+  int32_t status = 0;
+  if (nAvailble + nUnAvailble == 0 || nUnAvailble == 0) {
+    status = SHOW_STATUS_AVAILABLE;
+  } else if (nAvailble > 0 && nUnAvailble > 0) {
+    status = SHOW_STATUS_HALF_AVAILABLE;
+  } else {
+    status = SHOW_STATUS_NOT_AVAILABLE;
+  }
+
+  if (pStatus) {
+    *pStatus = status;
+  }
+  return TSDB_CODE_SUCCESS;
+}
+
+static int32_t setAliveResultIntoDataBlock(int64_t* pConnId, SSDataBlock* pBlock, char* dbName) {
+  blockDataEnsureCapacity(pBlock, 1);
+  pBlock->info.rows = 1;
+
+  SColumnInfoData* pCol1 = taosArrayGet(pBlock->pDataBlock, 0);
+  int32_t          status = 0;
+  int32_t          code = getAliveStatusFromApi(pConnId, dbName, &status);
+  if (code == TSDB_CODE_SUCCESS) {
+    colDataAppend(pCol1, 0, (const char*)&status, false);
+  }
+  return code;
+}
+
+static int32_t execShowAliveStatus(int64_t* pConnId, SShowAliveStmt* pStmt, SRetrieveTableRsp** pRsp) {
+  SSDataBlock* pBlock = NULL;
+  int32_t      code = buildAliveResultDataBlock(&pBlock);
+  if (TSDB_CODE_SUCCESS == code) {
+    code = setAliveResultIntoDataBlock(pConnId, pBlock, pStmt->dbName);
+  }
+  if (TSDB_CODE_SUCCESS == code) {
+    code = buildRetrieveTableRsp(pBlock, SHOW_ALIVE_RESULT_COLS, pRsp);
+  }
+  blockDataDestroy(pBlock);
+  return code;
+}
+
 static int32_t execShowCreateDatabase(SShowCreateDatabaseStmt* pStmt, SRetrieveTableRsp** pRsp) {
   SSDataBlock* pBlock = NULL;
   int32_t      code = buildCreateDBResultDataBlock(&pBlock);
@@ -736,7 +855,7 @@ static int32_t execSelectWithoutFrom(SSelectStmt* pSelect, SRetrieveTableRsp** p
   return code;
 }
 
-int32_t qExecCommand(bool sysInfoUser, SNode* pStmt, SRetrieveTableRsp** pRsp) {
+int32_t qExecCommand(int64_t* pConnId, bool sysInfoUser, SNode* pStmt, SRetrieveTableRsp** pRsp) {
   switch (nodeType(pStmt)) {
     case QUERY_NODE_DESCRIBE_STMT:
       return execDescribe(sysInfoUser, pStmt, pRsp);
@@ -754,6 +873,9 @@ int32_t qExecCommand(bool sysInfoUser, SNode* pStmt, SRetrieveTableRsp** pRsp) {
       return execShowLocalVariables(pRsp);
     case QUERY_NODE_SELECT_STMT:
       return execSelectWithoutFrom((SSelectStmt*)pStmt, pRsp);
+    case QUERY_NODE_SHOW_DB_ALIVE_STMT:
+    case QUERY_NODE_SHOW_CLUSTER_ALIVE_STMT:
+      return execShowAliveStatus(pConnId, (SShowAliveStmt*)pStmt, pRsp);       
     default:
       break;
   }

source/libs/command/src/explain.c

@@ -1666,7 +1666,6 @@ int32_t qExplainGetRspFromCtx(void *ctx, SRetrieveTableRsp **pRsp) {
   rsp->numOfRows = htobe64((int64_t)rowNum);
 
   int32_t len = blockEncode(pBlock, rsp->data, taosArrayGetSize(pBlock->pDataBlock));
-  ASSERT(len == rspSize - sizeof(SRetrieveTableRsp));
 
   rsp->compLen = htonl(len);
 

source/libs/executor/inc/executorimpl.h

@@ -138,6 +138,7 @@ typedef struct {
   SSchemaWrapper*     schema;
   char                tbName[TSDB_TABLE_NAME_LEN];
   int8_t              recoverStep;
+  int8_t              recoverScanFinished;
   SQueryTableDataCond tableCond;
   int64_t             fillHistoryVer1;
   int64_t             fillHistoryVer2;
@@ -471,6 +472,11 @@ typedef struct SStreamScanInfo {
   SNodeList* pGroupTags;
   SNode*     pTagCond;
   SNode*     pTagIndexCond;
+
+  // recover
+  int32_t blockRecoverContiCnt;
+  int32_t blockRecoverTotCnt;
+
 } SStreamScanInfo;
 
 typedef struct {
@@ -659,7 +665,6 @@ typedef struct SStreamFillOperatorInfo {
   SSDataBlock*          pRes;
   SSDataBlock*          pSrcBlock;
   int32_t               srcRowIndex;
-  SSDataBlock*          pPrevSrcBlock;
   SSDataBlock*          pSrcDelBlock;
   int32_t               srcDelRowIndex;
   SSDataBlock*          pDelRes;

source/libs/executor/src/cachescanoperator.c

@@ -90,7 +90,7 @@ SOperatorInfo* createCacherowsScanOperator(SLastRowScanPhysiNode* pScanNode, SRe
 
     uint64_t suid = tableListGetSuid(pTableList);
     code = tsdbCacherowsReaderOpen(pInfo->readHandle.vnode, pInfo->retrieveType, pList, totalTables,
-                                   taosArrayGetSize(pInfo->matchInfo.pList), suid, &pInfo->pLastrowReader);
+                                   taosArrayGetSize(pInfo->matchInfo.pList), suid, &pInfo->pLastrowReader, pTaskInfo->id.str);
     if (code != TSDB_CODE_SUCCESS) {
       goto _error;
     }
@@ -216,7 +216,7 @@ SSDataBlock* doScanCache(SOperatorInfo* pOperator) {
       }
 
       tsdbCacherowsReaderOpen(pInfo->readHandle.vnode, pInfo->retrieveType, pList, num,
-                              taosArrayGetSize(pInfo->matchInfo.pList), suid, &pInfo->pLastrowReader);
+                              taosArrayGetSize(pInfo->matchInfo.pList), suid, &pInfo->pLastrowReader, pTaskInfo->id.str);
       taosArrayClear(pInfo->pUidList);
 
       code = tsdbRetrieveCacheRows(pInfo->pLastrowReader, pInfo->pRes, pInfo->pSlotIds, pInfo->pUidList);

source/libs/executor/src/dataDeleter.c

@@ -62,8 +62,8 @@ static void toDataCacheEntry(SDataDeleterHandle* pHandle, const SInputData* pInp
   pEntry->numOfCols = taosArrayGetSize(pInput->pData->pDataBlock);
   pEntry->dataLen = sizeof(SDeleterRes);
 
-  ASSERT(1 == pEntry->numOfRows);
+//  ASSERT(1 == pEntry->numOfRows);
-  ASSERT(3 == pEntry->numOfCols);
+//  ASSERT(3 == pEntry->numOfCols);
 
   pBuf->useSize = sizeof(SDataCacheEntry);
 
@@ -167,7 +167,6 @@ static void getDataLength(SDataSinkHandle* pHandle, int64_t* pLen, bool* pQueryE
 
   SDataDeleterBuf* pBuf = NULL;
   taosReadQitem(pDeleter->pDataBlocks, (void**)&pBuf);
-  ASSERT(NULL != pBuf);
   memcpy(&pDeleter->nextOutput, pBuf, sizeof(SDataDeleterBuf));
   taosFreeQitem(pBuf);
 

source/libs/executor/src/dataDispatcher.c

@@ -77,8 +77,8 @@ static void toDataCacheEntry(SDataDispatchHandle* pHandle, const SInputData* pIn
 
   pBuf->useSize = sizeof(SDataCacheEntry);
   pEntry->dataLen = blockEncode(pInput->pData, pEntry->data, numOfCols);
-  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
+//  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
-  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
+//  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
 
   pBuf->useSize += pEntry->dataLen;
 
@@ -162,15 +162,14 @@ static void getDataLength(SDataSinkHandle* pHandle, int64_t* pLen, bool* pQueryE
 
   SDataDispatchBuf* pBuf = NULL;
   taosReadQitem(pDispatcher->pDataBlocks, (void**)&pBuf);
-  ASSERT(NULL != pBuf);
   memcpy(&pDispatcher->nextOutput, pBuf, sizeof(SDataDispatchBuf));
   taosFreeQitem(pBuf);
 
   SDataCacheEntry* pEntry = (SDataCacheEntry*)pDispatcher->nextOutput.pData;
   *pLen = pEntry->dataLen;
 
-  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
+//  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
-  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
+//  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
 
   *pQueryEnd = pDispatcher->queryEnd;
   qDebug("got data len %" PRId64 ", row num %d in sink", *pLen,
@@ -193,8 +192,8 @@ static int32_t getDataBlock(SDataSinkHandle* pHandle, SOutputData* pOutput) {
   pOutput->numOfCols = pEntry->numOfCols;
   pOutput->compressed = pEntry->compressed;
 
-  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
+//  ASSERT(pEntry->numOfRows == *(int32_t*)(pEntry->data + 8));
-  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
+//  ASSERT(pEntry->numOfCols == *(int32_t*)(pEntry->data + 8 + 4));
 
   atomic_sub_fetch_64(&pDispatcher->cachedSize, pEntry->dataLen);
   atomic_sub_fetch_64(&gDataSinkStat.cachedSize, pEntry->dataLen);