Merge branch '3.0' into feature/TD-14761

CMakeLists.txt

@@ -34,7 +34,7 @@ endif(${BUILD_TEST})
 
 add_subdirectory(source)
 add_subdirectory(tools)
-add_subdirectory(tests)
+add_subdirectory(utils)
 add_subdirectory(examples/c)
 
 # docs 

cmake/taostools_CMakeLists.txt.in

@@ -2,7 +2,7 @@
 # taos-tools
 ExternalProject_Add(taos-tools
         GIT_REPOSITORY https://github.com/taosdata/taos-tools.git
-        GIT_TAG 9cb965f
+        GIT_TAG a4d9b92
         SOURCE_DIR "${TD_SOURCE_DIR}/tools/taos-tools"
         BINARY_DIR ""
         #BUILD_IN_SOURCE TRUE

cmake/taosws_CMakeLists.txt.in

@@ -2,7 +2,7 @@
 # taosws-rs
 ExternalProject_Add(taosws-rs
         GIT_REPOSITORY https://github.com/taosdata/taos-connector-rust.git
-        GIT_TAG 7a54d21
+        GIT_TAG 6fc47d7
         SOURCE_DIR "${TD_SOURCE_DIR}/tools/taosws-rs"
         BINARY_DIR ""
         #BUILD_IN_SOURCE TRUE

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

@@ -71,9 +71,9 @@ 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.
-- 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 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.
+- 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 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.
+- 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 is 0. A value of 0 indicates that a new WAL file is created only after the previous WAL file was written to disk.
+- 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.
 - WAL_SEGMENT_SIZE: specifies the maximum size of a WAL file. After the current WAL file reaches this size, a new WAL file is created. The default value is 0. A value of 0 indicates that a new WAL file is created only after the previous WAL file was written to disk.
 
 ### Example Statement

docs/en/12-taos-sql/06-select.md

@@ -52,11 +52,6 @@ window_clause: {
   | STATE_WINDOW(col)
   | INTERVAL(interval_val [, interval_offset]) [SLIDING (sliding_val)] [WATERMARK(watermark_val)] [FILL(fill_mod_and_val)]
 
-changes_option: {
-    DURATION duration_val
-  | ROWS rows_val
-}
-
 group_by_clause:
     GROUP BY expr [, expr] ... HAVING condition
 
@@ -126,7 +121,6 @@ SELECT DISTINCT col_name [, col_name ...] FROM tb_name;
 
 1. Configuration parameter `maxNumOfDistinctRes` in `taos.cfg` is used to control the number of rows to output. The minimum configurable value is 100,000, the maximum configurable value is 100,000,000, the default value is 1,000,000. If the actual number of rows exceeds the value of this parameter, only the number of rows specified by this parameter will be output.
 2. It can't be guaranteed that the results selected by using `DISTINCT` on columns of `FLOAT` or `DOUBLE` are exactly unique because of the precision errors in floating point numbers.
-3. `DISTINCT` can't be used in the sub-query of a nested query statement, and can't be used together with aggregate functions, `GROUP BY` or `JOIN` in the same SQL statement.
 
 :::
 

docs/en/12-taos-sql/10-function.md

@@ -613,6 +613,7 @@ SELECT APERCENTILE(field_name, P[, algo_type]) FROM { tb_name | stb_name } [WHER
 **Explanations**：
 - _P_ is in range [0,100], when _P_ is 0, the result is same as using function MIN; when _P_ is 100, the result is same as function MAX.
 - `algo_type` can only be input as `default` or `t-digest` Enter `default` to use a histogram-based algorithm. Enter `t-digest` to use the t-digest algorithm to calculate the approximation of the quantile. `default` is used by default.
+- The approximation result of `t-digest` algorithm is sensitive to input data order. For example, when querying STable with different input data order there might be minor differences in calculated results.
 
 ### AVG
 
@@ -916,7 +917,7 @@ SELECT MAX(field_name) FROM { tb_name | stb_name } [WHERE clause];
 
 **Return value type**:Same as the data type of the column being operated upon
 
-**Applicable data types**: Numeric
+**Applicable data types**: Numeric, Timestamp
 
 **Applicable table types**: standard tables and supertables
 
@@ -931,7 +932,7 @@ SELECT MIN(field_name) FROM {tb_name | stb_name} [WHERE clause];
 
 **Return value type**:Same as the data type of the column being operated upon
 
-**Applicable data types**: Numeric
+**Applicable data types**: Numeric, Timestamp
 
 **Applicable table types**: standard tables and supertables
 

docs/en/12-taos-sql/14-stream.md

@@ -44,7 +44,7 @@ For example, the following SQL statement creates a stream and automatically crea
 
 ```sql
 CREATE STREAM avg_vol_s INTO avg_vol AS
-SELECT _wstartts, count(*), avg(voltage) FROM meters PARTITION BY tbname INTERVAL(1m) SLIDING(30s);
+SELECT _wstart, count(*), avg(voltage) FROM meters PARTITION BY tbname INTERVAL(1m) SLIDING(30s);
 ```
 
 ## Delete a Stream

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

@@ -5,16 +5,6 @@ title: SHOW Statement for Metadata
 
 `SHOW` command can be used to get brief system information. To get details about metatadata, information, and status in the system, please use `select` to query the tables in database `INFORMATION_SCHEMA`. 
 
-## SHOW ACCOUNTS
-
-```sql
-SHOW ACCOUNTS;
-```
-
-Shows information about tenants on the system.
-
-Note: TDengine Enterprise Edition only.
-
 ## SHOW APPS
 
 ```sql

docs/zh/07-develop/_sub_java.mdx

@@ -1,7 +1,5 @@
 ```java
 {{#include docs/examples/java/src/main/java/com/taos/example/SubscribeDemo.java}}
-{{#include docs/examples/java/src/main/java/com/taos/example/MetersDeserializer.java}}
-{{#include docs/examples/java/src/main/java/com/taos/example/Meters.java}}
 ```
 ```java
 {{#include docs/examples/java/src/main/java/com/taos/example/MetersDeserializer.java}}

docs/zh/08-connector/09-csharp.mdx

@@ -169,9 +169,9 @@ namespace TDengineExample
 
 ### 第三方驱动
 
-`Maikebing.Data.Taos` 是一个 TDengine 的 ADO.NET 连接器，支持 Linux，Windows 平台。该连接器由社区贡献者`麦壳饼@@maikebing` 提供，具体请参考:
+[`IoTSharp.Data.Taos`](https://github.com/IoTSharp/EntityFrameworkCore.Taos) 是一个 TDengine 的 ADO.NET 连接器,其中包含了用于EntityFrameworkCore 的提供程序 IoTSharp.EntityFrameworkCore.Taos 和健康检查组件 IoTSharp.HealthChecks.Taos ，支持 Linux，Windows 平台。该连接器由社区贡献者`麦壳饼@@maikebing` 提供，具体请参考:
 
-* 接口下载:<https://github.com/maikebing/Maikebing.EntityFrameworkCore.Taos>
+* 接口下载: <https://github.com/IoTSharp/EntityFrameworkCore.Taos>
 * 用法说明:<https://www.taosdata.com/blog/2020/11/02/1901.html>
 
 ## 常见问题

docs/zh/12-taos-sql/01-data-type.md

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 支持的数据类型
+sidebar_label: 数据类型
-title: 支持的数据类型
+title: 数据类型
 description: "TDengine 支持的数据类型: 时间戳、浮点型、JSON 类型等"
 ---
 

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

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 数据库管理
+sidebar_label: 数据库
-title: 数据库管理
+title: 数据库
 description: "创建、删除数据库，查看、修改数据库参数"
 ---
 
@@ -71,9 +71,9 @@ 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。
-- 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。默认为 0，即落盘后立即删除。-1 表示不删除。
+- 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。默认为 0，即落盘后立即删除。-1 表示不删除。
+- 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。默认为 0，即仅在落盘时创建新文件。
+- 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。
 - WAL_SEGMENT_SIZE：wal 单个文件大小，单位为 KB。当前写入文件大小超过上限后会自动创建一个新的 wal 文件。默认为 0，即仅在落盘时创建新文件。
 
 ### 创建数据库示例

docs/zh/12-taos-sql/03-table.md

@@ -1,5 +1,5 @@
 ---
-title: 表管理
+title: 表
 sidebar_label: 表
 description: 对表的各种管理操作
 ---

docs/zh/12-taos-sql/04-stable.md

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 超级表管理
+sidebar_label: 超级表
-title: 超级表 STable 管理
+title: 超级表
 description: 对超级表的各种管理操作
 ---
 

docs/zh/12-taos-sql/06-select.md

@@ -53,11 +53,6 @@ window_clause: {
   | STATE_WINDOW(col)
   | INTERVAL(interval_val [, interval_offset]) [SLIDING (sliding_val)] [WATERMARK(watermark_val)] [FILL(fill_mod_and_val)]
 
-changes_option: {
-    DURATION duration_val
-  | ROWS rows_val
-}
-
 group_by_clause:
     GROUP BY expr [, expr] ... HAVING condition
 
@@ -127,7 +122,6 @@ SELECT DISTINCT col_name [, col_name ...] FROM tb_name;
 
 1. cfg 文件中的配置参数 maxNumOfDistinctRes 将对 DISTINCT 能够输出的数据行数进行限制。其最小值是 100000，最大值是 100000000，默认值是 10000000。如果实际计算结果超出了这个限制，那么会仅输出这个数量范围内的部分。
 2. 由于浮点数天然的精度机制原因，在特定情况下，对 FLOAT 和 DOUBLE 列使用 DISTINCT 并不能保证输出值的完全唯一性。
-3. 在当前版本下，DISTINCT 不能在嵌套查询的子查询中使用，也不能与聚合函数、GROUP BY、或 JOIN 在同一条语句中混用。
 
 :::
 

docs/zh/12-taos-sql/10-function.md

@@ -614,6 +614,7 @@ SELECT APERCENTILE(field_name, P[, algo_type]) FROM { tb_name | stb_name } [WHER
 **说明**：
 - P值范围是[0,100]，当为0时等同于MIN，为100时等同于MAX。
 - algo_type 取值为 "default" 或 "t-digest"。 输入为 "default" 时函数使用基于直方图算法进行计算。输入为 "t-digest" 时使用t-digest算法计算分位数的近似结果。如果不指定 algo_type 则使用 "default" 算法。
+- "t-digest"算法的近似结果对于输入数据顺序敏感，对超级表查询时不同的输入排序结果可能会有微小的误差。
 
 ### AVG
 

docs/zh/12-taos-sql/12-distinguished.md

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 时序数据特色查询
+sidebar_label: 特色查询
-title: 时序数据特色查询
+title: 特色查询
 description: TDengine 提供的时序数据特有的查询功能
 ---
 

docs/zh/12-taos-sql/14-stream.md

@@ -44,7 +44,7 @@ window_clause: {
 
 ```sql
 CREATE STREAM avg_vol_s INTO avg_vol AS
-SELECT _wstartts, count(*), avg(voltage) FROM meters PARTITION BY tbname INTERVAL(1m) SLIDING(30s);
+SELECT _wstart, count(*), avg(voltage) FROM meters PARTITION BY tbname INTERVAL(1m) SLIDING(30s);
 ```
 
 ## 流式计算的 partition

docs/zh/12-taos-sql/17-json.md

@@ -1,6 +1,6 @@
 ---
-sidebar_label: JSON 类型使用说明
+sidebar_label: JSON 类型
-title: JSON 类型使用说明
+title: JSON 类型
 description: 对 JSON 类型如何使用的详细说明
 ---
 

docs/zh/12-taos-sql/18-escape.md

@@ -1,5 +1,5 @@
 ---
-title: 转义字符说明
+title: 转义字符
 sidebar_label: 转义字符
 description: TDengine 中使用转义字符的详细规则
 ---

docs/zh/12-taos-sql/19-limit.md

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 命名与边界限制
+sidebar_label: 命名与边界
-title: 命名与边界限制
+title: 命名与边界
 description: 合法字符集和命名中的限制规则
 ---
 

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

@@ -1,6 +1,6 @@
 ---
 sidebar_label: 保留关键字
-title: TDengine 保留关键字
+title: 保留关键字
 description: TDengine 保留关键字的详细列表
 ---
 

docs/zh/12-taos-sql/22-meta.md

@@ -1,6 +1,6 @@
 ---
 sidebar_label: 元数据
-title: 存储元数据的 Information_Schema 数据库
+title: 元数据
 description: Information_Schema 数据库中存储了系统中所有的元数据信息
 ---
 

docs/zh/12-taos-sql/23-perf.md

@@ -1,6 +1,6 @@
 ---
 sidebar_label: 统计数据
-title: 存储统计数据的 Performance_Schema 数据库
+title: 统计数据
 description: Performance_Schema 数据库中存储了系统中的各种统计信息
 ---
 

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

@@ -1,21 +1,11 @@
 ---
 sidebar_label: SHOW 命令
-title: 使用 SHOW 命令查看系统元数据
+title: SHOW 命令
 description: SHOW 命令的完整列表
 ---
 
 SHOW 命令可以用来获取简要的系统信息。若想获取系统中详细的各种元数据、系统信息和状态，请使用 select 语句查询 INFORMATION_SCHEMA 数据库中的表。
 
-## SHOW ACCOUNTS
-
-```sql
-SHOW ACCOUNTS;
-```
-
-显示当前系统中所有租户的信息。
-
-注：企业版独有
-
 ## SHOW APPS
 
 ```sql

docs/zh/12-taos-sql/26-udf.md

@@ -1,6 +1,6 @@
 ---
 sidebar_label: 自定义函数
-title: 用户自定义函数
+title: 自定义函数
 description: 使用 UDF 的详细指南
 ---
 

docs/zh/12-taos-sql/27-index.md

@@ -1,6 +1,6 @@
 ---
 sidebar_label: 索引
-title: 使用索引
+title: 索引
 description: 索引功能的使用细节
 ---
 

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

@@ -1,6 +1,6 @@
 ---
-sidebar_label: 3.0 版本语法变更
+sidebar_label: 语法变更
-title: 3.0 版本语法变更
+title: 语法变更
 description: "TDengine 3.0 版本的语法变更说明"
 ---
 

include/common/tdatablock.h

@@ -184,7 +184,8 @@ static FORCE_INLINE void colDataAppendDouble(SColumnInfoData* pColumnInfoData, u
 int32_t getJsonValueLen(const char* data);
 
 int32_t colDataAppend(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData, bool isNull);
-int32_t colDataAppendNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData, uint32_t numOfRows);
+int32_t colDataAppendNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData,
+                            uint32_t numOfRows);
 int32_t colDataMergeCol(SColumnInfoData* pColumnInfoData, int32_t numOfRow1, int32_t* capacity,
                         const SColumnInfoData* pSource, int32_t numOfRow2);
 int32_t colDataAssign(SColumnInfoData* pColumnInfoData, const SColumnInfoData* pSource, int32_t numOfRows,
@@ -232,6 +233,7 @@ SSDataBlock* createDataBlock();
 void*        blockDataDestroy(SSDataBlock* pBlock);
 void         blockDataFreeRes(SSDataBlock* pBlock);
 SSDataBlock* createOneDataBlock(const SSDataBlock* pDataBlock, bool copyData);
+SSDataBlock* createSpecialDataBlock(EStreamType type);
 
 int32_t blockDataAppendColInfo(SSDataBlock* pBlock, SColumnInfoData* pColInfoData);
 
@@ -249,7 +251,6 @@ char* dumpBlockData(SSDataBlock* pDataBlock, const char* flag, char** dumpBuf);
 int32_t buildSubmitReqFromDataBlock(SSubmitReq** pReq, const SSDataBlock* pDataBlocks, STSchema* pTSchema, int32_t vgId,
                                     tb_uid_t suid);
 
-
 char* buildCtbNameByGroupId(const char* stbName, uint64_t groupId);
 
 static FORCE_INLINE int32_t blockGetEncodeSize(const SSDataBlock* pBlock) {

include/common/tglobal.h

@@ -94,6 +94,7 @@ extern int64_t tsQueryBufferSizeBytes;   // maximum allowed usage buffer size in
 // query client
 extern int32_t tsQueryPolicy;
 extern int32_t tsQuerySmaOptimize;
+extern bool    tsQueryPlannerTrace;
 
 // client
 extern int32_t tsMinSlidingTime;

include/common/tmsg.h

@@ -784,6 +784,9 @@ typedef struct {
   int64_t walRetentionSize;
   int32_t walRollPeriod;
   int64_t walSegmentSize;
+  int32_t sstTrigger;
+  int16_t hashPrefix;
+  int16_t hashSuffix;
 } SCreateDbReq;
 
 int32_t tSerializeSCreateDbReq(void* buf, int32_t bufLen, SCreateDbReq* pReq);
@@ -805,6 +808,7 @@ typedef struct {
   int8_t  strict;
   int8_t  cacheLast;
   int8_t  replications;
+  int32_t sstTrigger;
 } SAlterDbReq;
 
 int32_t tSerializeSAlterDbReq(void* buf, int32_t bufLen, SAlterDbReq* pReq);
@@ -1190,6 +1194,9 @@ typedef struct {
   int64_t  walRetentionSize;
   int32_t  walRollPeriod;
   int64_t  walSegmentSize;
+  int16_t  sstTrigger;
+  int16_t  hashPrefix;
+  int16_t  hashSuffix;
 } SCreateVnodeReq;
 
 int32_t tSerializeSCreateVnodeReq(void* buf, int32_t bufLen, SCreateVnodeReq* pReq);

include/common/ttokendef.h

@@ -102,224 +102,227 @@
 #define TK_WAL_RETENTION_SIZE   84
 #define TK_WAL_ROLL_PERIOD      85
 #define TK_WAL_SEGMENT_SIZE     86
-#define TK_NK_COLON             87
+#define TK_SST_TRIGGER          87
-#define TK_TABLE                88
+#define TK_TABLE_PREFIX         88
-#define TK_NK_LP                89
+#define TK_TABLE_SUFFIX         89
-#define TK_NK_RP                90
+#define TK_NK_COLON             90
-#define TK_STABLE               91
+#define TK_TABLE                91
-#define TK_ADD                  92
+#define TK_NK_LP                92
-#define TK_COLUMN               93
+#define TK_NK_RP                93
-#define TK_MODIFY               94
+#define TK_STABLE               94
-#define TK_RENAME               95
+#define TK_ADD                  95
-#define TK_TAG                  96
+#define TK_COLUMN               96
-#define TK_SET                  97
+#define TK_MODIFY               97
-#define TK_NK_EQ                98
+#define TK_RENAME               98
-#define TK_USING                99
+#define TK_TAG                  99
-#define TK_TAGS                 100
+#define TK_SET                  100
-#define TK_COMMENT              101
+#define TK_NK_EQ                101
-#define TK_BOOL                 102
+#define TK_USING                102
-#define TK_TINYINT              103
+#define TK_TAGS                 103
-#define TK_SMALLINT             104
+#define TK_COMMENT              104
-#define TK_INT                  105
+#define TK_BOOL                 105
-#define TK_INTEGER              106
+#define TK_TINYINT              106
-#define TK_BIGINT               107
+#define TK_SMALLINT             107
-#define TK_FLOAT                108
+#define TK_INT                  108
-#define TK_DOUBLE               109
+#define TK_INTEGER              109
-#define TK_BINARY               110
+#define TK_BIGINT               110
-#define TK_TIMESTAMP            111
+#define TK_FLOAT                111
-#define TK_NCHAR                112
+#define TK_DOUBLE               112
-#define TK_UNSIGNED             113
+#define TK_BINARY               113
-#define TK_JSON                 114
+#define TK_TIMESTAMP            114
-#define TK_VARCHAR              115
+#define TK_NCHAR                115
-#define TK_MEDIUMBLOB           116
+#define TK_UNSIGNED             116
-#define TK_BLOB                 117
+#define TK_JSON                 117
-#define TK_VARBINARY            118
+#define TK_VARCHAR              118
-#define TK_DECIMAL              119
+#define TK_MEDIUMBLOB           119
-#define TK_MAX_DELAY            120
+#define TK_BLOB                 120
-#define TK_WATERMARK            121
+#define TK_VARBINARY            121
-#define TK_ROLLUP               122
+#define TK_DECIMAL              122
-#define TK_TTL                  123
+#define TK_MAX_DELAY            123
-#define TK_SMA                  124
+#define TK_WATERMARK            124
-#define TK_FIRST                125
+#define TK_ROLLUP               125
-#define TK_LAST                 126
+#define TK_TTL                  126
-#define TK_SHOW                 127
+#define TK_SMA                  127
-#define TK_DATABASES            128
+#define TK_FIRST                128
-#define TK_TABLES               129
+#define TK_LAST                 129
-#define TK_STABLES              130
+#define TK_SHOW                 130
-#define TK_MNODES               131
+#define TK_DATABASES            131
-#define TK_MODULES              132
+#define TK_TABLES               132
-#define TK_QNODES               133
+#define TK_STABLES              133
-#define TK_FUNCTIONS            134
+#define TK_MNODES               134
-#define TK_INDEXES              135
+#define TK_MODULES              135
-#define TK_ACCOUNTS             136
+#define TK_QNODES               136
-#define TK_APPS                 137
+#define TK_FUNCTIONS            137
-#define TK_CONNECTIONS          138
+#define TK_INDEXES              138
-#define TK_LICENCES             139
+#define TK_ACCOUNTS             139
-#define TK_GRANTS               140
+#define TK_APPS                 140
-#define TK_QUERIES              141
+#define TK_CONNECTIONS          141
-#define TK_SCORES               142
+#define TK_LICENCES             142
-#define TK_TOPICS               143
+#define TK_GRANTS               143
-#define TK_VARIABLES            144
+#define TK_QUERIES              144
-#define TK_BNODES               145
+#define TK_SCORES               145
-#define TK_SNODES               146
+#define TK_TOPICS               146
-#define TK_CLUSTER              147
+#define TK_VARIABLES            147
-#define TK_TRANSACTIONS         148
+#define TK_BNODES               148
-#define TK_DISTRIBUTED          149
+#define TK_SNODES               149
-#define TK_CONSUMERS            150
+#define TK_CLUSTER              150
-#define TK_SUBSCRIPTIONS        151
+#define TK_TRANSACTIONS         151
-#define TK_LIKE                 152
+#define TK_DISTRIBUTED          152
-#define TK_INDEX                153
+#define TK_CONSUMERS            153
-#define TK_FUNCTION             154
+#define TK_SUBSCRIPTIONS        154
-#define TK_INTERVAL             155
+#define TK_VNODES               155
-#define TK_TOPIC                156
+#define TK_LIKE                 156
-#define TK_AS                   157
+#define TK_INDEX                157
-#define TK_WITH                 158
+#define TK_FUNCTION             158
-#define TK_META                 159
+#define TK_INTERVAL             159
-#define TK_CONSUMER             160
+#define TK_TOPIC                160
-#define TK_GROUP                161
+#define TK_AS                   161
-#define TK_DESC                 162
+#define TK_WITH                 162
-#define TK_DESCRIBE             163
+#define TK_META                 163
-#define TK_RESET                164
+#define TK_CONSUMER             164
-#define TK_QUERY                165
+#define TK_GROUP                165
-#define TK_CACHE                166
+#define TK_DESC                 166
-#define TK_EXPLAIN              167
+#define TK_DESCRIBE             167
-#define TK_ANALYZE              168
+#define TK_RESET                168
-#define TK_VERBOSE              169
+#define TK_QUERY                169
-#define TK_NK_BOOL              170
+#define TK_CACHE                170
-#define TK_RATIO                171
+#define TK_EXPLAIN              171
-#define TK_NK_FLOAT             172
+#define TK_ANALYZE              172
-#define TK_OUTPUTTYPE           173
+#define TK_VERBOSE              173
-#define TK_AGGREGATE            174
+#define TK_NK_BOOL              174
-#define TK_BUFSIZE              175
+#define TK_RATIO                175
-#define TK_STREAM               176
+#define TK_NK_FLOAT             176
-#define TK_INTO                 177
+#define TK_OUTPUTTYPE           177
-#define TK_TRIGGER              178
+#define TK_AGGREGATE            178
-#define TK_AT_ONCE              179
+#define TK_BUFSIZE              179
-#define TK_WINDOW_CLOSE         180
+#define TK_STREAM               180
-#define TK_IGNORE               181
+#define TK_INTO                 181
-#define TK_EXPIRED              182
+#define TK_TRIGGER              182
-#define TK_KILL                 183
+#define TK_AT_ONCE              183
-#define TK_CONNECTION           184
+#define TK_WINDOW_CLOSE         184
-#define TK_TRANSACTION          185
+#define TK_IGNORE               185
-#define TK_BALANCE              186
+#define TK_EXPIRED              186
-#define TK_VGROUP               187
+#define TK_KILL                 187
-#define TK_MERGE                188
+#define TK_CONNECTION           188
-#define TK_REDISTRIBUTE         189
+#define TK_TRANSACTION          189
-#define TK_SPLIT                190
+#define TK_BALANCE              190
-#define TK_DELETE               191
+#define TK_VGROUP               191
-#define TK_INSERT               192
+#define TK_MERGE                192
-#define TK_NULL                 193
+#define TK_REDISTRIBUTE         193
-#define TK_NK_QUESTION          194
+#define TK_SPLIT                194
-#define TK_NK_ARROW             195
+#define TK_DELETE               195
-#define TK_ROWTS                196
+#define TK_INSERT               196
-#define TK_TBNAME               197
+#define TK_NULL                 197
-#define TK_QSTART               198
+#define TK_NK_QUESTION          198
-#define TK_QEND                 199
+#define TK_NK_ARROW             199
-#define TK_QDURATION            200
+#define TK_ROWTS                200
-#define TK_WSTART               201
+#define TK_TBNAME               201
-#define TK_WEND                 202
+#define TK_QSTART               202
-#define TK_WDURATION            203
+#define TK_QEND                 203
-#define TK_CAST                 204
+#define TK_QDURATION            204
-#define TK_NOW                  205
+#define TK_WSTART               205
-#define TK_TODAY                206
+#define TK_WEND                 206
-#define TK_TIMEZONE             207
+#define TK_WDURATION            207
-#define TK_CLIENT_VERSION       208
+#define TK_CAST                 208
-#define TK_SERVER_VERSION       209
+#define TK_NOW                  209
-#define TK_SERVER_STATUS        210
+#define TK_TODAY                210
-#define TK_CURRENT_USER         211
+#define TK_TIMEZONE             211
-#define TK_COUNT                212
+#define TK_CLIENT_VERSION       212
-#define TK_LAST_ROW             213
+#define TK_SERVER_VERSION       213
-#define TK_BETWEEN              214
+#define TK_SERVER_STATUS        214
-#define TK_IS                   215
+#define TK_CURRENT_USER         215
-#define TK_NK_LT                216
+#define TK_COUNT                216
-#define TK_NK_GT                217
+#define TK_LAST_ROW             217
-#define TK_NK_LE                218
+#define TK_BETWEEN              218
-#define TK_NK_GE                219
+#define TK_IS                   219
-#define TK_NK_NE                220
+#define TK_NK_LT                220
-#define TK_MATCH                221
+#define TK_NK_GT                221
-#define TK_NMATCH               222
+#define TK_NK_LE                222
-#define TK_CONTAINS             223
+#define TK_NK_GE                223
-#define TK_IN                   224
+#define TK_NK_NE                224
-#define TK_JOIN                 225
+#define TK_MATCH                225
-#define TK_INNER                226
+#define TK_NMATCH               226
-#define TK_SELECT               227
+#define TK_CONTAINS             227
-#define TK_DISTINCT             228
+#define TK_IN                   228
-#define TK_WHERE                229
+#define TK_JOIN                 229
-#define TK_PARTITION            230
+#define TK_INNER                230
-#define TK_BY                   231
+#define TK_SELECT               231
-#define TK_SESSION              232
+#define TK_DISTINCT             232
-#define TK_STATE_WINDOW         233
+#define TK_WHERE                233
-#define TK_SLIDING              234
+#define TK_PARTITION            234
-#define TK_FILL                 235
+#define TK_BY                   235
-#define TK_VALUE                236
+#define TK_SESSION              236
-#define TK_NONE                 237
+#define TK_STATE_WINDOW         237
-#define TK_PREV                 238
+#define TK_SLIDING              238
-#define TK_LINEAR               239
+#define TK_FILL                 239
-#define TK_NEXT                 240
+#define TK_VALUE                240
-#define TK_HAVING               241
+#define TK_NONE                 241
-#define TK_RANGE                242
+#define TK_PREV                 242
-#define TK_EVERY                243
+#define TK_LINEAR               243
-#define TK_ORDER                244
+#define TK_NEXT                 244
-#define TK_SLIMIT               245
+#define TK_HAVING               245
-#define TK_SOFFSET              246
+#define TK_RANGE                246
-#define TK_LIMIT                247
+#define TK_EVERY                247
-#define TK_OFFSET               248
+#define TK_ORDER                248
-#define TK_ASC                  249
+#define TK_SLIMIT               249
-#define TK_NULLS                250
+#define TK_SOFFSET              250
-#define TK_ABORT                251
+#define TK_LIMIT                251
-#define TK_AFTER                252
+#define TK_OFFSET               252
-#define TK_ATTACH               253
+#define TK_ASC                  253
-#define TK_BEFORE               254
+#define TK_NULLS                254
-#define TK_BEGIN                255
+#define TK_ABORT                255
-#define TK_BITAND               256
+#define TK_AFTER                256
-#define TK_BITNOT               257
+#define TK_ATTACH               257
-#define TK_BITOR                258
+#define TK_BEFORE               258
-#define TK_BLOCKS               259
+#define TK_BEGIN                259
-#define TK_CHANGE               260
+#define TK_BITAND               260
-#define TK_COMMA                261
+#define TK_BITNOT               261
-#define TK_COMPACT              262
+#define TK_BITOR                262
-#define TK_CONCAT               263
+#define TK_BLOCKS               263
-#define TK_CONFLICT             264
+#define TK_CHANGE               264
-#define TK_COPY                 265
+#define TK_COMMA                265
-#define TK_DEFERRED             266
+#define TK_COMPACT              266
-#define TK_DELIMITERS           267
+#define TK_CONCAT               267
-#define TK_DETACH               268
+#define TK_CONFLICT             268
-#define TK_DIVIDE               269
+#define TK_COPY                 269
-#define TK_DOT                  270
+#define TK_DEFERRED             270
-#define TK_EACH                 271
+#define TK_DELIMITERS           271
-#define TK_END                  272
+#define TK_DETACH               272
-#define TK_FAIL                 273
+#define TK_DIVIDE               273
-#define TK_FILE                 274
+#define TK_DOT                  274
-#define TK_FOR                  275
+#define TK_EACH                 275
-#define TK_GLOB                 276
+#define TK_END                  276
-#define TK_ID                   277
+#define TK_FAIL                 277
-#define TK_IMMEDIATE            278
+#define TK_FILE                 278
-#define TK_IMPORT               279
+#define TK_FOR                  279
-#define TK_INITIALLY            280
+#define TK_GLOB                 280
-#define TK_INSTEAD              281
+#define TK_ID                   281
-#define TK_ISNULL               282
+#define TK_IMMEDIATE            282
-#define TK_KEY                  283
+#define TK_IMPORT               283
-#define TK_NK_BITNOT            284
+#define TK_INITIALLY            284
-#define TK_NK_SEMI              285
+#define TK_INSTEAD              285
-#define TK_NOTNULL              286
+#define TK_ISNULL               286
-#define TK_OF                   287
+#define TK_KEY                  287
-#define TK_PLUS                 288
+#define TK_NK_BITNOT            288
-#define TK_PRIVILEGE            289
+#define TK_NK_SEMI              289
-#define TK_RAISE                290
+#define TK_NOTNULL              290
-#define TK_REPLACE              291
+#define TK_OF                   291
-#define TK_RESTRICT             292
+#define TK_PLUS                 292
-#define TK_ROW                  293
+#define TK_PRIVILEGE            293
-#define TK_SEMI                 294
+#define TK_RAISE                294
-#define TK_STAR                 295
+#define TK_REPLACE              295
-#define TK_STATEMENT            296
+#define TK_RESTRICT             296
-#define TK_STRING               297
+#define TK_ROW                  297
-#define TK_TIMES                298
+#define TK_SEMI                 298
-#define TK_UPDATE               299
+#define TK_STAR                 299
-#define TK_VALUES               300
+#define TK_STATEMENT            300
-#define TK_VARIABLE             301
+#define TK_STRING               301
-#define TK_VIEW                 302
+#define TK_TIMES                302
-#define TK_VNODES               303
+#define TK_UPDATE               303
-#define TK_WAL                  304
+#define TK_VALUES               304
+#define TK_VARIABLE             305
+#define TK_VIEW                 306
+#define TK_WAL                  307
 
 #define TK_NK_SPACE   300
 #define TK_NK_COMMENT 301

include/libs/nodes/cmdnodes.h

@@ -78,6 +78,12 @@ typedef struct SDatabaseOptions {
   int32_t     walRetentionSize;
   int32_t     walRollPeriod;
   int32_t     walSegmentSize;
+  bool        walRetentionPeriodIsSet;
+  bool        walRetentionSizeIsSet;
+  bool        walRollPeriodIsSet;
+  int32_t     sstTrigger;
+  int32_t     tablePrefix;
+  int32_t     tableSuffix;
 } SDatabaseOptions;
 
 typedef struct SCreateDatabaseStmt {
@@ -268,6 +274,12 @@ typedef struct SShowDnodeVariablesStmt {
   SNode*    pDnodeId;
 } SShowDnodeVariablesStmt;
 
+typedef struct SShowVnodesStmt {
+  ENodeType type;
+  SNode*    pDnodeId;
+  SNode*    pDnodeEndpoint;
+} SShowVnodesStmt;
+
 typedef enum EIndexType { INDEX_TYPE_SMA = 1, INDEX_TYPE_FULLTEXT } EIndexType;
 
 typedef struct SIndexOptions {

include/libs/nodes/nodes.h

@@ -183,12 +183,12 @@ typedef enum ENodeType {
   QUERY_NODE_SHOW_DNODE_VARIABLES_STMT,
   QUERY_NODE_SHOW_TRANSACTIONS_STMT,
   QUERY_NODE_SHOW_SUBSCRIPTIONS_STMT,
+  QUERY_NODE_SHOW_VNODES_STMT,
   QUERY_NODE_SHOW_CREATE_DATABASE_STMT,
   QUERY_NODE_SHOW_CREATE_TABLE_STMT,
   QUERY_NODE_SHOW_CREATE_STABLE_STMT,
   QUERY_NODE_SHOW_TABLE_DISTRIBUTED_STMT,
   QUERY_NODE_SHOW_LOCAL_VARIABLES_STMT,
-  QUERY_NODE_SHOW_VNODES_STMT,
   QUERY_NODE_SHOW_SCORES_STMT,
   QUERY_NODE_KILL_CONNECTION_STMT,
   QUERY_NODE_KILL_QUERY_STMT,
@@ -244,6 +244,7 @@ typedef enum ENodeType {
   QUERY_NODE_PHYSICAL_PLAN_MERGE_STATE,
   QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE,
   QUERY_NODE_PHYSICAL_PLAN_PARTITION,
+  QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION,
   QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC,
   QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC,
   QUERY_NODE_PHYSICAL_PLAN_DISPATCH,

include/libs/nodes/plannodes.h

@@ -488,6 +488,8 @@ typedef struct SPartitionPhysiNode {
   SNodeList* pTargets;
 } SPartitionPhysiNode;
 
+typedef SPartitionPhysiNode SStreamPartitionPhysiNode;
+
 typedef struct SDataSinkNode {
   ENodeType           type;
   SDataBlockDescNode* pInputDataBlockDesc;

include/os/osDir.h

@@ -56,6 +56,7 @@ void    taosRemoveDir(const char *dirname);
 bool    taosDirExist(const char *dirname);
 int32_t taosMkDir(const char *dirname);
 int32_t taosMulMkDir(const char *dirname);
+int32_t taosMulModeMkDir(const char *dirname, int mode);
 void    taosRemoveOldFiles(const char *dirname, int32_t keepDays);
 int32_t taosExpandDir(const char *dirname, char *outname, int32_t maxlen);
 int32_t taosRealPath(char *dirname, char *realPath, int32_t maxlen);

include/util/tdef.h

@@ -359,13 +359,25 @@ typedef enum ELogicConditionType {
 #define TSDB_DB_SCHEMALESS_ON           1
 #define TSDB_DB_SCHEMALESS_OFF          0
 #define TSDB_DEFAULT_DB_SCHEMALESS      TSDB_DB_SCHEMALESS_OFF
+#define TSDB_MIN_SST_TRIGGER            1
+#define TSDB_MAX_SST_TRIGGER            128
+#define TSDB_DEFAULT_SST_TRIGGER        8
+#define TSDB_MIN_HASH_PREFIX            0
+#define TSDB_MAX_HASH_PREFIX            128
+#define TSDB_DEFAULT_HASH_PREFIX        0
+#define TSDB_MIN_HASH_SUFFIX            0
+#define TSDB_MAX_HASH_SUFFIX            128
+#define TSDB_DEFAULT_HASH_SUFFIX        0
 
 #define TSDB_DB_MIN_WAL_RETENTION_PERIOD -1
-#define TSDB_DEFAULT_DB_WAL_RETENTION_PERIOD (24 * 60 * 60 * 4)
+#define TSDB_REP_DEF_DB_WAL_RET_PERIOD   0
+#define TSDB_REPS_DEF_DB_WAL_RET_PERIOD  (24 * 60 * 60 * 4)
 #define TSDB_DB_MIN_WAL_RETENTION_SIZE   -1
-#define TSDB_DEFAULT_DB_WAL_RETENTION_SIZE   -1
+#define TSDB_REP_DEF_DB_WAL_RET_SIZE     0
+#define TSDB_REPS_DEF_DB_WAL_RET_SIZE    -1
 #define TSDB_DB_MIN_WAL_ROLL_PERIOD      0
-#define TSDB_DEFAULT_DB_WAL_ROLL_PERIOD      (24 * 60 * 60 * 1)
+#define TSDB_REP_DEF_DB_WAL_ROLL_PERIOD  0
+#define TSDB_REPS_DEF_DB_WAL_ROLL_PERIOD (24 * 60 * 60 * 1)
 #define TSDB_DB_MIN_WAL_SEGMENT_SIZE     0
 #define TSDB_DEFAULT_DB_WAL_SEGMENT_SIZE 0
 

include/util/tutil.h

@@ -20,6 +20,7 @@
 #include "tcrc32c.h"
 #include "tdef.h"
 #include "tmd5.h"
+#include "thash.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -68,6 +69,8 @@ static FORCE_INLINE void taosEncryptPass_c(uint8_t *inBuf, size_t len, char *tar
   memcpy(target, buf, TSDB_PASSWORD_LEN);
 }
 
+#define taosGetTbHashVal(tbname, tblen, method, prefix, suffix) MurmurHash3_32((tbname), (tblen))
+
 #ifdef __cplusplus
 }
 #endif

packaging/deb/makedeb.sh

@@ -45,6 +45,7 @@ mkdir -p ${pkg_dir}${install_home_path}/include
 mkdir -p ${pkg_dir}${install_home_path}/script
 
 cp ${compile_dir}/../packaging/cfg/taos.cfg         ${pkg_dir}${install_home_path}/cfg
+cp ${compile_dir}/../packaging/cfg/taosd.service    ${pkg_dir}${install_home_path}/cfg
 if [ -f "${compile_dir}/test/cfg/taosadapter.toml" ]; then
     cp ${compile_dir}/test/cfg/taosadapter.toml		${pkg_dir}${install_home_path}/cfg || :
 fi

packaging/tools/install.sh

@@ -33,6 +33,7 @@ adapterName="taosadapter"
 benchmarkName="taosBenchmark"
 dumpName="taosdump"
 demoName="taosdemo"
+xname="taosx"
 
 data_dir=${dataDir}
 log_dir=${logDir}
@@ -199,6 +200,7 @@ function install_bin() {
   ${csudo}rm -f ${bin_link_dir}/${demoName} || :
   ${csudo}rm -f ${bin_link_dir}/${benchmarkName} || :
   ${csudo}rm -f ${bin_link_dir}/${dumpName} || :
+  ${csudo}rm -f ${bin_link_dir}/${xname} || :
   ${csudo}rm -f ${bin_link_dir}/set_core || :
   ${csudo}rm -f ${bin_link_dir}/TDinsight.sh || :
 
@@ -212,6 +214,7 @@ function install_bin() {
   [ -x ${install_main_dir}/bin/${benchmarkName} ] && ${csudo}ln -s ${install_main_dir}/bin/${benchmarkName} ${bin_link_dir}/${demoName} || :
   [ -x ${install_main_dir}/bin/${benchmarkName} ] && ${csudo}ln -s ${install_main_dir}/bin/${benchmarkName} ${bin_link_dir}/${benchmarkName} || :
   [ -x ${install_main_dir}/bin/${dumpName} ] && ${csudo}ln -s ${install_main_dir}/bin/${dumpName} ${bin_link_dir}/${dumpName} || :
+  [ -x ${install_main_dir}/bin/${xname} ] && ${csudo}ln -s ${install_main_dir}/bin/${dumpName} ${bin_link_dir}/${xname} || :
   [ -x ${install_main_dir}/bin/TDinsight.sh ] && ${csudo}ln -s ${install_main_dir}/bin/TDinsight.sh ${bin_link_dir}/TDinsight.sh || :
   [ -x ${install_main_dir}/bin/remove.sh ] && ${csudo}ln -s ${install_main_dir}/bin/remove.sh ${bin_link_dir}/${uninstallScript} || :
   [ -x ${install_main_dir}/bin/set_core.sh ] && ${csudo}ln -s ${install_main_dir}/bin/set_core.sh ${bin_link_dir}/set_core || :

packaging/tools/make_install.sh

@@ -172,6 +172,7 @@ function install_bin() {
   ${csudo}rm -f ${bin_link_dir}/udfd || :
   ${csudo}rm -f ${bin_link_dir}/taosdemo || :
   ${csudo}rm -f ${bin_link_dir}/taosdump || :
+  ${csudo}rm -f ${bin_link_dir}/taosx || :
 
   if [ "$osType" != "Darwin" ]; then
     ${csudo}rm -f ${bin_link_dir}/perfMonitor || :
@@ -184,6 +185,7 @@ function install_bin() {
     [ -f ${binary_dir}/build/bin/taosdump ] && ${csudo}cp -r ${binary_dir}/build/bin/taosdump ${install_main_dir}/bin || :
     [ -f ${binary_dir}/build/bin/taosadapter ] && ${csudo}cp -r ${binary_dir}/build/bin/taosadapter ${install_main_dir}/bin || :
     [ -f ${binary_dir}/build/bin/udfd ] && ${csudo}cp -r ${binary_dir}/build/bin/udfd ${install_main_dir}/bin || :
+    [ -f ${binary_dir}/build/bin/taosx ] && ${csudo}cp -r ${binary_dir}/build/bin/taosx ${install_main_dir}/bin || :
     ${csudo}cp -r ${binary_dir}/build/bin/${serverName} ${install_main_dir}/bin || :
 
     ${csudo}cp -r ${script_dir}/taosd-dump-cfg.gdb ${install_main_dir}/bin || :
@@ -199,6 +201,7 @@ function install_bin() {
     [ -x ${install_main_dir}/bin/udfd ] && ${csudo}ln -s ${install_main_dir}/bin/udfd ${bin_link_dir}/udfd || :
     [ -x ${install_main_dir}/bin/taosdump ] && ${csudo}ln -s ${install_main_dir}/bin/taosdump ${bin_link_dir}/taosdump || :
     [ -x ${install_main_dir}/bin/taosdemo ] && ${csudo}ln -s ${install_main_dir}/bin/taosdemo ${bin_link_dir}/taosdemo || :
+    [ -x ${install_main_dir}/bin/taosx ] && ${csudo}ln -s ${install_main_dir}/bin/taosx ${bin_link_dir}/taosx || :
     [ -x ${install_main_dir}/bin/perfMonitor ] && ${csudo}ln -s ${install_main_dir}/bin/perfMonitor ${bin_link_dir}/perfMonitor || :
     [ -x ${install_main_dir}/set_core.sh ] && ${csudo}ln -s ${install_main_dir}/bin/set_core.sh ${bin_link_dir}/set_core || :
     [ -x ${install_main_dir}/bin/remove.sh ] && ${csudo}ln -s ${install_main_dir}/bin/remove.sh ${bin_link_dir}/${uninstallScript} || :
@@ -215,6 +218,7 @@ function install_bin() {
     [ -x ${install_main_dir}/bin/udfd ] || [ -x ${install_main_2_dir}/bin/udfd ] && ${csudo}ln -s ${install_main_dir}/bin/udfd ${bin_link_dir}/udfd || ${csudo}ln -s ${install_main_2_dir}/bin/udfd || :
     [ -x ${install_main_dir}/bin/taosdump ] || [ -x ${install_main_2_dir}/bin/taosdump ] && ${csudo}ln -s ${install_main_dir}/bin/taosdump ${bin_link_dir}/taosdump || ln -s ${install_main_2_dir}/bin/taosdump ${bin_link_dir}/taosdump || :
     [ -x ${install_main_dir}/bin/taosdemo ] || [ -x ${install_main_2_dir}/bin/taosdemo ] && ${csudo}ln -s ${install_main_dir}/bin/taosdemo ${bin_link_dir}/taosdemo || ln -s ${install_main_2_dir}/bin/taosdemo ${bin_link_dir}/taosdemo || :
+    [ -x ${install_main_dir}/bin/taosx ] || [ -x ${install_main_2_dir}/bin/taosx ] && ${csudo}ln -s ${install_main_dir}/bin/taosx ${bin_link_dir}/taosx || ln -s ${install_main_2_dir}/bin/taosx ${bin_link_dir}/taosx || :
   fi
 }
 

packaging/tools/makepkg.sh

@@ -80,10 +80,12 @@ else
       ${build_dir}/bin/taosBenchmark \
       ${build_dir}/bin/TDinsight.sh \
       $tdinsight_caches"
+  [ -f ${build_dir}/bin/taosx ] && taosx_bin="${build_dir}/bin/taosx"
 
   bin_files="${build_dir}/bin/${serverName} \
       ${build_dir}/bin/${clientName} \
       ${taostools_bin_files} \
+      ${taosx_bin} \
       ${build_dir}/bin/taosadapter \
       ${build_dir}/bin/udfd \
       ${script_dir}/remove.sh \

source/client/src/clientRawBlockWrite.c

@@ -779,7 +779,8 @@ static int32_t taosCreateTable(TAOS* taos, void* meta, int32_t metaLen) {
 
       for (int32_t i = 0; i < taosArrayGetSize(pCreateReq->ctb.tagName); i++) {
         char* tName = taosArrayGet(pCreateReq->ctb.tagName, i);
-        for(int32_t j = pTableMeta->tableInfo.numOfColumns; j < pTableMeta->tableInfo.numOfColumns + pTableMeta->tableInfo.numOfTags; j++){
+        for (int32_t j = pTableMeta->tableInfo.numOfColumns;
+             j < pTableMeta->tableInfo.numOfColumns + pTableMeta->tableInfo.numOfTags; j++) {
           SSchema* tag = &pTableMeta->schema[j];
           if (strcmp(tag->name, tName) == 0 && tag->type != TSDB_DATA_TYPE_JSON) {
             tTagSetCid((STag*)pCreateReq->ctb.pTag, i, tag->colId);
@@ -1323,7 +1324,7 @@ end:
   return code;
 }
 
-static int32_t tmqWriteRaw(TAOS* taos, void* data, int32_t dataLen) {
+static int32_t tmqWriteRawDataImpl(TAOS* taos, void* data, int32_t dataLen) {
   int32_t     code = TSDB_CODE_SUCCESS;
   SHashObj*   pVgHash = NULL;
   SQuery*     pQuery = NULL;
@@ -1669,7 +1670,7 @@ int32_t tmq_write_raw(TAOS* taos, tmq_raw_data raw) {
   } else if (raw.raw_type == TDMT_VND_DELETE) {
     return taosDeleteData(taos, raw.raw, raw.raw_len);
   } else if (raw.raw_type == RES_TYPE__TMQ) {
-    return tmqWriteRaw(taos, raw.raw, raw.raw_len);
+    return tmqWriteRawDataImpl(taos, raw.raw, raw.raw_len);
   }
   return TSDB_CODE_INVALID_PARA;
 }

source/client/src/clientSml.c

@@ -547,6 +547,8 @@ static int32_t smlModifyDBSchemas(SSmlHandle *info) {
         goto end;
       }
       needCheckMeta = true;
+      taosHashCleanup(hashTmp);
+      hashTmp = NULL;
     } else {
       uError("SML:0x%" PRIx64 " load table meta error: %s", info->id, tstrerror(code));
       goto end;
@@ -576,7 +578,6 @@ static int32_t smlModifyDBSchemas(SSmlHandle *info) {
     sTableData->tableMeta = pTableMeta;
 
     tableMetaSml = (SSmlSTableMeta **)taosHashIterate(info->superTables, tableMetaSml);
-    taosHashCleanup(hashTmp);
   }
   return 0;
 

source/common/src/systable.c

@@ -90,7 +90,7 @@ static const SSysDbTableSchema userDBSchema[] = {
     {.name = "minrows", .bytes = 4, .type = TSDB_DATA_TYPE_INT, .sysInfo = true},
     {.name = "maxrows", .bytes = 4, .type = TSDB_DATA_TYPE_INT, .sysInfo = true},
     {.name = "comp", .bytes = 1, .type = TSDB_DATA_TYPE_TINYINT, .sysInfo = true},
-    {.name = "precision", .bytes = 2 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = true},
+    {.name = "precision", .bytes = 2 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = false},
     {.name = "status", .bytes = 10 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = false},
     {.name = "retentions", .bytes = 60 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = true},
     {.name = "single_stable", .bytes = 1, .type = TSDB_DATA_TYPE_BOOL, .sysInfo = true},
@@ -102,6 +102,7 @@ static const SSysDbTableSchema userDBSchema[] = {
     {.name = "wal_retention_size", .bytes = 8, .type = TSDB_DATA_TYPE_BIGINT, .sysInfo = true},
     {.name = "wal_roll_period", .bytes = 4, .type = TSDB_DATA_TYPE_INT, .sysInfo = true},
     {.name = "wal_segment_size", .bytes = 8, .type = TSDB_DATA_TYPE_BIGINT, .sysInfo = true},
+    {.name = "sst_trigger", .bytes = 2, .type = TSDB_DATA_TYPE_SMALLINT, .sysInfo = true},
 };
 
 static const SSysDbTableSchema userFuncSchema[] = {
@@ -256,6 +257,15 @@ static const SSysDbTableSchema subscriptionSchema[] = {
     {.name = "consumer_id", .bytes = 8, .type = TSDB_DATA_TYPE_BIGINT, .sysInfo = false},
 };
 
+static const SSysDbTableSchema vnodesSchema[] = {
+    {.name = "vgroup_id", .bytes = 4, .type = TSDB_DATA_TYPE_INT, .sysInfo = true},
+    {.name = "replica", .bytes = 1, .type = TSDB_DATA_TYPE_TINYINT, .sysInfo = true},
+    {.name = "status", .bytes = 9 + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = true},
+    {.name = "db_name", .bytes = SYSTABLE_SCH_DB_NAME_LEN, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = true},
+    {.name = "dnode_id", .bytes = 4, .type = TSDB_DATA_TYPE_INT, .sysInfo = true},
+    {.name = "dnode_ep", .bytes = TSDB_EP_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR, .sysInfo = true},
+};
+
 static const SSysTableMeta infosMeta[] = {
     {TSDB_INS_TABLE_DNODES, dnodesSchema, tListLen(dnodesSchema), true},
     {TSDB_INS_TABLE_MNODES, mnodesSchema, tListLen(mnodesSchema), true},
@@ -279,6 +289,7 @@ static const SSysTableMeta infosMeta[] = {
     {TSDB_INS_TABLE_TOPICS, topicSchema, tListLen(topicSchema), false},
     {TSDB_INS_TABLE_SUBSCRIPTIONS, subscriptionSchema, tListLen(subscriptionSchema), false},
     {TSDB_INS_TABLE_STREAMS, streamSchema, tListLen(streamSchema), false},
+    {TSDB_INS_TABLE_VNODES, vnodesSchema, tListLen(vnodesSchema), true},
 };
 
 static const SSysDbTableSchema connectionsSchema[] = {

source/common/src/tdatablock.c

@@ -140,7 +140,8 @@ int32_t colDataReserve(SColumnInfoData* pColumnInfoData, size_t newSize) {
   return TSDB_CODE_SUCCESS;
 }
 
-static void doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t currentRow, const char* pData, int32_t itemLen, int32_t numOfRows) {
+static void doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t currentRow, const char* pData,
+                         int32_t itemLen, int32_t numOfRows) {
   ASSERT(pColumnInfoData->info.bytes >= itemLen);
   size_t start = 1;
 
@@ -151,14 +152,16 @@ static void doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t curren
   int32_t count = log(numOfRows) / log(2);
   while (t < count) {
     int32_t xlen = 1 << t;
-    memcpy(pColumnInfoData->pData + start * itemLen + pColumnInfoData->varmeta.length, pColumnInfoData->pData, xlen * itemLen);
+    memcpy(pColumnInfoData->pData + start * itemLen + pColumnInfoData->varmeta.length, pColumnInfoData->pData,
+           xlen * itemLen);
     t += 1;
     start += xlen;
   }
 
   // the tail part
   if (numOfRows > start) {
-    memcpy(pColumnInfoData->pData + start * itemLen + currentRow * itemLen, pColumnInfoData->pData, (numOfRows - start) * itemLen);
+    memcpy(pColumnInfoData->pData + start * itemLen + currentRow * itemLen, pColumnInfoData->pData,
+           (numOfRows - start) * itemLen);
   }
 
   if (IS_VAR_DATA_TYPE(pColumnInfoData->info.type)) {
@@ -170,7 +173,8 @@ static void doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t curren
   }
 }
 
-int32_t colDataAppendNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData, uint32_t numOfRows) {
+int32_t colDataAppendNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData,
+                            uint32_t numOfRows) {
   ASSERT(pData != NULL && pColumnInfoData != NULL);
 
   int32_t len = pColumnInfoData->info.bytes;
@@ -676,7 +680,8 @@ size_t blockDataGetRowSize(SSDataBlock* pBlock) {
  * @return
  */
 size_t blockDataGetSerialMetaSize(uint32_t numOfCols) {
-  // | version | total length | total rows | total columns | flag seg| block group id | column schema | each column length |
+  // | version | total length | total rows | total columns | flag seg| block group id | column schema | each column
+  // length |
   return sizeof(int32_t) + sizeof(int32_t) + sizeof(int32_t) + sizeof(int32_t) + sizeof(int32_t) + sizeof(uint64_t) +
          numOfCols * (sizeof(int8_t) + sizeof(int32_t)) + numOfCols * sizeof(int32_t);
 }
@@ -1302,6 +1307,40 @@ int32_t copyDataBlock(SSDataBlock* dst, const SSDataBlock* src) {
   return TSDB_CODE_SUCCESS;
 }
 
+SSDataBlock* createSpecialDataBlock(EStreamType type) {
+  SSDataBlock* pBlock = taosMemoryCalloc(1, sizeof(SSDataBlock));
+  pBlock->info.hasVarCol = false;
+  pBlock->info.groupId = 0;
+  pBlock->info.rows = 0;
+  pBlock->info.type = type;
+  pBlock->info.rowSize =
+      sizeof(TSKEY) + sizeof(TSKEY) + sizeof(uint64_t) + sizeof(uint64_t) + sizeof(TSKEY) + sizeof(TSKEY);
+  pBlock->info.watermark = INT64_MIN;
+
+  pBlock->pDataBlock = taosArrayInit(6, sizeof(SColumnInfoData));
+  SColumnInfoData infoData = {0};
+  infoData.info.type = TSDB_DATA_TYPE_TIMESTAMP;
+  infoData.info.bytes = sizeof(TSKEY);
+  // window start ts
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+  // window end ts
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+
+  infoData.info.type = TSDB_DATA_TYPE_UBIGINT;
+  infoData.info.bytes = sizeof(uint64_t);
+  // uid
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+  // group id
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+
+  // calculate start ts
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+  // calculate end ts
+  taosArrayPush(pBlock->pDataBlock, &infoData);
+
+  return pBlock;
+}
+
 SSDataBlock* createOneDataBlock(const SSDataBlock* pDataBlock, bool copyData) {
   if (pDataBlock == NULL) {
     return NULL;

source/common/src/tglobal.c

@@ -91,6 +91,7 @@ bool tsSmlDataFormat =
 // query
 int32_t tsQueryPolicy = 1;
 int32_t tsQuerySmaOptimize = 0;
+bool    tsQueryPlannerTrace = false;
 
 /*
  * denote if the server needs to compress response message at the application layer to client, including query rsp,
@@ -286,6 +287,7 @@ static int32_t taosAddClientCfg(SConfig *pCfg) {
   if (cfgAddInt32(pCfg, "compressColData", tsCompressColData, -1, 100000000, 1) != 0) return -1;
   if (cfgAddInt32(pCfg, "queryPolicy", tsQueryPolicy, 1, 3, 1) != 0) return -1;
   if (cfgAddInt32(pCfg, "querySmaOptimize", tsQuerySmaOptimize, 0, 1, 1) != 0) return -1;
+  if (cfgAddBool(pCfg, "queryPlannerTrace", tsQueryPlannerTrace, true) != 0) return -1;
   if (cfgAddString(pCfg, "smlChildTableName", "", 1) != 0) return -1;
   if (cfgAddString(pCfg, "smlTagName", tsSmlTagName, 1) != 0) return -1;
   if (cfgAddBool(pCfg, "smlDataFormat", tsSmlDataFormat, 1) != 0) return -1;
@@ -572,7 +574,6 @@ static int32_t taosUpdateServerCfg(SConfig *pCfg) {
   return 0;
 }
 
-
 static void taosSetClientLogCfg(SConfig *pCfg) {
   SConfigItem *pItem = cfgGetItem(pCfg, "logDir");
   tstrncpy(tsLogDir, cfgGetItem(pCfg, "logDir")->str, PATH_MAX);
@@ -643,6 +644,7 @@ static int32_t taosSetClientCfg(SConfig *pCfg) {
   tsNumOfTaskQueueThreads = cfgGetItem(pCfg, "numOfTaskQueueThreads")->i32;
   tsQueryPolicy = cfgGetItem(pCfg, "queryPolicy")->i32;
   tsQuerySmaOptimize = cfgGetItem(pCfg, "querySmaOptimize")->i32;
+  tsQueryPlannerTrace = cfgGetItem(pCfg, "queryPlannerTrace")->bval;
   return 0;
 }
 
@@ -972,6 +974,8 @@ int32_t taosSetCfg(SConfig *pCfg, char *name) {
         tsQnodeShmSize = cfgGetItem(pCfg, "qnodeShmSize")->i32;
       } else if (strcasecmp("qDebugFlag", name) == 0) {
         qDebugFlag = cfgGetItem(pCfg, "qDebugFlag")->i32;
+      } else if (strcasecmp("queryPlannerTrace", name) == 0) {
+        tsQueryPlannerTrace = cfgGetItem(pCfg, "queryPlannerTrace")->bval;
       }
       break;
     }
@@ -1129,7 +1133,7 @@ int32_t taosCreateLog(const char *logname, int32_t logFileNum, const char *cfgDi
 
   taosSetAllDebugFlag(cfgGetItem(pCfg, "debugFlag")->i32, false);
 
-  if (taosMulMkDir(tsLogDir) != 0) {
+  if (taosMulModeMkDir(tsLogDir, 0777) != 0) {
     uError("failed to create dir:%s since %s", tsLogDir, terrstr());
     cfgCleanup(pCfg);
     return -1;
@@ -1241,16 +1245,14 @@ void taosCfgDynamicOptions(const char *option, const char *value) {
   }
 
   const char *options[] = {
-      "dDebugFlag",   "vDebugFlag",  "mDebugFlag",   "wDebugFlag",   "sDebugFlag",   "tsdbDebugFlag",
+      "dDebugFlag",   "vDebugFlag",   "mDebugFlag",   "wDebugFlag",    "sDebugFlag",   "tsdbDebugFlag", "tqDebugFlag",
-      "tqDebugFlag",  "fsDebugFlag", "udfDebugFlag", "smaDebugFlag", "idxDebugFlag", "tdbDebugFlag",
+      "fsDebugFlag",  "udfDebugFlag", "smaDebugFlag", "idxDebugFlag",  "tdbDebugFlag", "tmrDebugFlag",  "uDebugFlag",
-      "tmrDebugFlag", "uDebugFlag",  "smaDebugFlag", "rpcDebugFlag", "qDebugFlag",   "metaDebugFlag",
+      "smaDebugFlag", "rpcDebugFlag", "qDebugFlag",   "metaDebugFlag", "jniDebugFlag",
-      "jniDebugFlag",
   };
   int32_t *optionVars[] = {
-      &dDebugFlag,   &vDebugFlag,  &mDebugFlag,   &wDebugFlag,   &sDebugFlag,   &tsdbDebugFlag,
+      &dDebugFlag,   &vDebugFlag,   &mDebugFlag,   &wDebugFlag,    &sDebugFlag,   &tsdbDebugFlag, &tqDebugFlag,
-      &tqDebugFlag,  &fsDebugFlag, &udfDebugFlag, &smaDebugFlag, &idxDebugFlag, &tdbDebugFlag,
+      &fsDebugFlag,  &udfDebugFlag, &smaDebugFlag, &idxDebugFlag,  &tdbDebugFlag, &tmrDebugFlag,  &uDebugFlag,
-      &tmrDebugFlag, &uDebugFlag,  &smaDebugFlag, &rpcDebugFlag, &qDebugFlag,   &metaDebugFlag,
+      &smaDebugFlag, &rpcDebugFlag, &qDebugFlag,   &metaDebugFlag, &jniDebugFlag,
-      &jniDebugFlag,
   };
 
   int32_t optionSize = tListLen(options);

source/common/src/tmsg.c

@@ -2024,6 +2024,9 @@ int32_t tSerializeSCreateDbReq(void *buf, int32_t bufLen, SCreateDbReq *pReq) {
   if (tEncodeI64(&encoder, pReq->walRetentionSize) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->walRollPeriod) < 0) return -1;
   if (tEncodeI64(&encoder, pReq->walSegmentSize) < 0) return -1;
+  if (tEncodeI32(&encoder, pReq->sstTrigger) < 0) return -1;
+  if (tEncodeI16(&encoder, pReq->hashPrefix) < 0) return -1;
+  if (tEncodeI16(&encoder, pReq->hashSuffix) < 0) return -1;
   if (tEncodeI8(&encoder, pReq->ignoreExist) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->numOfRetensions) < 0) return -1;
   for (int32_t i = 0; i < pReq->numOfRetensions; ++i) {
@@ -2070,6 +2073,9 @@ int32_t tDeserializeSCreateDbReq(void *buf, int32_t bufLen, SCreateDbReq *pReq)
   if (tDecodeI64(&decoder, &pReq->walRetentionSize) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->walRollPeriod) < 0) return -1;
   if (tDecodeI64(&decoder, &pReq->walSegmentSize) < 0) return -1;
+  if (tDecodeI32(&decoder, &pReq->sstTrigger) < 0) return -1;
+  if (tDecodeI16(&decoder, &pReq->hashPrefix) < 0) return -1;
+  if (tDecodeI16(&decoder, &pReq->hashSuffix) < 0) return -1;
   if (tDecodeI8(&decoder, &pReq->ignoreExist) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->numOfRetensions) < 0) return -1;
   pReq->pRetensions = taosArrayInit(pReq->numOfRetensions, sizeof(SRetention));
@@ -2120,6 +2126,7 @@ int32_t tSerializeSAlterDbReq(void *buf, int32_t bufLen, SAlterDbReq *pReq) {
   if (tEncodeI8(&encoder, pReq->strict) < 0) return -1;
   if (tEncodeI8(&encoder, pReq->cacheLast) < 0) return -1;
   if (tEncodeI8(&encoder, pReq->replications) < 0) return -1;
+  if (tEncodeI32(&encoder, pReq->sstTrigger) < 0) return -1;
   tEndEncode(&encoder);
 
   int32_t tlen = encoder.pos;
@@ -2146,6 +2153,7 @@ int32_t tDeserializeSAlterDbReq(void *buf, int32_t bufLen, SAlterDbReq *pReq) {
   if (tDecodeI8(&decoder, &pReq->strict) < 0) return -1;
   if (tDecodeI8(&decoder, &pReq->cacheLast) < 0) return -1;
   if (tDecodeI8(&decoder, &pReq->replications) < 0) return -1;
+  if (tDecodeI32(&decoder, &pReq->sstTrigger) < 0) return -1;
   tEndDecode(&decoder);
 
   tDecoderClear(&decoder);
@@ -3762,6 +3770,9 @@ int32_t tSerializeSCreateVnodeReq(void *buf, int32_t bufLen, SCreateVnodeReq *pR
   if (tEncodeI64(&encoder, pReq->walRetentionSize) < 0) return -1;
   if (tEncodeI32(&encoder, pReq->walRollPeriod) < 0) return -1;
   if (tEncodeI64(&encoder, pReq->walSegmentSize) < 0) return -1;
+  if (tEncodeI16(&encoder, pReq->sstTrigger) < 0) return -1;
+  if (tEncodeI16(&encoder, pReq->hashPrefix) < 0) return -1;
+  if (tEncodeI16(&encoder, pReq->hashSuffix) < 0) return -1;
 
   tEndEncode(&encoder);
 
@@ -3834,6 +3845,9 @@ int32_t tDeserializeSCreateVnodeReq(void *buf, int32_t bufLen, SCreateVnodeReq *
   if (tDecodeI64(&decoder, &pReq->walRetentionSize) < 0) return -1;
   if (tDecodeI32(&decoder, &pReq->walRollPeriod) < 0) return -1;
   if (tDecodeI64(&decoder, &pReq->walSegmentSize) < 0) return -1;
+  if (tDecodeI16(&decoder, &pReq->sstTrigger) < 0) return -1;
+  if (tDecodeI16(&decoder, &pReq->hashPrefix) < 0) return -1;
+  if (tDecodeI16(&decoder, &pReq->hashSuffix) < 0) return -1;
 
   tEndDecode(&decoder);
   tDecoderClear(&decoder);

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

@@ -167,9 +167,12 @@ static void vmGenerateVnodeCfg(SCreateVnodeReq *pCreate, SVnodeCfg *pCfg) {
   pCfg->walCfg.segSize = pCreate->walSegmentSize;
   pCfg->walCfg.level = pCreate->walLevel;
 
+  pCfg->sstTrigger = pCreate->sstTrigger;
   pCfg->hashBegin = pCreate->hashBegin;
   pCfg->hashEnd = pCreate->hashEnd;
   pCfg->hashMethod = pCreate->hashMethod;
+  pCfg->hashPrefix = pCreate->hashPrefix;
+  pCfg->hashSuffix = pCreate->hashSuffix;
 
   pCfg->standby = pCfg->standby;
   pCfg->syncCfg.myIndex = pCreate->selfIndex;
@@ -219,8 +222,11 @@ int32_t vmProcessCreateVnodeReq(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) {
     return -1;
   }
 
-  dDebug("vgId:%d, start to create vnode, tsma:%d standby:%d cacheLast:%d cacheLastSize:%d", createReq.vgId,
+  dInfo("vgId:%d, start to create vnode, tsma:%d standby:%d cacheLast:%d cacheLastSize:%d sstTrigger:%d",
-         createReq.isTsma, createReq.standby, createReq.cacheLast, createReq.cacheLastSize);
+         createReq.vgId, createReq.isTsma, createReq.standby, createReq.cacheLast, createReq.cacheLastSize,
+         createReq.sstTrigger);
+  dInfo("vgId:%d, hashMethod:%d begin:%u end:%u prefix:%d surfix:%d", createReq.vgId, createReq.hashMethod,
+         createReq.hashBegin, createReq.hashEnd, createReq.hashPrefix, createReq.hashSuffix);
   vmGenerateVnodeCfg(&createReq, &vnodeCfg);
 
   if (vmTsmaAdjustDays(&vnodeCfg, &createReq) < 0) {

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

@@ -305,11 +305,14 @@ typedef struct {
   int8_t  hashMethod;  // default is 1
   int8_t  cacheLast;
   int8_t  schemaless;
+  int16_t hashPrefix;
+  int16_t hashSuffix;
+  int16_t sstTrigger;
   int32_t numOfRetensions;
   SArray* pRetensions;
   int32_t walRetentionPeriod;
-  int64_t walRetentionSize;
   int32_t walRollPeriod;
+  int64_t walRetentionSize;
   int64_t walSegmentSize;
 } SDbCfg;
 

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

@@ -30,7 +30,7 @@
 #include "systable.h"
 
 #define DB_VER_NUMBER   1
-#define DB_RESERVE_SIZE 64
+#define DB_RESERVE_SIZE 58
 
 static SSdbRaw *mndDbActionEncode(SDbObj *pDb);
 static SSdbRow *mndDbActionDecode(SSdbRaw *pRaw);
@@ -124,6 +124,9 @@ static SSdbRaw *mndDbActionEncode(SDbObj *pDb) {
   SDB_SET_INT64(pRaw, dataPos, pDb->cfg.walRetentionSize, _OVER)
   SDB_SET_INT32(pRaw, dataPos, pDb->cfg.walRollPeriod, _OVER)
   SDB_SET_INT64(pRaw, dataPos, pDb->cfg.walSegmentSize, _OVER)
+  SDB_SET_INT16(pRaw, dataPos, pDb->cfg.sstTrigger, _OVER)
+  SDB_SET_INT16(pRaw, dataPos, pDb->cfg.hashPrefix, _OVER)
+  SDB_SET_INT16(pRaw, dataPos, pDb->cfg.hashSuffix, _OVER)
 
   SDB_SET_RESERVE(pRaw, dataPos, DB_RESERVE_SIZE, _OVER)
   SDB_SET_DATALEN(pRaw, dataPos, _OVER)
@@ -207,6 +210,9 @@ static SSdbRow *mndDbActionDecode(SSdbRaw *pRaw) {
   SDB_GET_INT64(pRaw, dataPos, &pDb->cfg.walRetentionSize, _OVER)
   SDB_GET_INT32(pRaw, dataPos, &pDb->cfg.walRollPeriod, _OVER)
   SDB_GET_INT64(pRaw, dataPos, &pDb->cfg.walSegmentSize, _OVER)
+  SDB_GET_INT16(pRaw, dataPos, &pDb->cfg.sstTrigger, _OVER)
+  SDB_GET_INT16(pRaw, dataPos, &pDb->cfg.hashPrefix, _OVER)
+  SDB_GET_INT16(pRaw, dataPos, &pDb->cfg.hashSuffix, _OVER)
 
   SDB_GET_RESERVE(pRaw, dataPos, DB_RESERVE_SIZE, _OVER)
   taosInitRWLatch(&pDb->lock);
@@ -254,6 +260,7 @@ static int32_t mndDbActionUpdate(SSdb *pSdb, SDbObj *pOld, SDbObj *pNew) {
   pOld->cfg.strict = pNew->cfg.strict;
   pOld->cfg.cacheLast = pNew->cfg.cacheLast;
   pOld->cfg.replications = pNew->cfg.replications;
+  pOld->cfg.sstTrigger = pNew->cfg.sstTrigger;
   taosWUnLockLatch(&pOld->lock);
   return 0;
 }
@@ -330,6 +337,9 @@ static int32_t mndCheckDbCfg(SMnode *pMnode, SDbCfg *pCfg) {
   if (pCfg->walRetentionSize < TSDB_DB_MIN_WAL_RETENTION_SIZE) return -1;
   if (pCfg->walRollPeriod < TSDB_DB_MIN_WAL_ROLL_PERIOD) return -1;
   if (pCfg->walSegmentSize < TSDB_DB_MIN_WAL_SEGMENT_SIZE) return -1;
+  if (pCfg->sstTrigger < TSDB_MIN_SST_TRIGGER || pCfg->sstTrigger > TSDB_MAX_SST_TRIGGER) return -1;
+  if (pCfg->hashPrefix < TSDB_MIN_HASH_PREFIX || pCfg->hashPrefix > TSDB_MAX_HASH_PREFIX) return -1;
+  if (pCfg->hashSuffix < TSDB_MIN_HASH_SUFFIX || pCfg->hashSuffix > TSDB_MAX_HASH_SUFFIX) return -1;
 
   terrno = 0;
   return terrno;
@@ -358,11 +368,14 @@ static void mndSetDefaultDbCfg(SDbCfg *pCfg) {
   if (pCfg->numOfRetensions < 0) pCfg->numOfRetensions = 0;
   if (pCfg->schemaless < 0) pCfg->schemaless = TSDB_DB_SCHEMALESS_OFF;
   if (pCfg->walRetentionPeriod < 0 && pCfg->walRetentionPeriod != -1)
-    pCfg->walRetentionPeriod = TSDB_DEFAULT_DB_WAL_RETENTION_PERIOD;
+    pCfg->walRetentionPeriod = TSDB_REPS_DEF_DB_WAL_RET_PERIOD;
   if (pCfg->walRetentionSize < 0 && pCfg->walRetentionSize != -1)
-    pCfg->walRetentionSize = TSDB_DEFAULT_DB_WAL_RETENTION_SIZE;
+    pCfg->walRetentionSize = TSDB_REPS_DEF_DB_WAL_RET_SIZE;
-  if (pCfg->walRollPeriod < 0) pCfg->walRollPeriod = TSDB_DEFAULT_DB_WAL_ROLL_PERIOD;
+  if (pCfg->walRollPeriod < 0) pCfg->walRollPeriod = TSDB_REPS_DEF_DB_WAL_ROLL_PERIOD;
   if (pCfg->walSegmentSize < 0) pCfg->walSegmentSize = TSDB_DEFAULT_DB_WAL_SEGMENT_SIZE;
+  if (pCfg->sstTrigger <= 0) pCfg->sstTrigger = TSDB_DEFAULT_SST_TRIGGER;
+  if (pCfg->hashPrefix < 0) pCfg->hashPrefix = TSDB_DEFAULT_HASH_PREFIX;
+  if (pCfg->hashSuffix < 0) pCfg->hashSuffix = TSDB_DEFAULT_HASH_SUFFIX;
 }
 
 static int32_t mndSetCreateDbRedoLogs(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SVgObj *pVgroups) {
@@ -479,6 +492,9 @@ static int32_t mndCreateDb(SMnode *pMnode, SRpcMsg *pReq, SCreateDbReq *pCreate,
       .walRetentionSize = pCreate->walRetentionSize,
       .walRollPeriod = pCreate->walRollPeriod,
       .walSegmentSize = pCreate->walSegmentSize,
+      .sstTrigger = pCreate->sstTrigger,
+      .hashPrefix = pCreate->hashPrefix,
+      .hashSuffix = pCreate->hashSuffix,
   };
 
   dbObj.cfg.numOfRetensions = pCreate->numOfRetensions;
@@ -1536,6 +1552,24 @@ static void mndDumpDbInfoData(SMnode *pMnode, SSDataBlock *pBlock, SDbObj *pDb,
     STR_WITH_MAXSIZE_TO_VARSTR(buf, "NULL", bytes);
   }
 
+  const char *precStr = NULL;
+  switch (pDb->cfg.precision) {
+    case TSDB_TIME_PRECISION_MILLI:
+      precStr = TSDB_TIME_PRECISION_MILLI_STR;
+      break;
+    case TSDB_TIME_PRECISION_MICRO:
+      precStr = TSDB_TIME_PRECISION_MICRO_STR;
+      break;
+    case TSDB_TIME_PRECISION_NANO:
+      precStr = TSDB_TIME_PRECISION_NANO_STR;
+      break;
+    default:
+      precStr = "none";
+      break;
+  }
+  char precVstr[10] = {0};
+  STR_WITH_SIZE_TO_VARSTR(precVstr, precStr, 2);
+
   char *statusStr = "ready";
   if (objStatus == SDB_STATUS_CREATING) {
     statusStr = "creating";
@@ -1546,7 +1580,6 @@ static void mndDumpDbInfoData(SMnode *pMnode, SSDataBlock *pBlock, SDbObj *pDb,
       statusStr = "unsynced";
     }
   }
-
   char statusVstr[24] = {0};
   STR_WITH_SIZE_TO_VARSTR(statusVstr, statusStr, strlen(statusStr));
 
@@ -1557,6 +1590,8 @@ static void mndDumpDbInfoData(SMnode *pMnode, SSDataBlock *pBlock, SDbObj *pDb,
         colDataAppend(pColInfo, rows, buf, false);
       } else if (i == 3) {
         colDataAppend(pColInfo, rows, (const char *)&numOfTables, false);
+      } else if (i == 14) {
+        colDataAppend(pColInfo, rows, precVstr, false);
       } else if (i == 15) {
         colDataAppend(pColInfo, rows, statusVstr, false);
       } else {
@@ -1621,23 +1656,6 @@ static void mndDumpDbInfoData(SMnode *pMnode, SSDataBlock *pBlock, SDbObj *pDb,
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, rows, (const char *)&pDb->cfg.compression, false);
 
-    const char *precStr = NULL;
-    switch (pDb->cfg.precision) {
-      case TSDB_TIME_PRECISION_MILLI:
-        precStr = TSDB_TIME_PRECISION_MILLI_STR;
-        break;
-      case TSDB_TIME_PRECISION_MICRO:
-        precStr = TSDB_TIME_PRECISION_MICRO_STR;
-        break;
-      case TSDB_TIME_PRECISION_NANO:
-        precStr = TSDB_TIME_PRECISION_NANO_STR;
-        break;
-      default:
-        precStr = "none";
-        break;
-    }
-    char precVstr[10] = {0};
-    STR_WITH_SIZE_TO_VARSTR(precVstr, precStr, 2);
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, rows, (const char *)precVstr, false);
 
@@ -1682,6 +1700,9 @@ static void mndDumpDbInfoData(SMnode *pMnode, SSDataBlock *pBlock, SDbObj *pDb,
 
     pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
     colDataAppend(pColInfo, rows, (const char *)&pDb->cfg.walSegmentSize, false);
+
+    pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+    colDataAppend(pColInfo, rows, (const char *)&pDb->cfg.sstTrigger, false);
   }
 
   taosMemoryFree(buf);

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

@@ -234,6 +234,9 @@ void *mndBuildCreateVnodeReq(SMnode *pMnode, SDnodeObj *pDnode, SDbObj *pDb, SVg
   createReq.walRetentionSize = pDb->cfg.walRetentionSize;
   createReq.walRollPeriod = pDb->cfg.walRollPeriod;
   createReq.walSegmentSize = pDb->cfg.walSegmentSize;
+  createReq.sstTrigger = pDb->cfg.sstTrigger;
+  createReq.hashPrefix = pDb->cfg.hashPrefix;
+  createReq.hashSuffix = pDb->cfg.hashSuffix;
 
   for (int32_t v = 0; v < pVgroup->replica; ++v) {
     SReplica  *pReplica = &createReq.replicas[v];
@@ -792,31 +795,42 @@ static int32_t mndRetrieveVnodes(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pB
 
     for (int32_t i = 0; i < pVgroup->replica && numOfRows < rows; ++i) {
       SVnodeGid       *pVgid = &pVgroup->vnodeGid[i];
+      SColumnInfoData *pColInfo = NULL;
       cols = 0;
 
-      SColumnInfoData *pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+      pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
       colDataAppend(pColInfo, numOfRows, (const char *)&pVgroup->vgId, false);
 
-      SName name = {0};
-      char  db[TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
-      tNameFromString(&name, pVgroup->dbName, T_NAME_ACCT | T_NAME_DB);
-      tNameGetDbName(&name, varDataVal(db));
-      varDataSetLen(db, strlen(varDataVal(db)));
-
       pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-      colDataAppend(pColInfo, numOfRows, (const char *)db, false);
+      colDataAppend(pColInfo, numOfRows, (const char *)&pVgroup->replica, false);
-
-      uint32_t val = 0;
-      pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-      colDataAppend(pColInfo, numOfRows, (const char *)&val, false);
 
       char buf[20] = {0};
       STR_TO_VARSTR(buf, syncStr(pVgid->role));
       pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
       colDataAppend(pColInfo, numOfRows, (const char *)buf, false);
 
+      const char *dbname = mndGetDbStr(pVgroup->dbName);
+      char        b1[TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE] = {0};
+      if (dbname != NULL) {
+        STR_WITH_MAXSIZE_TO_VARSTR(b1, dbname, TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE);
+      } else {
+        STR_WITH_MAXSIZE_TO_VARSTR(b1, "NULL", TSDB_DB_NAME_LEN + VARSTR_HEADER_SIZE);
+      }
       pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
-      colDataAppend(pColInfo, numOfRows, (const char *)&pVgroup->replica, false);  // onlines
+      colDataAppend(pColInfo, numOfRows, (const char *)b1, false);
+
+      pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+      colDataAppend(pColInfo, numOfRows, (const char *)&pVgid->dnodeId, false);
+
+      SDnodeObj *pDnode = mndAcquireDnode(pMnode, pVgid->dnodeId);
+      char       b2[TSDB_EP_LEN + VARSTR_HEADER_SIZE] = {0};
+      if (pDnode != NULL) {
+        STR_WITH_MAXSIZE_TO_VARSTR(b2, pDnode->ep, TSDB_EP_LEN + VARSTR_HEADER_SIZE);
+      } else {
+        STR_WITH_MAXSIZE_TO_VARSTR(b2, "NULL", TSDB_EP_LEN + VARSTR_HEADER_SIZE);
+      }
+      pColInfo = taosArrayGet(pBlock->pDataBlock, cols++);
+      colDataAppend(pColInfo, numOfRows, (const char *)b2, false);
 
       numOfRows++;
     }

source/dnode/vnode/inc/vnode.h

@@ -287,6 +287,9 @@ struct SVnodeCfg {
   SVnodeStats vndStats;
   uint32_t    hashBegin;
   uint32_t    hashEnd;
+  int16_t     sstTrigger;
+  int16_t     hashPrefix;
+  int16_t     hashSuffix;
 };
 
 typedef struct {

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

@@ -173,7 +173,8 @@ int32_t tqProcessPollReq(STQ* pTq, SRpcMsg* pMsg);
 // tq-stream
 int32_t tqProcessTaskDeployReq(STQ* pTq, int64_t version, char* msg, int32_t msgLen);
 int32_t tqProcessTaskDropReq(STQ* pTq, int64_t version, char* msg, int32_t msgLen);
-int32_t tqProcessStreamTrigger(STQ* pTq, SSubmitReq* data, int64_t ver);
+int32_t tqProcessSubmitReq(STQ* pTq, SSubmitReq* data, int64_t ver);
+int32_t tqProcessDelReq(STQ* pTq, void* pReq, int32_t len, int64_t ver);
 int32_t tqProcessTaskRunReq(STQ* pTq, SRpcMsg* pMsg);
 int32_t tqProcessTaskDispatchReq(STQ* pTq, SRpcMsg* pMsg, bool exec);
 int32_t tqProcessTaskRecoverReq(STQ* pTq, SRpcMsg* pMsg);

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

@@ -129,10 +129,16 @@ _err:
 
 bool metaIsTableExist(SMeta *pMeta, tb_uid_t uid) {
   // query uid.idx
+  metaRLock(pMeta);
+
   if (tdbTbGet(pMeta->pUidIdx, &uid, sizeof(uid), NULL, NULL) < 0) {
+    metaULock(pMeta);
+
     return false;
   }
 
+  metaULock(pMeta);
+
   return true;
 }
 
@@ -182,9 +188,14 @@ tb_uid_t metaGetTableEntryUidByName(SMeta *pMeta, const char *name) {
 }
 
 int metaGetTableNameByUid(void *meta, uint64_t uid, char *tbName) {
+  int         code = 0;
   SMetaReader mr = {0};
   metaReaderInit(&mr, (SMeta *)meta, 0);
-  metaGetTableEntryByUid(&mr, uid);
+  code = metaGetTableEntryByUid(&mr, uid);
+  if (code < 0) {
+    metaReaderClear(&mr);
+    return -1;
+  }
 
   STR_TO_VARSTR(tbName, mr.me.name);
   metaReaderClear(&mr);

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

@@ -816,7 +816,86 @@ int32_t tqProcessTaskDeployReq(STQ* pTq, int64_t version, char* msg, int32_t msg
   return streamMetaAddSerializedTask(pTq->pStreamMeta, version, msg, msgLen);
 }
 
-int32_t tqProcessStreamTrigger(STQ* pTq, SSubmitReq* pReq, int64_t ver) {
+int32_t tqProcessDelReq(STQ* pTq, void* pReq, int32_t len, int64_t ver) {
+  bool        failed = false;
+  SDecoder*   pCoder = &(SDecoder){0};
+  SDeleteRes* pRes = &(SDeleteRes){0};
+
+  pRes->uidList = taosArrayInit(0, sizeof(tb_uid_t));
+  if (pRes->uidList == NULL) {
+    terrno = TSDB_CODE_OUT_OF_MEMORY;
+    failed = true;
+  }
+
+  tDecoderInit(pCoder, pReq, len);
+  tDecodeDeleteRes(pCoder, pRes);
+  tDecoderClear(pCoder);
+
+  int32_t sz = taosArrayGetSize(pRes->uidList);
+  if (sz == 0) {
+    taosArrayDestroy(pRes->uidList);
+    return 0;
+  }
+  SSDataBlock* pDelBlock = createSpecialDataBlock(STREAM_DELETE_DATA);
+  blockDataEnsureCapacity(pDelBlock, sz);
+  pDelBlock->info.rows = sz;
+  pDelBlock->info.version = ver;
+
+  for (int32_t i = 0; i < sz; i++) {
+    // start key column
+    SColumnInfoData* pStartCol = taosArrayGet(pDelBlock->pDataBlock, START_TS_COLUMN_INDEX);
+    colDataAppend(pStartCol, i, (const char*)&pRes->skey, false);  // end key column
+    SColumnInfoData* pEndCol = taosArrayGet(pDelBlock->pDataBlock, END_TS_COLUMN_INDEX);
+    colDataAppend(pEndCol, i, (const char*)&pRes->ekey, false);
+    // uid column
+    SColumnInfoData* pUidCol = taosArrayGet(pDelBlock->pDataBlock, UID_COLUMN_INDEX);
+    int64_t*         pUid = taosArrayGet(pRes->uidList, i);
+    colDataAppend(pUidCol, i, (const char*)pUid, false);
+
+    colDataAppendNULL(taosArrayGet(pDelBlock->pDataBlock, GROUPID_COLUMN_INDEX), i);
+    colDataAppendNULL(taosArrayGet(pDelBlock->pDataBlock, CALCULATE_START_TS_COLUMN_INDEX), i);
+    colDataAppendNULL(taosArrayGet(pDelBlock->pDataBlock, CALCULATE_END_TS_COLUMN_INDEX), i);
+  }
+
+  taosArrayDestroy(pRes->uidList);
+
+  void* pIter = NULL;
+  while (1) {
+    pIter = taosHashIterate(pTq->pStreamMeta->pTasks, pIter);
+    if (pIter == NULL) break;
+    SStreamTask* pTask = *(SStreamTask**)pIter;
+    if (pTask->taskLevel != TASK_LEVEL__SOURCE) continue;
+
+    qDebug("delete req enqueue stream task: %d, ver: %" PRId64, pTask->taskId, ver);
+
+    SStreamDataBlock* pStreamBlock = taosAllocateQitem(sizeof(SStreamDataBlock), DEF_QITEM);
+    pStreamBlock->type = STREAM_INPUT__DATA_BLOCK;
+    pStreamBlock->blocks = taosArrayInit(0, sizeof(SSDataBlock));
+    SSDataBlock block = {0};
+    assignOneDataBlock(&block, pDelBlock);
+    block.info.type = STREAM_DELETE_DATA;
+    taosArrayPush(pStreamBlock->blocks, &block);
+
+    if (!failed) {
+      if (streamTaskInput(pTask, (SStreamQueueItem*)pStreamBlock) < 0) {
+        qError("stream task input del failed, task id %d", pTask->taskId);
+        continue;
+      }
+
+      if (streamSchedExec(pTask) < 0) {
+        qError("stream task launch failed, task id %d", pTask->taskId);
+        continue;
+      }
+    } else {
+      streamTaskInputFail(pTask);
+    }
+  }
+  blockDataDestroy(pDelBlock);
+
+  return 0;
+}
+
+int32_t tqProcessSubmitReq(STQ* pTq, SSubmitReq* pReq, int64_t ver) {
   void*              pIter = NULL;
   bool               failed = false;
   SStreamDataSubmit* pSubmit = NULL;

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

@@ -241,6 +241,20 @@ int32_t tqTaosxScanLog(STQ* pTq, STqHandle* pHandle, SSubmitReq* pReq, STaosxRsp
           continue;
         }
       }
+      if (pHandle->fetchMeta) {
+        SSubmitBlk* pBlk = pReader->pBlock;
+        if (pBlk->schemaLen > 0) {
+          if (pRsp->createTableNum == 0) {
+            pRsp->createTableLen = taosArrayInit(0, sizeof(int32_t));
+            pRsp->createTableReq = taosArrayInit(0, sizeof(void*));
+          }
+          void* createReq = taosMemoryCalloc(1, pBlk->schemaLen);
+          memcpy(createReq, pBlk->data, pBlk->schemaLen);
+          taosArrayPush(pRsp->createTableLen, &pBlk->schemaLen);
+          taosArrayPush(pRsp->createTableReq, &createReq);
+          pRsp->createTableNum++;
+        }
+      }
       tqAddBlockDataToRsp(&block, (SMqDataRsp*)pRsp, taosArrayGetSize(block.pDataBlock));
       blockDataFreeRes(&block);
       tqAddBlockSchemaToRsp(pExec, (SMqDataRsp*)pRsp);
@@ -261,34 +275,25 @@ int32_t tqTaosxScanLog(STQ* pTq, STqHandle* pHandle, SSubmitReq* pReq, STaosxRsp
           continue;
         }
       }
+      if (pHandle->fetchMeta) {
+        SSubmitBlk* pBlk = pReader->pBlock;
+        if (pBlk->schemaLen > 0) {
+          if (pRsp->createTableNum == 0) {
+            pRsp->createTableLen = taosArrayInit(0, sizeof(int32_t));
+            pRsp->createTableReq = taosArrayInit(0, sizeof(void*));
+          }
+          void* createReq = taosMemoryCalloc(1, pBlk->schemaLen);
+          memcpy(createReq, pBlk->data, pBlk->schemaLen);
+          taosArrayPush(pRsp->createTableLen, &pBlk->schemaLen);
+          taosArrayPush(pRsp->createTableReq, &createReq);
+          pRsp->createTableNum++;
+        }
+      }
       tqAddBlockDataToRsp(&block, (SMqDataRsp*)pRsp, taosArrayGetSize(block.pDataBlock));
       blockDataFreeRes(&block);
       tqAddBlockSchemaToRsp(pExec, (SMqDataRsp*)pRsp);
       pRsp->blockNum++;
     }
-#if 1
-    if (pHandle->fetchMeta && pRsp->blockNum) {
-      SSubmitMsgIter iter = {0};
-      tInitSubmitMsgIter(pReq, &iter);
-      STaosxRsp* pXrsp = (STaosxRsp*)pRsp;
-      while (1) {
-        SSubmitBlk* pBlk = NULL;
-        if (tGetSubmitMsgNext(&iter, &pBlk) < 0) break;
-        if (pBlk == NULL) break;
-        if (pBlk->schemaLen > 0) {
-          if (pXrsp->createTableNum == 0) {
-            pXrsp->createTableLen = taosArrayInit(0, sizeof(int32_t));
-            pXrsp->createTableReq = taosArrayInit(0, sizeof(void*));
-          }
-          void* createReq = taosMemoryCalloc(1, pBlk->schemaLen);
-          memcpy(createReq, pBlk->data, pBlk->schemaLen);
-          taosArrayPush(pXrsp->createTableLen, &pBlk->schemaLen);
-          taosArrayPush(pXrsp->createTableReq, &createReq);
-          pXrsp->createTableNum++;
-        }
-      }
-    }
-#endif
   }
 
   if (pRsp->blockNum == 0) {

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

@@ -213,7 +213,8 @@ int32_t tqPushMsgNew(STQ* pTq, void* msg, int32_t msgLen, tmsg_t msgType, int64_
 #endif
 
 int tqPushMsg(STQ* pTq, void* msg, int32_t msgLen, tmsg_t msgType, int64_t ver) {
-  if (vnodeIsRoleLeader(pTq->pVnode) && msgType == TDMT_VND_SUBMIT) {
+  if (vnodeIsRoleLeader(pTq->pVnode)) {
+    if (msgType == TDMT_VND_SUBMIT) {
       if (taosHashGetSize(pTq->pStreamMeta->pTasks) == 0) return 0;
 
       void* data = taosMemoryMalloc(msgLen);
@@ -226,7 +227,11 @@ int tqPushMsg(STQ* pTq, void* msg, int32_t msgLen, tmsg_t msgType, int64_t ver)
       SSubmitReq* pReq = (SSubmitReq*)data;
       pReq->version = ver;
 
-    tqProcessStreamTrigger(pTq, data, ver);
+      tqProcessSubmitReq(pTq, data, ver);
+    }
+    if (msgType == TDMT_VND_DELETE) {
+      tqProcessDelReq(pTq, POINTER_SHIFT(msg, sizeof(SMsgHead)), msgLen - sizeof(SMsgHead), ver);
+    }
   }
 
   return 0;

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

@@ -184,9 +184,9 @@ static int32_t  doAppendRowFromFileBlock(SSDataBlock* pResBlock, STsdbReader* pR
 static void     setComposedBlockFlag(STsdbReader* pReader, bool composed);
 static bool     hasBeenDropped(const SArray* pDelList, int32_t* index, TSDBKEY* pKey, int32_t order);
 
-static void doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SArray* pDelList, STSRow** pTSRow,
+static int32_t doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SArray* pDelList, STSRow** pTSRow,
                              STsdbReader* pReader, bool* freeTSRow);
-static void doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* pBlockScanInfo, STsdbReader* pReader,
+static int32_t doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* pBlockScanInfo, STsdbReader* pReader,
                                STSRow** pTSRow);
 static int32_t mergeRowsInFileBlocks(SBlockData* pBlockData, STableBlockScanInfo* pBlockScanInfo, int64_t key, STsdbReader* pReader);
 
@@ -1395,7 +1395,11 @@ static int32_t doMergeBufAndFileRows(STsdbReader* pReader, STableBlockScanInfo*
     }
   }
 
-  tRowMergerGetRow(&merge, &pTSRow);
+  int32_t code = tRowMergerGetRow(&merge, &pTSRow);
+  if (code != TSDB_CODE_SUCCESS) {
+    return code;
+  }
+
   doAppendRowFromTSRow(pReader->pResBlock, pReader, pTSRow, pBlockScanInfo->uid);
 
   taosMemoryFree(pTSRow);
@@ -1406,6 +1410,7 @@ static int32_t doMergeBufAndFileRows(STsdbReader* pReader, STableBlockScanInfo*
 static int32_t doMergeFileBlockAndLastBlock(SLastBlockReader* pLastBlockReader, STsdbReader* pReader,
                                             STableBlockScanInfo* pBlockScanInfo, SBlockData* pBlockData,
                                             bool mergeBlockData) {
+  SFileBlockDumpInfo* pDumpInfo = &pReader->status.fBlockDumpInfo;
   SBlockData* pLastBlockData = &pLastBlockReader->lastBlockData;
   int64_t     tsLastBlock = getCurrentKeyInLastBlock(pLastBlockReader);
 
@@ -1418,11 +1423,15 @@ static int32_t doMergeFileBlockAndLastBlock(SLastBlockReader* pLastBlockReader,
   doMergeRowsInLastBlock(pLastBlockReader, pBlockScanInfo, tsLastBlock, &merge);
 
   // merge with block data if ts == key
-  if (mergeBlockData) {
+  if (mergeBlockData && (tsLastBlock == pBlockData->aTSKEY[pDumpInfo->rowIndex])) {
     doMergeRowsInFileBlocks(pBlockData, pBlockScanInfo, pReader, &merge);
   }
 
-  tRowMergerGetRow(&merge, &pTSRow);
+  int32_t code = tRowMergerGetRow(&merge, &pTSRow);
+  if (code != TSDB_CODE_SUCCESS) {
+    return code;
+  }
+
   doAppendRowFromTSRow(pReader->pResBlock, pReader, pTSRow, pBlockScanInfo->uid);
 
   taosMemoryFree(pTSRow);
@@ -1456,12 +1465,16 @@ static int32_t mergeFileBlockAndLastBlock(STsdbReader* pReader, SLastBlockReader
         doMergeRowsInFileBlocks(pBlockData, pBlockScanInfo, pReader, &merge);
         doMergeRowsInLastBlock(pLastBlockReader, pBlockScanInfo, ts, &merge);
 
-        tRowMergerGetRow(&merge, &pTSRow);
+        int32_t code = tRowMergerGetRow(&merge, &pTSRow);
+        if (code != TSDB_CODE_SUCCESS) {
+          return code;
+        }
+
         doAppendRowFromTSRow(pReader->pResBlock, pReader, pTSRow, pBlockScanInfo->uid);
 
         taosMemoryFree(pTSRow);
         tRowMergerClear(&merge);
-        return TSDB_CODE_SUCCESS;
+        return code;
       } else {
         ASSERT(0);
         return TSDB_CODE_SUCCESS;
@@ -1618,12 +1631,16 @@ static int32_t doMergeMultiLevelRows(STsdbReader* pReader, STableBlockScanInfo*
     }
   }
 
-  tRowMergerGetRow(&merge, &pTSRow);
+  int32_t code = tRowMergerGetRow(&merge, &pTSRow);
+  if (code != TSDB_CODE_SUCCESS) {
+    return code;
+  }
+
   doAppendRowFromTSRow(pReader->pResBlock, pReader, pTSRow, pBlockScanInfo->uid);
 
   taosMemoryFree(pTSRow);
   tRowMergerClear(&merge);
-  return TSDB_CODE_SUCCESS;
+  return code;
 }
 
 #if 0
@@ -1907,7 +1924,11 @@ int32_t mergeRowsInFileBlocks(SBlockData* pBlockData, STableBlockScanInfo* pBloc
 
     tRowMergerInit(&merge, &fRow, pReader->pSchema);
     doMergeRowsInFileBlocks(pBlockData, pBlockScanInfo, pReader, &merge);
-    tRowMergerGetRow(&merge, &pTSRow);
+    int32_t code = tRowMergerGetRow(&merge, &pTSRow);
+    if (code != TSDB_CODE_SUCCESS) {
+      return code;
+    }
+
     doAppendRowFromTSRow(pReader->pResBlock, pReader, pTSRow, pBlockScanInfo->uid);
 
     taosMemoryFree(pTSRow);
@@ -3026,7 +3047,7 @@ int32_t doMergeRowsInLastBlock(SLastBlockReader* pLastBlockReader, STableBlockSc
   return TSDB_CODE_SUCCESS;
 }
 
-void doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SArray* pDelList, STSRow** pTSRow,
+int32_t doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SArray* pDelList, STSRow** pTSRow,
                                  STsdbReader* pReader, bool* freeTSRow) {
   TSDBROW* pNextRow = NULL;
   TSDBROW  current = *pRow;
@@ -3037,19 +3058,19 @@ void doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SAr
     if (!pIter->hasVal) {
       *pTSRow = current.pTSRow;
       *freeTSRow = false;
-      return;
+      return TSDB_CODE_SUCCESS;
     } else {  // has next point in mem/imem
       pNextRow = getValidMemRow(pIter, pDelList, pReader);
       if (pNextRow == NULL) {
         *pTSRow = current.pTSRow;
         *freeTSRow = false;
-        return;
+        return TSDB_CODE_SUCCESS;
       }
 
       if (current.pTSRow->ts != pNextRow->pTSRow->ts) {
         *pTSRow = current.pTSRow;
         *freeTSRow = false;
-        return;
+        return TSDB_CODE_SUCCESS;
       }
     }
   }
@@ -3069,13 +3090,17 @@ void doMergeMemTableMultiRows(TSDBROW* pRow, uint64_t uid, SIterInfo* pIter, SAr
   tRowMergerAdd(&merge, pNextRow, pTSchema1);
 
   doMergeRowsInBuf(pIter, uid, current.pTSRow->ts, pDelList, &merge, pReader);
-  tRowMergerGetRow(&merge, pTSRow);
+  int32_t code = tRowMergerGetRow(&merge, pTSRow);
-  tRowMergerClear(&merge);
+  if (code != TSDB_CODE_SUCCESS) {
-
+    return code;
-  *freeTSRow = true;
   }
 
-void doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* pBlockScanInfo, STsdbReader* pReader,
+  tRowMergerClear(&merge);
+  *freeTSRow = true;
+  return TSDB_CODE_SUCCESS;
+}
+
+int32_t doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* pBlockScanInfo, STsdbReader* pReader,
                         STSRow** pTSRow) {
   SRowMerger merge = {0};
 
@@ -3100,7 +3125,8 @@ void doMergeMemIMemRows(TSDBROW* pRow, TSDBROW* piRow, STableBlockScanInfo* pBlo
     doMergeRowsInBuf(&pBlockScanInfo->iiter, pBlockScanInfo->uid, k.ts, pBlockScanInfo->delSkyline, &merge, pReader);
   }
 
-  tRowMergerGetRow(&merge, pTSRow);
+  int32_t code = tRowMergerGetRow(&merge, pTSRow);
+  return code;
 }
 
 int32_t tsdbGetNextRowInMem(STableBlockScanInfo* pBlockScanInfo, STsdbReader* pReader, STSRow** pTSRow, int64_t endKey,
@@ -3130,28 +3156,30 @@ int32_t tsdbGetNextRowInMem(STableBlockScanInfo* pBlockScanInfo, STsdbReader* pR
     TSDBKEY k = TSDBROW_KEY(pRow);
     TSDBKEY ik = TSDBROW_KEY(piRow);
 
+    int32_t code = TSDB_CODE_SUCCESS;
     if (ik.ts != k.ts) {
       if (((ik.ts < k.ts) && asc) || ((ik.ts > k.ts) && (!asc))) {  // ik.ts < k.ts
-        doMergeMemTableMultiRows(piRow, uid, &pBlockScanInfo->iiter, pDelList, pTSRow, pReader, freeTSRow);
+        code = doMergeMemTableMultiRows(piRow, uid, &pBlockScanInfo->iiter, pDelList, pTSRow, pReader, freeTSRow);
       } else if (((k.ts < ik.ts) && asc) || ((k.ts > ik.ts) && (!asc))) {
-        doMergeMemTableMultiRows(pRow, uid, &pBlockScanInfo->iter, pDelList, pTSRow, pReader, freeTSRow);
+        code = doMergeMemTableMultiRows(pRow, uid, &pBlockScanInfo->iter, pDelList, pTSRow, pReader, freeTSRow);
       }
     } else {  // ik.ts == k.ts
-      doMergeMemIMemRows(pRow, piRow, pBlockScanInfo, pReader, pTSRow);
       *freeTSRow = true;
+      code = doMergeMemIMemRows(pRow, piRow, pBlockScanInfo, pReader, pTSRow);
+      if (code != TSDB_CODE_SUCCESS) {
+        return code;
+      }
     }
 
-    return TSDB_CODE_SUCCESS;
+    return code;
   }
 
   if (pBlockScanInfo->iter.hasVal && pRow != NULL) {
-    doMergeMemTableMultiRows(pRow, pBlockScanInfo->uid, &pBlockScanInfo->iter, pDelList, pTSRow, pReader, freeTSRow);
+    return doMergeMemTableMultiRows(pRow, pBlockScanInfo->uid, &pBlockScanInfo->iter, pDelList, pTSRow, pReader, freeTSRow);
-    return TSDB_CODE_SUCCESS;
   }
 
   if (pBlockScanInfo->iiter.hasVal && piRow != NULL) {
-    doMergeMemTableMultiRows(piRow, uid, &pBlockScanInfo->iiter, pDelList, pTSRow, pReader, freeTSRow);
+    return doMergeMemTableMultiRows(piRow, uid, &pBlockScanInfo->iiter, pDelList, pTSRow, pReader, freeTSRow);
-    return TSDB_CODE_SUCCESS;
   }
 
   return TSDB_CODE_SUCCESS;

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

@@ -2039,7 +2039,7 @@ int32_t tsdbCmprColData(SColData *pColData, int8_t cmprAlg, SBlockCol *pBlockCol
   size += pBlockCol->szOffset;
 
   // value
-  if (pColData->flag != (HAS_NULL | HAS_NONE)) {
+  if ((pColData->flag != (HAS_NULL | HAS_NONE)) && pColData->nData) {
     code = tsdbCmprData((uint8_t *)pColData->pData, pColData->nData, pColData->type, cmprAlg, ppOut, nOut + size,
                         &pBlockCol->szValue, ppBuf);
     if (code) goto _exit;

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

@@ -14,6 +14,7 @@
  */
 
 #include "vnd.h"
+#include "tutil.h"
 
 const SVnodeCfg vnodeCfgDefault = {.vgId = -1,
                                    .dbname = "",
@@ -106,9 +107,12 @@ int vnodeEncodeConfig(const void *pObj, SJson *pJson) {
   if (tjsonAddIntegerToObject(pJson, "wal.retentionSize", pCfg->walCfg.retentionSize) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "wal.segSize", pCfg->walCfg.segSize) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "wal.level", pCfg->walCfg.level) < 0) return -1;
+  if (tjsonAddIntegerToObject(pJson, "sstTrigger", pCfg->sstTrigger) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "hashBegin", pCfg->hashBegin) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "hashEnd", pCfg->hashEnd) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "hashMethod", pCfg->hashMethod) < 0) return -1;
+  if (tjsonAddIntegerToObject(pJson, "hashPrefix", pCfg->hashPrefix) < 0) return -1;
+  if (tjsonAddIntegerToObject(pJson, "hashSuffix", pCfg->hashSuffix) < 0) return -1;
 
   if (tjsonAddIntegerToObject(pJson, "syncCfg.replicaNum", pCfg->syncCfg.replicaNum) < 0) return -1;
   if (tjsonAddIntegerToObject(pJson, "syncCfg.myIndex", pCfg->syncCfg.myIndex) < 0) return -1;
@@ -205,12 +209,18 @@ int vnodeDecodeConfig(const SJson *pJson, void *pObj) {
   if (code < 0) return -1;
   tjsonGetNumberValue(pJson, "wal.level", pCfg->walCfg.level, code);
   if (code < 0) return -1;
+  tjsonGetNumberValue(pJson, "sstTrigger", pCfg->sstTrigger, code);
+  if (code < 0) return -1;
   tjsonGetNumberValue(pJson, "hashBegin", pCfg->hashBegin, code);
   if (code < 0) return -1;
   tjsonGetNumberValue(pJson, "hashEnd", pCfg->hashEnd, code);
   if (code < 0) return -1;
   tjsonGetNumberValue(pJson, "hashMethod", pCfg->hashMethod, code);
   if (code < 0) return -1;
+  tjsonGetNumberValue(pJson, "hashPrefix", pCfg->hashPrefix, code);
+  if (code < 0) return -1;
+  tjsonGetNumberValue(pJson, "hashSuffix", pCfg->hashSuffix, code);
+  if (code < 0) return -1;
 
   tjsonGetNumberValue(pJson, "syncCfg.replicaNum", pCfg->syncCfg.replicaNum, code);
   if (code < 0) return -1;
@@ -247,7 +257,8 @@ int vnodeValidateTableHash(SVnode *pVnode, char *tableFName) {
 
   switch (pVnode->config.hashMethod) {
     default:
-      hashValue = MurmurHash3_32(tableFName, strlen(tableFName));
+      hashValue = taosGetTbHashVal(tableFName, strlen(tableFName), pVnode->config.hashMethod, pVnode->config.hashPrefix,
+                                   pVnode->config.hashSuffix);
       break;
   }
 

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

@@ -467,9 +467,13 @@ static int32_t vnodeGetStbColumnNum(SVnode *pVnode, tb_uid_t suid, int *num) {
   STSchema *pTSchema = metaGetTbTSchema(pVnode->pMeta, suid, -1);
   // metaGetTbTSchemaEx(pVnode->pMeta, suid, suid, -1, &pTSchema);
 
+  if (pTSchema) {
     *num = pTSchema->numOfCols;
 
     taosMemoryFree(pTSchema);
+  } else {
+    *num = 2;
+  }
 
   return TSDB_CODE_SUCCESS;
 }

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

@@ -496,6 +496,7 @@ static int32_t vnodeProcessCreateTbReq(SVnode *pVnode, int64_t version, void *pR
   // loop to create table
   for (int32_t iReq = 0; iReq < req.nReqs; iReq++) {
     pCreateReq = req.pReqs + iReq;
+    memset(&cRsp, 0, sizeof(cRsp));
 
     if ((terrno = grantCheck(TSDB_GRANT_TIMESERIES)) < 0) {
       rcode = -1;
@@ -1106,6 +1107,7 @@ static int32_t vnodeProcessDeleteReq(SVnode *pVnode, int64_t version, void *pReq
 
   tDecoderInit(pCoder, pReq, len);
   tDecodeDeleteRes(pCoder, pRes);
+  ASSERT(taosArrayGetSize(pRes->uidList) == 0 || (pRes->skey != 0 && pRes->ekey != 0));
 
   for (int32_t iUid = 0; iUid < taosArrayGetSize(pRes->uidList); iUid++) {
     code = tsdbDeleteTableData(pVnode->pTsdb, version, pRes->suid, *(uint64_t *)taosArrayGet(pRes->uidList, iUid),

source/libs/executor/inc/executorimpl.h

@@ -38,11 +38,11 @@ extern "C" {
 #include "tlockfree.h"
 #include "tmsg.h"
 #include "tpagedbuf.h"
-#include "tstreamUpdate.h"
 #include "tstream.h"
+#include "tstreamUpdate.h"
 
-#include "vnode.h"
 #include "executorInt.h"
+#include "vnode.h"
 
 typedef int32_t (*__block_search_fn_t)(char* data, int32_t num, int64_t key, int32_t order);
 
@@ -310,7 +310,8 @@ typedef struct SAggSupporter {
 } SAggSupporter;
 
 typedef struct {
-    // if the upstream is an interval operator, the interval info is also kept here to get the time window to check if current data block needs to be loaded.
+  // if the upstream is an interval operator, the interval info is also kept here to get the time window to check if
+  // current data block needs to be loaded.
   SInterval      interval;
   SAggSupporter* pAggSup;
   SExprSupp*     pExprSup;  // expr supporter of aggregate operator
@@ -410,6 +411,7 @@ typedef enum EStreamScanMode {
   STREAM_SCAN_FROM_READERHANDLE = 1,
   STREAM_SCAN_FROM_RES,
   STREAM_SCAN_FROM_UPDATERES,
+  STREAM_SCAN_FROM_DELETERES,
   STREAM_SCAN_FROM_DATAREADER_RETRIEVE,
   STREAM_SCAN_FROM_DATAREADER_RANGE,
 } EStreamScanMode;
@@ -438,12 +440,24 @@ typedef struct SStreamAggSupporter {
   SSDataBlock*   pScanBlock;
 } SStreamAggSupporter;
 
-typedef struct SessionWindowSupporter {
+typedef struct SWindowSupporter {
   SStreamAggSupporter* pStreamAggSup;
   int64_t              gap;
   uint16_t             parentType;
   SAggSupporter*       pIntervalAggSup;
-} SessionWindowSupporter;
+} SWindowSupporter;
+
+typedef struct SPartitionBySupporter {
+  SArray* pGroupCols;     // group by columns, SArray<SColumn>
+  SArray* pGroupColVals;  // current group column values, SArray<SGroupKeys>
+  char*   keyBuf;         // group by keys for hash
+  bool    needCalc;       // partition by column
+} SPartitionBySupporter;
+
+typedef struct SPartitionDataInfo {
+  uint64_t groupId;
+  SArray*  rowIds;
+} SPartitionDataInfo;
 
 typedef struct STimeWindowSupp {
   int8_t          calTrigger;
@@ -478,7 +492,9 @@ typedef struct SStreamScanInfo {
   SOperatorInfo*        pStreamScanOp;
   SOperatorInfo*        pTableScanOp;
   SArray*               childIds;
-  SessionWindowSupporter sessionSup;
+  SWindowSupporter      windowSup;
+  SPartitionBySupporter partitionSup;
+  SExprSupp*            pPartScalarSup;
   bool                  assignBlockUid;  // assign block uid to groupId, temporarily used for generating rollup SMA.
   int32_t               scanWinIndex;    // for state operator
   int32_t               pullDataResIndex;
@@ -691,7 +707,6 @@ typedef struct SPartitionOperatorInfo {
   SArray*        sortedGroupArray;  // SDataGroupInfo sorted by group id
   int32_t        groupIndex;        // group index
   int32_t        pageIndex;         // page index of current group
-  SSDataBlock*   pUpdateRes;
   SExprSupp      scalarSup;
 } SPartitionOperatorInfo;
 
@@ -743,8 +758,8 @@ typedef struct SStreamSessionAggOperatorInfo {
   SSDataBlock*        pWinBlock;   // window result
   SqlFunctionCtx*     pDummyCtx;   // for combine
   SSDataBlock*        pDelRes;     // delete result
-  bool                 returnDelete;
   SSDataBlock*        pUpdateRes;  // update window
+  bool                returnUpdate;
   SHashObj*           pStDeleted;
   void*               pDelIterator;
   SArray*             pChildren;  // cache for children's result; final stream operator
@@ -753,6 +768,16 @@ typedef struct SStreamSessionAggOperatorInfo {
   bool                ignoreExpiredData;
 } SStreamSessionAggOperatorInfo;
 
+typedef struct SStreamPartitionOperatorInfo {
+  SOptrBasicInfo        binfo;
+  SPartitionBySupporter partitionSup;
+  SExprSupp             scalarSup;
+  SHashObj*             pPartitions;
+  void*                 parIte;
+  SSDataBlock*          pInputDataBlock;
+  int32_t               tsColIndex;
+} SStreamPartitionOperatorInfo;
+
 typedef struct STimeSliceOperatorInfo {
   SSDataBlock*         pRes;
   STimeWindow          win;
@@ -875,7 +900,8 @@ void    destroyExprInfo(SExprInfo* pExpr, int32_t numOfExprs);
 int32_t initAggInfo(SExprSupp* pSup, SAggSupporter* pAggSup, SExprInfo* pExprInfo, int32_t numOfCols, size_t keyBufSize,
                     const char* pkey);
 void    initResultSizeInfo(SResultInfo* pResultInfo, int32_t numOfRows);
-void    doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SGroupResInfo* pGroupResInfo, SDiskbasedBuf* pBuf);
+void    doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SGroupResInfo* pGroupResInfo,
+                               SDiskbasedBuf* pBuf);
 int32_t handleLimitOffset(SOperatorInfo* pOperator, SLimitInfo* pLimitInfo, SSDataBlock* pBlock, bool holdDataInBuf);
 bool    hasLimitOffsetInfo(SLimitInfo* pLimitInfo);
 void    initLimitInfo(const SNode* pLimit, const SNode* pSLimit, SLimitInfo* pLimitInfo);
@@ -883,7 +909,8 @@ void    initLimitInfo(const SNode* pLimit, const SNode* pSLimit, SLimitInfo* pLi
 void doApplyFunctions(SExecTaskInfo* taskInfo, SqlFunctionCtx* pCtx, SColumnInfoData* pTimeWindowData, int32_t offset,
                       int32_t forwardStep, int32_t numOfTotal, int32_t numOfOutput);
 
-int32_t extractDataBlockFromFetchRsp(SSDataBlock* pRes, char* pData, int32_t numOfOutput, SArray* pColList, char** pNextStart);
+int32_t extractDataBlockFromFetchRsp(SSDataBlock* pRes, char* pData, int32_t numOfOutput, SArray* pColList,
+                                     char** pNextStart);
 void    updateLoadRemoteInfo(SLoadRemoteDataInfo* pInfo, int32_t numOfRows, int32_t dataLen, int64_t startTs,
                              SOperatorInfo* pOperator);
 
@@ -909,29 +936,37 @@ SSDataBlock* loadNextDataBlock(void* param);
 
 void setResultRowInitCtx(SResultRow* pResult, SqlFunctionCtx* pCtx, int32_t numOfOutput, int32_t* rowEntryInfoOffset);
 
-SResultRow* doSetResultOutBufByKey(SDiskbasedBuf* pResultBuf, SResultRowInfo* pResultRowInfo,
+SResultRow* doSetResultOutBufByKey(SDiskbasedBuf* pResultBuf, SResultRowInfo* pResultRowInfo, char* pData,
-                                   char* pData, int16_t bytes, bool masterscan, uint64_t groupId,
+                                   int16_t bytes, bool masterscan, uint64_t groupId, SExecTaskInfo* pTaskInfo,
-                                   SExecTaskInfo* pTaskInfo, bool isIntervalQuery, SAggSupporter* pSup);
+                                   bool isIntervalQuery, SAggSupporter* pSup);
 
 SOperatorInfo* createExchangeOperatorInfo(void* pTransporter, SExchangePhysiNode* pExNode, SExecTaskInfo* pTaskInfo);
 
-SOperatorInfo* createTableScanOperatorInfo(STableScanPhysiNode* pTableScanNode, SReadHandle* pHandle, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createTableScanOperatorInfo(STableScanPhysiNode* pTableScanNode, SReadHandle* pHandle,
+                                           SExecTaskInfo* pTaskInfo);
 SOperatorInfo* createTagScanOperatorInfo(SReadHandle* pReadHandle, STagScanPhysiNode* pPhyNode,
                                          STableListInfo* pTableListInfo, SExecTaskInfo* pTaskInfo);
-SOperatorInfo* createSysTableScanOperatorInfo(void* readHandle, SSystemTableScanPhysiNode *pScanPhyNode, const char* pUser, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createSysTableScanOperatorInfo(void* readHandle, SSystemTableScanPhysiNode* pScanPhyNode,
+                                              const char* pUser, SExecTaskInfo* pTaskInfo);
 
-SOperatorInfo* createAggregateOperatorInfo(SOperatorInfo* downstream, SExprInfo* pExprInfo, int32_t numOfCols, SSDataBlock* pResultBlock, SNode* pCondition, SExprInfo* pScalarExprInfo,
+SOperatorInfo* createAggregateOperatorInfo(SOperatorInfo* downstream, SExprInfo* pExprInfo, int32_t numOfCols,
+                                           SSDataBlock* pResultBlock, SNode* pCondition, SExprInfo* pScalarExprInfo,
                                            int32_t numOfScalarExpr, bool mergeResult, SExecTaskInfo* pTaskInfo);
 
-SOperatorInfo* createIndefinitOutputOperatorInfo(SOperatorInfo* downstream, SPhysiNode *pNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createIndefinitOutputOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pNode,
-SOperatorInfo* createProjectOperatorInfo(SOperatorInfo* downstream, SProjectPhysiNode* pProjPhyNode, SExecTaskInfo* pTaskInfo);
+                                                 SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createProjectOperatorInfo(SOperatorInfo* downstream, SProjectPhysiNode* pProjPhyNode,
+                                         SExecTaskInfo* pTaskInfo);
 SOperatorInfo* createSortOperatorInfo(SOperatorInfo* downstream, SSortPhysiNode* pSortNode, SExecTaskInfo* pTaskInfo);
-SOperatorInfo* createMultiwayMergeOperatorInfo(SOperatorInfo** dowStreams, size_t numStreams, SMergePhysiNode* pMergePhysiNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createMultiwayMergeOperatorInfo(SOperatorInfo** dowStreams, size_t numStreams,
-SOperatorInfo* createCacherowsScanOperator(SLastRowScanPhysiNode* pTableScanNode, SReadHandle* readHandle, SExecTaskInfo* pTaskInfo);
+                                               SMergePhysiNode* pMergePhysiNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createCacherowsScanOperator(SLastRowScanPhysiNode* pTableScanNode, SReadHandle* readHandle,
+                                           SExecTaskInfo* pTaskInfo);
 
 SOperatorInfo* createIntervalOperatorInfo(SOperatorInfo* downstream, SExprInfo* pExprInfo, int32_t numOfCols,
                                           SSDataBlock* pResBlock, SInterval* pInterval, int32_t primaryTsSlotId,
-                                          STimeWindowAggSupp* pTwAggSupp, SIntervalPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo, bool isStream);
+                                          STimeWindowAggSupp* pTwAggSupp, SIntervalPhysiNode* pPhyNode,
+                                          SExecTaskInfo* pTaskInfo, bool isStream);
 SOperatorInfo* createMergeIntervalOperatorInfo(SOperatorInfo* downstream, SMergeIntervalPhysiNode* pIntervalPhyNode,
                                                SExecTaskInfo* pTaskInfo);
 SOperatorInfo* createMergeAlignedIntervalOperatorInfo(SOperatorInfo* downstream, SMergeAlignedIntervalPhysiNode* pNode,
@@ -951,24 +986,33 @@ SOperatorInfo* createStreamScanOperatorInfo(SReadHandle* pHandle, STableScanPhys
 
 SOperatorInfo* createRawScanOperatorInfo(SReadHandle* pHandle, SExecTaskInfo* pTaskInfo);
 
-SOperatorInfo* createFillOperatorInfo(SOperatorInfo* downstream, SFillPhysiNode* pPhyFillNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createFillOperatorInfo(SOperatorInfo* downstream, SFillPhysiNode* pPhyFillNode,
-SOperatorInfo* createStatewindowOperatorInfo(SOperatorInfo* downstream, SStateWinodwPhysiNode* pStateNode, SExecTaskInfo* pTaskInfo);
+                                      SExecTaskInfo* pTaskInfo);
-SOperatorInfo* createPartitionOperatorInfo(SOperatorInfo* downstream, SPartitionPhysiNode* pPartNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createStatewindowOperatorInfo(SOperatorInfo* downstream, SStateWinodwPhysiNode* pStateNode,
-SOperatorInfo* createTimeSliceOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pNode, SExecTaskInfo* pTaskInfo);
+                                             SExecTaskInfo* pTaskInfo);
-SOperatorInfo* createMergeJoinOperatorInfo(SOperatorInfo** pDownstream, int32_t numOfDownstream, SSortMergeJoinPhysiNode* pJoinNode,
+SOperatorInfo* createPartitionOperatorInfo(SOperatorInfo* downstream, SPartitionPhysiNode* pPartNode,
                                            SExecTaskInfo* pTaskInfo);
 
-SOperatorInfo* createStreamSessionAggOperatorInfo(SOperatorInfo* downstream,
+SOperatorInfo* createStreamPartitionOperatorInfo(SOperatorInfo* downstream, SPartitionPhysiNode* pPartNode,
-    SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo);
+                                                 SExecTaskInfo* pTaskInfo);
-SOperatorInfo* createStreamFinalSessionAggOperatorInfo(SOperatorInfo* downstream,
-    SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo, int32_t numOfChild);
 
-SOperatorInfo* createStreamStateAggOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createTimeSliceOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pNode, SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createMergeJoinOperatorInfo(SOperatorInfo** pDownstream, int32_t numOfDownstream,
+                                           SSortMergeJoinPhysiNode* pJoinNode, SExecTaskInfo* pTaskInfo);
+
+SOperatorInfo* createStreamSessionAggOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pPhyNode,
+                                                  SExecTaskInfo* pTaskInfo);
+SOperatorInfo* createStreamFinalSessionAggOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pPhyNode,
+                                                       SExecTaskInfo* pTaskInfo, int32_t numOfChild);
+
+SOperatorInfo* createStreamStateAggOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pPhyNode,
+                                                SExecTaskInfo* pTaskInfo);
 
 int32_t projectApplyFunctions(SExprInfo* pExpr, SSDataBlock* pResult, SSDataBlock* pSrcBlock, SqlFunctionCtx* pCtx,
                               int32_t numOfOutput, SArray* pPseudoList);
 
-void setInputDataBlock(SOperatorInfo* pOperator, SqlFunctionCtx* pCtx, SSDataBlock* pBlock, int32_t order, int32_t scanFlag, bool createDummyCol);
+void setInputDataBlock(SOperatorInfo* pOperator, SqlFunctionCtx* pCtx, SSDataBlock* pBlock, int32_t order,
+                       int32_t scanFlag, bool createDummyCol);
 
 bool    isTaskKilled(SExecTaskInfo* pTaskInfo);
 int32_t checkForQueryBuf(size_t numOfTables);
@@ -1013,24 +1057,27 @@ int32_t binarySearchForKey(char* pValue, int num, TSKEY key, int order);
 int32_t initStreamAggSupporter(SStreamAggSupporter* pSup, const char* pKey, SqlFunctionCtx* pCtx, int32_t numOfOutput,
                                int32_t size);
 SResultRow*        getNewResultRow(SDiskbasedBuf* pResultBuf, int32_t* currentPageId, int32_t interBufSize);
-SResultWindowInfo* getSessionTimeWindow(SStreamAggSupporter* pAggSup, TSKEY startTs,
+SResultWindowInfo* getSessionTimeWindow(SStreamAggSupporter* pAggSup, TSKEY startTs, TSKEY endTs, uint64_t groupId,
-    TSKEY endTs, uint64_t groupId, int64_t gap, int32_t* pIndex);
+                                        int64_t gap, int32_t* pIndex);
-SResultWindowInfo* getCurSessionWindow(SStreamAggSupporter* pAggSup, TSKEY startTs,
+SResultWindowInfo* getCurSessionWindow(SStreamAggSupporter* pAggSup, TSKEY startTs, TSKEY endTs, uint64_t groupId,
-    TSKEY endTs, uint64_t groupId, int64_t gap, int32_t* pIndex);
+                                       int64_t gap, int32_t* pIndex);
 bool               isInTimeWindow(STimeWindow* pWin, TSKEY ts, int64_t gap);
 bool               functionNeedToExecute(SqlFunctionCtx* pCtx);
 bool               isOverdue(TSKEY ts, STimeWindowAggSupp* pSup);
 bool               isCloseWindow(STimeWindow* pWin, STimeWindowAggSupp* pSup);
 bool               isDeletedWindow(STimeWindow* pWin, uint64_t groupId, SAggSupporter* pSup);
-void appendOneRow(SSDataBlock* pBlock, TSKEY* pStartTs, TSKEY* pEndTs, int32_t uidCol, uint64_t* pID);
+void               appendOneRow(SSDataBlock* pBlock, TSKEY* pStartTs, TSKEY* pEndTs, uint64_t* pUid, uint64_t* pGp);
 void               printDataBlock(SSDataBlock* pBlock, const char* flag);
+uint64_t calGroupIdByData(SPartitionBySupporter* pParSup, SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t rowId);
 
 int32_t finalizeResultRowIntoResultDataBlock(SDiskbasedBuf* pBuf, SResultRowPosition* resultRowPosition,
-                                       SqlFunctionCtx* pCtx, SExprInfo* pExprInfo, int32_t numOfExprs, const int32_t* rowCellOffset,
+                                             SqlFunctionCtx* pCtx, SExprInfo* pExprInfo, int32_t numOfExprs,
-                                       SSDataBlock* pBlock, SExecTaskInfo* pTaskInfo);
+                                             const int32_t* rowCellOffset, SSDataBlock* pBlock,
+                                             SExecTaskInfo* pTaskInfo);
 
 int32_t createScanTableListInfo(SScanPhysiNode* pScanNode, SNodeList* pGroupTags, bool groupSort, SReadHandle* pHandle,
-                                STableListInfo* pTableListInfo, SNode* pTagCond, SNode* pTagIndexCond, const char* idstr);
+                                STableListInfo* pTableListInfo, SNode* pTagCond, SNode* pTagIndexCond,
+                                const char* idstr);
 
 SOperatorInfo* createGroupSortOperatorInfo(SOperatorInfo* downstream, SGroupSortPhysiNode* pSortPhyNode,
                                            SExecTaskInfo* pTaskInfo);
@@ -1041,7 +1088,6 @@ void copyUpdateDataBlock(SSDataBlock* pDest, SSDataBlock* pSource, int32_t tsCol
 
 bool    groupbyTbname(SNodeList* pGroupList);
 int32_t generateGroupIdMap(STableListInfo* pTableListInfo, SReadHandle* pHandle, SNodeList* groupKey);
-SSDataBlock* createSpecialDataBlock(EStreamType type);
 void*   destroySqlFunctionCtx(SqlFunctionCtx* pCtx, int32_t numOfOutput);
 
 #ifdef __cplusplus

source/libs/executor/src/executil.c

@@ -938,6 +938,7 @@ SArray* extractColMatchInfo(SNodeList* pNodeList, SDataBlockDescNode* pOutputNod
 
   for (int32_t i = 0; i < numOfCols; ++i) {
     STargetNode* pNode = (STargetNode*)nodesListGetNode(pNodeList, i);
+    if (nodeType(pNode->pExpr) == QUERY_NODE_COLUMN) {
       SColumnNode* pColNode = (SColumnNode*)pNode->pExpr;
 
       SColMatchInfo c = {0};
@@ -948,6 +949,7 @@ SArray* extractColMatchInfo(SNodeList* pNodeList, SDataBlockDescNode* pOutputNod
       c.targetSlotId = pNode->slotId;
       taosArrayPush(pList, &c);
     }
+  }
 
   *numOfOutputCols = 0;
   int32_t num = LIST_LENGTH(pOutputNodeList->pSlots);

source/libs/executor/src/executor.c

@@ -52,7 +52,11 @@ static int32_t doSetStreamBlock(SOperatorInfo* pOperator, void* input, size_t nu
     // TODO: if a block was set but not consumed,
     // prevent setting a different type of block
     pInfo->validBlockIndex = 0;
+    if (pInfo->blockType == STREAM_INPUT__DATA_BLOCK) {
+      taosArrayClearP(pInfo->pBlockLists, taosMemoryFree);
+    } else {
       taosArrayClear(pInfo->pBlockLists);
+    }
 
     if (type == STREAM_INPUT__MERGED_SUBMIT) {
       // ASSERT(numOfBlocks > 1);
@@ -810,7 +814,7 @@ int32_t qStreamPrepareScan(qTaskInfo_t tinfo, STqOffsetVal* pOffset, int8_t subT
         }
       }
 
-      // TODO after dropping table, table may be not found
+      // TODO after dropping table, table may not found
       ASSERT(found);
 
       if (pTableScanInfo->dataReader == NULL) {

source/libs/executor/src/executorimpl.c

@@ -263,8 +263,6 @@ SResultRow* doSetResultOutBufByKey(SDiskbasedBuf* pResultBuf, SResultRowInfo* pR
     ASSERT(pSup->resultRowSize > 0);
     pResult = getNewResultRow(pResultBuf, &pSup->currentPageId, pSup->resultRowSize);
 
-    initResultRow(pResult);
-
     // add a new result set for a new group
     SResultRowPosition pos = {.pageId = pResult->pageId, .offset = pResult->offset};
     tSimpleHashPut(pSup->pResultRowHashTable, pSup->keyBuf, GET_RES_WINDOW_KEY_LEN(bytes), &pos,
@@ -817,13 +815,6 @@ void setBlockSMAInfo(SqlFunctionCtx* pCtx, SExprInfo* pExprInfo, SSDataBlock* pB
   } else {
     pInput->colDataAggIsSet = false;
   }
-
-  // set the statistics data for primary time stamp column
-  //  if (pCtx->functionId == FUNCTION_SPREAD && pColumn->colId == PRIMARYKEY_TIMESTAMP_COL_ID) {
-  //    pCtx->isAggSet = true;
-  //    pCtx->agg.min = pBlock->info.window.skey;
-  //    pCtx->agg.max = pBlock->info.window.ekey;
-  //  }
 }
 
 bool isTaskKilled(SExecTaskInfo* pTaskInfo) {
@@ -860,146 +851,6 @@ STimeWindow getAlignQueryTimeWindow(SInterval* pInterval, int32_t precision, int
   return win;
 }
 
-#if 0
-static int32_t updateBlockLoadStatus(STaskAttr* pQuery, int32_t status) {
-
-  bool hasFirstLastFunc = false;
-  bool hasOtherFunc = false;
-
-  if (status == BLK_DATA_DATA_LOAD || status == BLK_DATA_FILTEROUT) {
-    return status;
-  }
-
-  for (int32_t i = 0; i < pQuery->numOfOutput; ++i) {
-    int32_t functionId = getExprFunctionId(&pQuery->pExpr1[i]);
-
-    if (functionId == FUNCTION_TS || functionId == FUNCTION_TS_DUMMY || functionId == FUNCTION_TAG ||
-        functionId == FUNCTION_TAG_DUMMY) {
-      continue;
-    }
-
-    if (functionId == FUNCTION_FIRST_DST || functionId == FUNCTION_LAST_DST) {
-      hasFirstLastFunc = true;
-    } else {
-      hasOtherFunc = true;
-    }
-
-  }
-
-  if (hasFirstLastFunc && status == BLK_DATA_NOT_LOAD) {
-    if (!hasOtherFunc) {
-      return BLK_DATA_FILTEROUT;
-    } else {
-      return BLK_DATA_DATA_LOAD;
-    }
-  }
-
-  return status;
-}
-
-#endif
-
-// static void updateDataCheckOrder(SQInfo *pQInfo, SQueryTableReq* pQueryMsg, bool stableQuery) {
-//   STaskAttr* pQueryAttr = pQInfo->runtimeEnv.pQueryAttr;
-//
-//   // in case of point-interpolation query, use asc order scan
-//   char msg[] = "QInfo:0x%"PRIx64" scan order changed for %s query, old:%d, new:%d, qrange exchanged, old qrange:%"
-//   PRId64
-//                "-%" PRId64 ", new qrange:%" PRId64 "-%" PRId64;
-//
-//   // todo handle the case the the order irrelevant query type mixed up with order critical query type
-//   // descending order query for last_row query
-//   if (isFirstLastRowQuery(pQueryAttr)) {
-//     //qDebug("QInfo:0x%"PRIx64" scan order changed for last_row query, old:%d, new:%d", pQInfo->qId,
-//     pQueryAttr->order.order, TSDB_ORDER_ASC);
-//
-//     pQueryAttr->order.order = TSDB_ORDER_ASC;
-//     if (pQueryAttr->window.skey > pQueryAttr->window.ekey) {
-//       TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//     }
-//
-//     pQueryAttr->needReverseScan = false;
-//     return;
-//   }
-//
-//   if (pQueryAttr->groupbyColumn && pQueryAttr->order.order == TSDB_ORDER_DESC) {
-//     pQueryAttr->order.order = TSDB_ORDER_ASC;
-//     if (pQueryAttr->window.skey > pQueryAttr->window.ekey) {
-//       TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//     }
-//
-//     pQueryAttr->needReverseScan = false;
-//     doUpdateLastKey(pQueryAttr);
-//     return;
-//   }
-//
-//   if (pQueryAttr->pointInterpQuery && pQueryAttr->interval.interval == 0) {
-//     if (!QUERY_IS_ASC_QUERY(pQueryAttr)) {
-//       //qDebug(msg, pQInfo->qId, "interp", pQueryAttr->order.order, TSDB_ORDER_ASC, pQueryAttr->window.skey,
-//       pQueryAttr->window.ekey, pQueryAttr->window.ekey, pQueryAttr->window.skey); TSWAP(pQueryAttr->window.skey,
-//       pQueryAttr->window.ekey, TSKEY);
-//     }
-//
-//     pQueryAttr->order.order = TSDB_ORDER_ASC;
-//     return;
-//   }
-//
-//   if (pQueryAttr->interval.interval == 0) {
-//     if (onlyFirstQuery(pQueryAttr)) {
-//       if (!QUERY_IS_ASC_QUERY(pQueryAttr)) {
-//         //qDebug(msg, pQInfo->qId, "only-first", pQueryAttr->order.order, TSDB_ORDER_ASC, pQueryAttr->window.skey,
-////               pQueryAttr->window.ekey, pQueryAttr->window.ekey, pQueryAttr->window.skey);
-//
-//        TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//        doUpdateLastKey(pQueryAttr);
-//      }
-//
-//      pQueryAttr->order.order = TSDB_ORDER_ASC;
-//      pQueryAttr->needReverseScan = false;
-//    } else if (onlyLastQuery(pQueryAttr) && notContainSessionOrStateWindow(pQueryAttr)) {
-//      if (QUERY_IS_ASC_QUERY(pQueryAttr)) {
-//        //qDebug(msg, pQInfo->qId, "only-last", pQueryAttr->order.order, TSDB_ORDER_DESC, pQueryAttr->window.skey,
-////               pQueryAttr->window.ekey, pQueryAttr->window.ekey, pQueryAttr->window.skey);
-//
-//        TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//        doUpdateLastKey(pQueryAttr);
-//      }
-//
-//      pQueryAttr->order.order = TSDB_ORDER_DESC;
-//      pQueryAttr->needReverseScan = false;
-//    }
-//
-//  } else {  // interval query
-//    if (stableQuery) {
-//      if (onlyFirstQuery(pQueryAttr)) {
-//        if (!QUERY_IS_ASC_QUERY(pQueryAttr)) {
-//          //qDebug(msg, pQInfo->qId, "only-first stable", pQueryAttr->order.order, TSDB_ORDER_ASC,
-////                 pQueryAttr->window.skey, pQueryAttr->window.ekey, pQueryAttr->window.ekey,
-/// pQueryAttr->window.skey);
-//
-//          TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//          doUpdateLastKey(pQueryAttr);
-//        }
-//
-//        pQueryAttr->order.order = TSDB_ORDER_ASC;
-//        pQueryAttr->needReverseScan = false;
-//      } else if (onlyLastQuery(pQueryAttr)) {
-//        if (QUERY_IS_ASC_QUERY(pQueryAttr)) {
-//          //qDebug(msg, pQInfo->qId, "only-last stable", pQueryAttr->order.order, TSDB_ORDER_DESC,
-////                 pQueryAttr->window.skey, pQueryAttr->window.ekey, pQueryAttr->window.ekey,
-/// pQueryAttr->window.skey);
-//
-//          TSWAP(pQueryAttr->window.skey, pQueryAttr->window.ekey);
-//          doUpdateLastKey(pQueryAttr);
-//        }
-//
-//        pQueryAttr->order.order = TSDB_ORDER_DESC;
-//        pQueryAttr->needReverseScan = false;
-//      }
-//    }
-//  }
-//}
-
 #if 0
 static bool overlapWithTimeWindow(STaskAttr* pQueryAttr, SDataBlockInfo* pBlockInfo) {
   STimeWindow w = {0};
@@ -1225,24 +1076,6 @@ static void updateTableQueryInfoForReverseScan(STableQueryInfo* pTableQueryInfo)
   if (pTableQueryInfo == NULL) {
     return;
   }
-
-  //  TSWAP(pTableQueryInfo->win.skey, pTableQueryInfo->win.ekey);
-  //  pTableQueryInfo->lastKey = pTableQueryInfo->win.skey;
-
-  //  SWITCH_ORDER(pTableQueryInfo->cur.order);
-  //  pTableQueryInfo->cur.vgroupIndex = -1;
-
-  // set the index to be the end slot of result rows array
-  //  SResultRowInfo* pResultRowInfo = &pTableQueryInfo->resInfo;
-  //  if (pResultRowInfo->size > 0) {
-  //    pResultRowInfo->curPos = pResultRowInfo->size - 1;
-  //  } else {
-  //    pResultRowInfo->curPos = -1;
-  //  }
-}
-
-void initResultRow(SResultRow* pResultRow) {
-  //  pResultRow->pEntryInfo = (struct SResultRowEntryInfo*)((char*)pResultRow + sizeof(SResultRow));
 }
 
 void setTaskStatus(SExecTaskInfo* pTaskInfo, int8_t status) {
@@ -1255,15 +1088,6 @@ void setTaskStatus(SExecTaskInfo* pTaskInfo, int8_t status) {
   }
 }
 
-void destroyTableQueryInfoImpl(STableQueryInfo* pTableQueryInfo) {
-  if (pTableQueryInfo == NULL) {
-    return;
-  }
-
-  //  taosVariantDestroy(&pTableQueryInfo->tag);
-  //  cleanupResultRowInfo(&pTableQueryInfo->resInfo);
-}
-
 void setResultRowInitCtx(SResultRow* pResult, SqlFunctionCtx* pCtx, int32_t numOfOutput, int32_t* rowEntryInfoOffset) {
   bool init = false;
   for (int32_t i = 0; i < numOfOutput; ++i) {
@@ -3022,7 +2846,6 @@ int32_t aggDecodeResultRow(SOperatorInfo* pOperator, char* result) {
     resultRow->offset = pOffset;
     offset += valueLen;
 
-    initResultRow(resultRow);
     pInfo->resultRowInfo.cur = (SResultRowPosition){.pageId = resultRow->pageId, .offset = resultRow->offset};
     // releaseBufPage(pSup->pResultBuf, getBufPage(pSup->pResultBuf, pageId));
   }
@@ -4149,6 +3972,8 @@ SOperatorInfo* createOperatorTree(SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo
     pOptr = createStreamFinalSessionAggOperatorInfo(ops[0], pPhyNode, pTaskInfo, children);
   } else if (QUERY_NODE_PHYSICAL_PLAN_PARTITION == type) {
     pOptr = createPartitionOperatorInfo(ops[0], (SPartitionPhysiNode*)pPhyNode, pTaskInfo);
+  } else if (QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION == type) {
+    pOptr = createStreamPartitionOperatorInfo(ops[0], (SPartitionPhysiNode*)pPhyNode, pTaskInfo);
   } else if (QUERY_NODE_PHYSICAL_PLAN_MERGE_STATE == type) {
     SStateWinodwPhysiNode* pStateNode = (SStateWinodwPhysiNode*)pPhyNode;
     pOptr = createStatewindowOperatorInfo(ops[0], pStateNode, pTaskInfo);

source/libs/executor/src/groupoperator.c

@@ -830,3 +830,205 @@ int32_t setGroupResultOutputBuf(SOperatorInfo* pOperator, SOptrBasicInfo* binfo,
   setResultRowInitCtx(pResultRow, pCtx, numOfCols, pOperator->exprSupp.rowEntryInfoOffset);
   return TSDB_CODE_SUCCESS;
 }
+
+uint64_t calGroupIdByData(SPartitionBySupporter* pParSup, SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t rowId) {
+  if (pExprSup->pExprInfo != NULL) {
+    int32_t code = projectApplyFunctions(pExprSup->pExprInfo, pBlock, pBlock, pExprSup->pCtx, pExprSup->numOfExprs, NULL);
+    if (code != TSDB_CODE_SUCCESS) {
+      qError("calaculate group id error, code:%d", code);
+    }
+  }
+  recordNewGroupKeys(pParSup->pGroupCols, pParSup->pGroupColVals, pBlock, rowId);
+  int32_t len = buildGroupKeys(pParSup->keyBuf, pParSup->pGroupColVals);
+  uint64_t groupId = calcGroupId(pParSup->keyBuf, len);
+  return groupId;
+}
+
+static bool hasRemainPartion(SStreamPartitionOperatorInfo* pInfo) {
+  return pInfo->parIte != NULL;
+}
+
+static SSDataBlock* buildStreamPartitionResult(SOperatorInfo* pOperator) {
+  SStreamPartitionOperatorInfo* pInfo = pOperator->info;
+  SSDataBlock* pDest = pInfo->binfo.pRes;
+  ASSERT(hasRemainPartion(pInfo));
+  SPartitionDataInfo* pParInfo = (SPartitionDataInfo*)pInfo->parIte;
+  blockDataCleanup(pDest);
+  int32_t rows = taosArrayGetSize(pParInfo->rowIds);
+  SSDataBlock* pSrc = pInfo->pInputDataBlock;
+  for (int32_t i = 0; i < rows; i++) {
+    int32_t rowIndex = *(int32_t*)taosArrayGet(pParInfo->rowIds, i);
+    for (int32_t j = 0; j < pOperator->exprSupp.numOfExprs; j++) {
+      int32_t slotId = pOperator->exprSupp.pExprInfo[j].base.pParam[0].pCol->slotId;
+      SColumnInfoData* pSrcCol = taosArrayGet(pSrc->pDataBlock, slotId);
+      SColumnInfoData* pDestCol = taosArrayGet(pDest->pDataBlock, j);
+      bool isNull = colDataIsNull(pSrcCol, pSrc->info.rows, rowIndex, NULL);
+      char* pSrcData = colDataGetData(pSrcCol, rowIndex);
+      colDataAppend(pDestCol, pDest->info.rows, pSrcData, isNull);
+    }
+    pDest->info.rows++;
+  }
+  blockDataUpdateTsWindow(pDest, pInfo->tsColIndex);
+  pDest->info.groupId = pParInfo->groupId;
+  pOperator->resultInfo.totalRows += pDest->info.rows;
+  pInfo->parIte = taosHashIterate(pInfo->pPartitions, pInfo->parIte);
+  ASSERT(pDest->info.rows > 0);
+  printDataBlock(pDest, "stream partitionby");
+  return pDest;
+}
+
+static void doStreamHashPartitionImpl(SStreamPartitionOperatorInfo* pInfo, SSDataBlock* pBlock) {
+  pInfo->pInputDataBlock = pBlock;
+  for (int32_t i = 0; i < pBlock->info.rows; ++i) {
+    recordNewGroupKeys(pInfo->partitionSup.pGroupCols, pInfo->partitionSup.pGroupColVals, pBlock, i);
+    int32_t keyLen = buildGroupKeys(pInfo->partitionSup.keyBuf, pInfo->partitionSup.pGroupColVals);
+    SPartitionDataInfo* pParData = 
+        (SPartitionDataInfo*) taosHashGet(pInfo->pPartitions, pInfo->partitionSup.keyBuf, keyLen);
+    if (pParData) {
+      taosArrayPush(pParData->rowIds, &i);
+    } else {
+      SPartitionDataInfo newParData = {0};
+      newParData.groupId = calcGroupId(pInfo->partitionSup.keyBuf, keyLen);
+      newParData.rowIds = taosArrayInit(64, sizeof(int32_t));
+      taosArrayPush(newParData.rowIds, &i);
+      taosHashPut(pInfo->pPartitions, pInfo->partitionSup.keyBuf, keyLen, &newParData,
+          sizeof(SPartitionDataInfo));
+    }
+  }
+}
+
+static SSDataBlock* doStreamHashPartition(SOperatorInfo* pOperator) {
+  if (pOperator->status == OP_EXEC_DONE) {
+    return NULL;
+  }
+
+  SExecTaskInfo* pTaskInfo = pOperator->pTaskInfo;
+  SStreamPartitionOperatorInfo* pInfo = pOperator->info;
+  if (hasRemainPartion(pInfo)) {
+    return buildStreamPartitionResult(pOperator);
+  }
+
+  int64_t st = taosGetTimestampUs();
+  SOperatorInfo* downstream = pOperator->pDownstream[0];
+  {
+    pInfo->pInputDataBlock = NULL;
+    SSDataBlock* pBlock = downstream->fpSet.getNextFn(downstream);
+    if (pBlock == NULL) {
+      doSetOperatorCompleted(pOperator);
+      return NULL;
+    }
+    printDataBlock(pBlock, "stream partitionby recv");
+    switch (pBlock->info.type) {
+      case STREAM_NORMAL:
+      case STREAM_PULL_DATA:
+      case STREAM_INVALID:
+        pInfo->binfo.pRes->info.type = pBlock->info.type;
+        break;
+      default:
+        return pBlock;
+    }
+
+    // there is an scalar expression that needs to be calculated right before apply the group aggregation.
+    if (pInfo->scalarSup.pExprInfo != NULL) {
+      pTaskInfo->code = projectApplyFunctions(pInfo->scalarSup.pExprInfo, pBlock, pBlock,
+          pInfo->scalarSup.pCtx, pInfo->scalarSup.numOfExprs, NULL);
+      if (pTaskInfo->code != TSDB_CODE_SUCCESS) {
+        longjmp(pTaskInfo->env, pTaskInfo->code);
+      }
+    }
+    taosHashClear(pInfo->pPartitions);
+    doStreamHashPartitionImpl(pInfo, pBlock);
+  }
+  pOperator->cost.openCost = (taosGetTimestampUs() - st) / 1000.0;
+  
+  pInfo->parIte = taosHashIterate(pInfo->pPartitions, NULL);
+  return buildStreamPartitionResult(pOperator);
+}
+
+static void destroyStreamPartitionOperatorInfo(void* param) {
+  SStreamPartitionOperatorInfo* pInfo = (SStreamPartitionOperatorInfo*)param;
+  cleanupBasicInfo(&pInfo->binfo);
+  taosArrayDestroy(pInfo->partitionSup.pGroupCols);
+
+  for(int i = 0; i < taosArrayGetSize(pInfo->partitionSup.pGroupColVals); i++){
+    SGroupKeys key = *(SGroupKeys*)taosArrayGet(pInfo->partitionSup.pGroupColVals, i);
+    taosMemoryFree(key.pData);
+  }
+  taosArrayDestroy(pInfo->partitionSup.pGroupColVals);
+
+  taosMemoryFree(pInfo->partitionSup.keyBuf);
+  cleanupExprSupp(&pInfo->scalarSup);
+  taosMemoryFreeClear(param);
+}
+
+void initParDownStream(SOperatorInfo* downstream, SPartitionBySupporter* pParSup, SExprSupp* pExpr) {
+  if (downstream->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) {
+    return;
+  }
+  SStreamScanInfo* pScanInfo = downstream->info;
+  pScanInfo->partitionSup = *pParSup;
+  pScanInfo->pPartScalarSup = pExpr;
+}
+
+SOperatorInfo* createStreamPartitionOperatorInfo(SOperatorInfo* downstream, SStreamPartitionPhysiNode* pPartNode, SExecTaskInfo* pTaskInfo) {
+  SStreamPartitionOperatorInfo* pInfo = taosMemoryCalloc(1, sizeof(SStreamPartitionOperatorInfo));
+  SOperatorInfo*                pOperator = taosMemoryCalloc(1, sizeof(SOperatorInfo));
+  if (pInfo == NULL || pOperator == NULL) {
+    goto _error;
+  }
+  int32_t code = TSDB_CODE_SUCCESS;
+  pInfo->partitionSup.pGroupCols = extractPartitionColInfo(pPartNode->pPartitionKeys);
+
+  if (pPartNode->pExprs != NULL) {
+    int32_t num = 0;
+    SExprInfo* pCalExprInfo = createExprInfo(pPartNode->pExprs, NULL, &num);
+    code = initExprSupp(&pInfo->scalarSup, pCalExprInfo, num);
+    if (code != TSDB_CODE_SUCCESS) {
+      goto _error;
+    }
+  }
+
+  int32_t keyLen = 0;
+  code = initGroupOptrInfo(&pInfo->partitionSup.pGroupColVals, &keyLen, &pInfo->partitionSup.keyBuf, pInfo->partitionSup.pGroupCols);
+  if (code != TSDB_CODE_SUCCESS) {
+    goto _error;
+  }
+  pInfo->partitionSup.needCalc = true;
+
+  SSDataBlock* pResBlock = createResDataBlock(pPartNode->node.pOutputDataBlockDesc);
+  if (!pResBlock) {
+    goto _error;
+  }
+  blockDataEnsureCapacity(pResBlock, 4096);
+  pInfo->binfo.pRes       = pResBlock;
+  pInfo->parIte           = NULL;
+  pInfo->pInputDataBlock  = NULL;
+  _hash_fn_t hashFn = taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY);
+  pInfo->pPartitions      = taosHashInit(1024, hashFn, false, HASH_NO_LOCK);
+  pInfo->tsColIndex       = 0;
+
+  int32_t numOfCols = 0;
+  SExprInfo* pExprInfo = createExprInfo(pPartNode->pTargets, NULL, &numOfCols);
+
+  pOperator->name         = "StreamPartitionOperator";
+  pOperator->blocking     = false;
+  pOperator->status       = OP_NOT_OPENED;
+  pOperator->operatorType = QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION;
+  pOperator->exprSupp.numOfExprs   = numOfCols;
+  pOperator->exprSupp.pExprInfo    = pExprInfo;
+  pOperator->info         = pInfo;
+  pOperator->pTaskInfo    = pTaskInfo;
+  pOperator->fpSet = createOperatorFpSet(operatorDummyOpenFn, doStreamHashPartition, NULL, NULL, destroyStreamPartitionOperatorInfo,
+                                         NULL, NULL, NULL);
+
+  initParDownStream(downstream, &pInfo->partitionSup, &pInfo->scalarSup);
+  code = appendDownstream(pOperator, &downstream, 1);
+  return pOperator;
+
+  _error:
+  pTaskInfo->code = TSDB_CODE_OUT_OF_MEMORY;
+  taosMemoryFreeClear(pInfo);
+  taosMemoryFreeClear(pOperator);
+  return NULL;
+}
+

source/libs/executor/src/scanoperator.c

@@ -920,49 +920,28 @@ static void doClearBufferedBlocks(SStreamScanInfo* pInfo) {
 }
 
 static bool isSessionWindow(SStreamScanInfo* pInfo) {
-  return pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION ||
+  return pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION ||
-         pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION;
+         pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION;
 }
 
 static bool isStateWindow(SStreamScanInfo* pInfo) {
-  return pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE;
+  return pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE;
 }
 
 static bool isIntervalWindow(SStreamScanInfo* pInfo) {
-  return pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL ||
+  return pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL ||
-         pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SEMI_INTERVAL ||
+         pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SEMI_INTERVAL ||
-         pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_FINAL_INTERVAL;
+         pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_FINAL_INTERVAL;
 }
 
 static bool isSignleIntervalWindow(SStreamScanInfo* pInfo) {
-  return pInfo->sessionSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL;
+  return pInfo->windowSup.parentType == QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL;
 }
 
 static bool isSlidingWindow(SStreamScanInfo* pInfo) {
   return isIntervalWindow(pInfo) && pInfo->interval.interval != pInfo->interval.sliding;
 }
 
-static uint64_t getGroupId(SOperatorInfo* pOperator, uint64_t uid) {
-  uint64_t* groupId = taosHashGet(pOperator->pTaskInfo->tableqinfoList.map, &uid, sizeof(int64_t));
-  if (groupId) {
-    return *groupId;
-  }
-  return 0;
-  /* Todo(liuyao) for partition by column
-  recordNewGroupKeys(pTableScanInfo->pGroupCols, pTableScanInfo->pGroupColVals, pBlock, rowId);
-  int32_t len = buildGroupKeys(pTableScanInfo->keyBuf, pTableScanInfo->pGroupColVals);
-  uint64_t resId = 0;
-  uint64_t* groupId = taosHashGet(pTableScanInfo->pGroupSet, pTableScanInfo->keyBuf, len);
-  if (groupId) {
-    return *groupId;
-  } else if (len != 0) {
-    resId = calcGroupId(pTableScanInfo->keyBuf, len);
-    taosHashPut(pTableScanInfo->pGroupSet, pTableScanInfo->keyBuf, len, &resId, sizeof(uint64_t));
-  }
-  return resId;
-  */
-}
-
 static void setGroupId(SStreamScanInfo* pInfo, SSDataBlock* pBlock, int32_t groupColIndex, int32_t rowIndex) {
   SColumnInfoData* pColInfo = taosArrayGet(pBlock->pDataBlock, groupColIndex);
   uint64_t*        groupCol = (uint64_t*)pColInfo->pData;
@@ -976,6 +955,61 @@ void resetTableScanInfo(STableScanInfo* pTableScanInfo, STimeWindow* pWin) {
   pTableScanInfo->currentGroupId = -1;
 }
 
+static void freeArray(void* array) { taosArrayDestroy(array); }
+
+static void resetTableScanOperator(SOperatorInfo* pTableScanOp) {
+  STableScanInfo* pTableScanInfo = pTableScanOp->info;
+  pTableScanInfo->cond.startVersion = -1;
+  pTableScanInfo->cond.endVersion = -1;
+  SArray* gpTbls = pTableScanOp->pTaskInfo->tableqinfoList.pGroupList;
+  SArray* allTbls = pTableScanOp->pTaskInfo->tableqinfoList.pTableList;
+  taosArrayClearP(gpTbls, freeArray);
+  taosArrayPush(gpTbls, &allTbls);
+  STimeWindow win = {.skey = INT64_MIN, .ekey = INT64_MAX};
+  resetTableScanInfo(pTableScanOp->info, &win);
+}
+
+static SSDataBlock* readPreVersionData(SOperatorInfo* pTableScanOp, uint64_t tbUid, TSKEY startTs, TSKEY endTs,
+                                       int64_t maxVersion) {
+  SArray* gpTbls = pTableScanOp->pTaskInfo->tableqinfoList.pGroupList;
+  taosArrayClear(gpTbls);
+  STableKeyInfo tblInfo = {.uid = tbUid, .groupId = 0};
+  SArray*       tbls = taosArrayInit(1, sizeof(STableKeyInfo));
+  taosArrayPush(tbls, &tblInfo);
+  taosArrayPush(gpTbls, &tbls);
+
+  STimeWindow     win = {.skey = startTs, .ekey = endTs};
+  STableScanInfo* pTableScanInfo = pTableScanOp->info;
+  pTableScanInfo->cond.startVersion = -1;
+  pTableScanInfo->cond.endVersion = maxVersion;
+  resetTableScanInfo(pTableScanOp->info, &win);
+  SSDataBlock* pRes = doTableScan(pTableScanOp);
+  resetTableScanOperator(pTableScanOp);
+  return pRes;
+}
+
+static uint64_t getGroupIdByCol(SStreamScanInfo* pInfo, uint64_t uid, TSKEY ts, int64_t maxVersion) {
+  SSDataBlock* pPreRes = readPreVersionData(pInfo->pTableScanOp, uid, ts, ts, maxVersion);
+  if (!pPreRes || pPreRes->info.rows == 0) {
+    return 0;
+  }
+  ASSERT(pPreRes->info.rows == 1);
+  return calGroupIdByData(&pInfo->partitionSup, pInfo->pPartScalarSup, pPreRes, 0);
+}
+
+static uint64_t getGroupIdByData(SStreamScanInfo* pInfo, uint64_t uid, TSKEY ts, int64_t maxVersion) {
+  if (pInfo->partitionSup.needCalc) {
+    return getGroupIdByCol(pInfo, uid, ts, maxVersion);
+  }
+
+  SHashObj* map = pInfo->pTableScanOp->pTaskInfo->tableqinfoList.map;
+  uint64_t* groupId = taosHashGet(map, &uid, sizeof(int64_t));
+  if (groupId) {
+    return *groupId;
+  }
+  return 0;
+}
+
 static bool prepareRangeScan(SStreamScanInfo* pInfo, SSDataBlock* pBlock, int32_t* pRowIndex) {
   if ((*pRowIndex) == pBlock->info.rows) {
     return false;
@@ -987,6 +1021,9 @@ static bool prepareRangeScan(SStreamScanInfo* pInfo, SSDataBlock* pBlock, int32_
   SColumnInfoData* pEndTsCol = taosArrayGet(pBlock->pDataBlock, END_TS_COLUMN_INDEX);
   TSKEY*           endData = (TSKEY*)pEndTsCol->pData;
   STimeWindow      win = {.skey = startData[*pRowIndex], .ekey = endData[*pRowIndex]};
+  SColumnInfoData* pGpCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
+  uint64_t*        gpData = (uint64_t*)pGpCol->pData;
+  uint64_t         groupId = gpData[*pRowIndex];
 
   SColumnInfoData* pCalStartTsCol = taosArrayGet(pBlock->pDataBlock, CALCULATE_START_TS_COLUMN_INDEX);
   TSKEY*           calStartData = (TSKEY*)pCalStartTsCol->pData;
@@ -1001,11 +1038,11 @@ static bool prepareRangeScan(SStreamScanInfo* pInfo, SSDataBlock* pBlock, int32_
   (*pRowIndex)++;
 
   for (; *pRowIndex < pBlock->info.rows; (*pRowIndex)++) {
-    if (win.skey == startData[*pRowIndex]) {
+    if (win.skey == startData[*pRowIndex] && groupId == gpData[*pRowIndex]) {
       win.ekey = TMAX(win.ekey, endData[*pRowIndex]);
       continue;
     }
-    if (win.skey == endData[*pRowIndex]) {
+    if (win.skey == endData[*pRowIndex] && groupId == gpData[*pRowIndex]) {
       win.skey = TMIN(win.skey, startData[*pRowIndex]);
       continue;
     }
@@ -1020,15 +1057,19 @@ static bool prepareRangeScan(SStreamScanInfo* pInfo, SSDataBlock* pBlock, int32_
 }
 
 static STimeWindow getSlidingWindow(TSKEY* tsCol, SInterval* pInterval, SDataBlockInfo* pDataBlockInfo,
-                                    int32_t* pRowIndex) {
+                                    int32_t* pRowIndex, bool hasGroup) {
   SResultRowInfo dumyInfo;
   dumyInfo.cur.pageId = -1;
   STimeWindow win = getActiveTimeWindow(NULL, &dumyInfo, tsCol[*pRowIndex], pInterval, TSDB_ORDER_ASC);
   STimeWindow endWin = win;
   STimeWindow preWin = win;
   while (1) {
+    if (hasGroup) {
+      (*pRowIndex) += 1;
+    } else {
       (*pRowIndex) += getNumOfRowsInTimeWindow(pDataBlockInfo, tsCol, *pRowIndex, endWin.ekey, binarySearchForKey, NULL,
                                                TSDB_ORDER_ASC);
+    }
     do {
       preWin = endWin;
       getNextTimeWindow(pInterval, &endWin, TSDB_ORDER_ASC);
@@ -1060,7 +1101,26 @@ static SSDataBlock* doRangeScan(SStreamScanInfo* pInfo, SSDataBlock* pSDB, int32
       return NULL;
     }
 
-    if (pResult->info.groupId == pInfo->groupId) {
+    if (pInfo->partitionSup.needCalc) {
+      SSDataBlock* tmpBlock = createOneDataBlock(pResult, true);
+      blockDataCleanup(pResult);
+      for (int32_t i = 0; i < tmpBlock->info.rows; i++) {
+        if (calGroupIdByData(&pInfo->partitionSup, pInfo->pPartScalarSup, tmpBlock, i) == pInfo->groupId) {
+          for (int32_t j = 0; j < pInfo->pTableScanOp->exprSupp.numOfExprs; j++) {
+            SColumnInfoData* pSrcCol = taosArrayGet(tmpBlock->pDataBlock, j);
+            SColumnInfoData* pDestCol = taosArrayGet(pResult->pDataBlock, j);
+            bool             isNull = colDataIsNull(pSrcCol, tmpBlock->info.rows, i, NULL);
+            char*            pSrcData = colDataGetData(pSrcCol, i);
+            colDataAppend(pDestCol, pResult->info.rows, pSrcData, isNull);
+          }
+          pResult->info.rows++;
+        }
+      }
+      if (pResult->info.rows > 0) {
+        pResult->info.calWin = pInfo->updateWin;
+        return pResult;
+      }
+    } else if (pResult->info.groupId == pInfo->groupId) {
       pResult->info.calWin = pInfo->updateWin;
       return pResult;
     }
@@ -1091,17 +1151,18 @@ static int32_t generateSessionScanRange(SStreamScanInfo* pInfo, SSDataBlock* pSr
   SColumnInfoData* pDestCalStartTsCol = taosArrayGet(pDestBlock->pDataBlock, CALCULATE_START_TS_COLUMN_INDEX);
   SColumnInfoData* pDestCalEndTsCol = taosArrayGet(pDestBlock->pDataBlock, CALCULATE_END_TS_COLUMN_INDEX);
   int32_t          dummy = 0;
+  int64_t          version = pSrcBlock->info.version - 1;
   for (int32_t i = 0; i < pSrcBlock->info.rows; i++) {
-    uint64_t groupId = getGroupId(pInfo->pTableScanOp, uidCol[i]);
+    uint64_t groupId = getGroupIdByData(pInfo, uidCol[i], startData[i], version);
     // gap must be 0.
     SResultWindowInfo* pStartWin =
-        getCurSessionWindow(pInfo->sessionSup.pStreamAggSup, startData[i], endData[i], groupId, 0, &dummy);
+        getCurSessionWindow(pInfo->windowSup.pStreamAggSup, startData[i], endData[i], groupId, 0, &dummy);
     if (!pStartWin) {
       // window has been closed.
       continue;
     }
     SResultWindowInfo* pEndWin =
-        getCurSessionWindow(pInfo->sessionSup.pStreamAggSup, endData[i], endData[i], groupId, 0, &dummy);
+        getCurSessionWindow(pInfo->windowSup.pStreamAggSup, endData[i], endData[i], groupId, 0, &dummy);
     ASSERT(pEndWin);
     TSKEY ts = INT64_MIN;
     colDataAppend(pDestStartCol, i, (const char*)&pStartWin->win.skey, false);
@@ -1121,34 +1182,49 @@ static int32_t generateIntervalScanRange(SStreamScanInfo* pInfo, SSDataBlock* pS
   if (rows == 0) {
     return TSDB_CODE_SUCCESS;
   }
-  int32_t code = blockDataEnsureCapacity(pDestBlock, rows);
+  int32_t code = blockDataEnsureCapacity(pDestBlock, rows * 2);
   if (code != TSDB_CODE_SUCCESS) {
     return code;
   }
 
-  SColumnInfoData* pTsCol = (SColumnInfoData*)taosArrayGet(pSrcBlock->pDataBlock, START_TS_COLUMN_INDEX);
+  SColumnInfoData* pSrcTsCol = (SColumnInfoData*)taosArrayGet(pSrcBlock->pDataBlock, START_TS_COLUMN_INDEX);
-  SColumnInfoData* pUidCol = taosArrayGet(pSrcBlock->pDataBlock, UID_COLUMN_INDEX);
+  SColumnInfoData* pSrcUidCol = taosArrayGet(pSrcBlock->pDataBlock, UID_COLUMN_INDEX);
-  uint64_t*        uidCol = (uint64_t*)pUidCol->pData;
+  uint64_t*        srcUidData = (uint64_t*)pSrcUidCol->pData;
-  ASSERT(pTsCol->info.type == TSDB_DATA_TYPE_TIMESTAMP);
+  SColumnInfoData* pSrcGpCol = taosArrayGet(pSrcBlock->pDataBlock, GROUPID_COLUMN_INDEX);
-  TSKEY*           tsCol = (TSKEY*)pTsCol->pData;
+  uint64_t*        srcGp = (uint64_t*)pSrcGpCol->pData;
+  ASSERT(pSrcTsCol->info.type == TSDB_DATA_TYPE_TIMESTAMP);
+  TSKEY*           tsCol = (TSKEY*)pSrcTsCol->pData;
   SColumnInfoData* pStartTsCol = taosArrayGet(pDestBlock->pDataBlock, START_TS_COLUMN_INDEX);
   SColumnInfoData* pEndTsCol = taosArrayGet(pDestBlock->pDataBlock, END_TS_COLUMN_INDEX);
+  SColumnInfoData* pDeUidCol = taosArrayGet(pDestBlock->pDataBlock, UID_COLUMN_INDEX);
   SColumnInfoData* pGpCol = taosArrayGet(pDestBlock->pDataBlock, GROUPID_COLUMN_INDEX);
   SColumnInfoData* pCalStartTsCol = taosArrayGet(pDestBlock->pDataBlock, CALCULATE_START_TS_COLUMN_INDEX);
   SColumnInfoData* pCalEndTsCol = taosArrayGet(pDestBlock->pDataBlock, CALCULATE_END_TS_COLUMN_INDEX);
-  uint64_t         groupId = getGroupId(pInfo->pTableScanOp, uidCol[0]);
+  int64_t          version = pSrcBlock->info.version - 1;
   for (int32_t i = 0; i < rows;) {
-    colDataAppend(pCalStartTsCol, pDestBlock->info.rows, (const char*)(tsCol + i), false);
+    uint64_t srcUid = srcUidData[i];
-    STimeWindow win = getSlidingWindow(tsCol, &pInfo->interval, &pSrcBlock->info, &i);
+    uint64_t groupId = getGroupIdByData(pInfo, srcUid, tsCol[i], version);
-    colDataAppend(pCalEndTsCol, pDestBlock->info.rows, (const char*)(tsCol + i - 1), false);
+    uint64_t srcGpId = srcGp[i];
-
+    TSKEY    calStartTs = tsCol[i];
+    colDataAppend(pCalStartTsCol, pDestBlock->info.rows, (const char*)(&calStartTs), false);
+    STimeWindow win = getSlidingWindow(tsCol, &pInfo->interval, &pSrcBlock->info, &i, pInfo->partitionSup.needCalc);
+    TSKEY       calEndTs = tsCol[i - 1];
+    colDataAppend(pCalEndTsCol, pDestBlock->info.rows, (const char*)(&calEndTs), false);
+    colDataAppend(pDeUidCol, pDestBlock->info.rows, (const char*)(&srcUid), false);
     colDataAppend(pStartTsCol, pDestBlock->info.rows, (const char*)(&win.skey), false);
     colDataAppend(pEndTsCol, pDestBlock->info.rows, (const char*)(&win.ekey), false);
     colDataAppend(pGpCol, pDestBlock->info.rows, (const char*)(&groupId), false);
     pDestBlock->info.rows++;
+    if (pInfo->partitionSup.needCalc && srcGpId != 0 && groupId != srcGpId) {
+      colDataAppend(pCalStartTsCol, pDestBlock->info.rows, (const char*)(&calStartTs), false);
+      colDataAppend(pCalEndTsCol, pDestBlock->info.rows, (const char*)(&calEndTs), false);
+      colDataAppend(pDeUidCol, pDestBlock->info.rows, (const char*)(&srcUid), false);
+      colDataAppend(pStartTsCol, pDestBlock->info.rows, (const char*)(&win.skey), false);
+      colDataAppend(pEndTsCol, pDestBlock->info.rows, (const char*)(&win.ekey), false);
+      colDataAppend(pGpCol, pDestBlock->info.rows, (const char*)(&srcGpId), false);
+      pDestBlock->info.rows++;
+    }
   }
-  // all rows have same group id
-  pDestBlock->info.groupId = groupId;
   return TSDB_CODE_SUCCESS;
 }
 
@@ -1160,17 +1236,20 @@ static int32_t generateScanRange(SStreamScanInfo* pInfo, SSDataBlock* pSrcBlock,
     code = generateSessionScanRange(pInfo, pSrcBlock, pDestBlock);
   }
   pDestBlock->info.type = STREAM_CLEAR;
+  pDestBlock->info.version = pSrcBlock->info.version;
   blockDataUpdateTsWindow(pDestBlock, 0);
   return code;
 }
 
-void appendOneRow(SSDataBlock* pBlock, TSKEY* pStartTs, TSKEY* pEndTs, int32_t uidCol, uint64_t* pID) {
+void appendOneRow(SSDataBlock* pBlock, TSKEY* pStartTs, TSKEY* pEndTs, uint64_t* pUid, uint64_t* pGp) {
   SColumnInfoData* pStartTsCol = taosArrayGet(pBlock->pDataBlock, START_TS_COLUMN_INDEX);
   SColumnInfoData* pEndTsCol = taosArrayGet(pBlock->pDataBlock, END_TS_COLUMN_INDEX);
-  SColumnInfoData* pUidCol = taosArrayGet(pBlock->pDataBlock, uidCol);
+  SColumnInfoData* pUidCol = taosArrayGet(pBlock->pDataBlock, UID_COLUMN_INDEX);
+  SColumnInfoData* pGpCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
   colDataAppend(pStartTsCol, pBlock->info.rows, (const char*)pStartTs, false);
   colDataAppend(pEndTsCol, pBlock->info.rows, (const char*)pEndTs, false);
-  colDataAppend(pUidCol, pBlock->info.rows, (const char*)pID, false);
+  colDataAppend(pUidCol, pBlock->info.rows, (const char*)pUid, false);
+  colDataAppend(pGpCol, pBlock->info.rows, (const char*)pGp, false);
   pBlock->info.rows++;
 }
 
@@ -1195,24 +1274,18 @@ static void checkUpdateData(SStreamScanInfo* pInfo, bool invertible, SSDataBlock
     // must check update info first.
     bool update = updateInfoIsUpdated(pInfo->pUpdateInfo, pBlock->info.uid, tsCol[rowId]);
     bool closedWin = isClosed && isSignleIntervalWindow(pInfo) &&
-                     isDeletedWindow(&win, pBlock->info.groupId, pInfo->sessionSup.pIntervalAggSup);
+                     isDeletedWindow(&win, pBlock->info.groupId, pInfo->windowSup.pIntervalAggSup);
     if ((update || closedWin) && out) {
-      appendOneRow(pInfo->pUpdateDataRes, tsCol + rowId, tsCol + rowId, UID_COLUMN_INDEX, &pBlock->info.uid);
+      uint64_t gpId = closedWin && pInfo->partitionSup.needCalc
+                          ? calGroupIdByData(&pInfo->partitionSup, pInfo->pPartScalarSup, pBlock, rowId)
+                          : 0;
+      appendOneRow(pInfo->pUpdateDataRes, tsCol + rowId, tsCol + rowId, &pBlock->info.uid, &gpId);
     }
   }
-  if (out) {
+  if (out && pInfo->pUpdateDataRes->info.rows > 0) {
+    pInfo->pUpdateDataRes->info.version = pBlock->info.version;
     blockDataUpdateTsWindow(pInfo->pUpdateDataRes, 0);
-    pInfo->pUpdateDataRes->info.type = STREAM_CLEAR;
+    pInfo->pUpdateDataRes->info.type = pInfo->partitionSup.needCalc ? STREAM_DELETE_DATA : STREAM_CLEAR;
-  }
-}
-
-static void setBlockGroupId(SOperatorInfo* pOperator, SSDataBlock* pBlock, int32_t uidColIndex) {
-  ASSERT(taosArrayGetSize(pBlock->pDataBlock) >= 3);
-  SColumnInfoData* pColDataInfo = taosArrayGet(pBlock->pDataBlock, uidColIndex);
-  uint64_t*        uidCol = (uint64_t*)pColDataInfo->pData;
-  ASSERT(pBlock->info.rows > 0);
-  for (int32_t i = 0; i < pBlock->info.rows; i++) {
-    uidCol[i] = getGroupId(pOperator, uidCol[i]);
   }
 }
 
@@ -1441,12 +1514,14 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
         generateScanRange(pInfo, pBlock, pInfo->pUpdateRes);
         prepareRangeScan(pInfo, pInfo->pUpdateRes, &pInfo->updateResIndex);
         copyDataBlock(pInfo->pDeleteDataRes, pInfo->pUpdateRes);
+        pInfo->pDeleteDataRes->info.type = STREAM_DELETE_DATA;
         pInfo->scanMode = STREAM_SCAN_FROM_DATAREADER_RANGE;
         return pInfo->pDeleteDataRes;
       } break;
       default:
         break;
     }
+    // printDataBlock(pBlock, "stream scan recv");
     return pBlock;
   } else if (pInfo->blockType == STREAM_INPUT__DATA_SUBMIT) {
     qDebug("scan mode %d", pInfo->scanMode);
@@ -1456,6 +1531,14 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
         pInfo->scanMode = STREAM_SCAN_FROM_READERHANDLE;
         return pInfo->pRes;
       } break;
+      case STREAM_SCAN_FROM_DELETERES: {
+        generateScanRange(pInfo, pInfo->pUpdateDataRes, pInfo->pUpdateRes);
+        prepareRangeScan(pInfo, pInfo->pUpdateRes, &pInfo->updateResIndex);
+        pInfo->scanMode = STREAM_SCAN_FROM_DATAREADER_RANGE;
+        copyDataBlock(pInfo->pDeleteDataRes, pInfo->pUpdateRes);
+        pInfo->pDeleteDataRes->info.type = STREAM_DELETE_DATA;
+        return pInfo->pDeleteDataRes;
+      } break;
       case STREAM_SCAN_FROM_UPDATERES: {
         generateScanRange(pInfo, pInfo->pUpdateDataRes, pInfo->pUpdateRes);
         prepareRangeScan(pInfo, pInfo->pUpdateRes, &pInfo->updateResIndex);
@@ -1471,6 +1554,7 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
           updateInfoSetScanRange(pInfo->pUpdateInfo, &pTableScanInfo->cond.twindows, pInfo->groupId, version);
           pSDB->info.type = pInfo->scanMode == STREAM_SCAN_FROM_DATAREADER_RANGE ? STREAM_NORMAL : STREAM_PULL_DATA;
           checkUpdateData(pInfo, true, pSDB, false);
+          // printDataBlock(pSDB, "stream scan update");
           return pSDB;
         }
         pInfo->scanMode = STREAM_SCAN_FROM_READERHANDLE;
@@ -1479,7 +1563,7 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
         break;
     }
 
-    SStreamAggSupporter* pSup = pInfo->sessionSup.pStreamAggSup;
+    SStreamAggSupporter* pSup = pInfo->windowSup.pStreamAggSup;
     if (isStateWindow(pInfo) && pSup->pScanBlock->info.rows > 0) {
       pInfo->scanMode = STREAM_SCAN_FROM_DATAREADER_RANGE;
       pInfo->updateResIndex = 0;
@@ -1545,7 +1629,7 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
     // record the scan action.
     pInfo->numOfExec++;
     pOperator->resultInfo.totalRows += pBlockInfo->rows;
-    printDataBlock(pInfo->pRes, "stream scan");
+    // printDataBlock(pInfo->pRes, "stream scan");
 
     if (pBlockInfo->rows == 0) {
       updateInfoDestoryColseWinSBF(pInfo->pUpdateInfo);
@@ -1554,19 +1638,20 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
       checkUpdateData(pInfo, true, pInfo->pRes, true);
       pInfo->twAggSup.maxTs = TMAX(pInfo->twAggSup.maxTs, pBlockInfo->window.ekey);
       if (pInfo->pUpdateDataRes->info.rows > 0) {
-        if (pInfo->pUpdateDataRes->info.type == STREAM_CLEAR) {
         pInfo->updateResIndex = 0;
+        if (pInfo->pUpdateDataRes->info.type == STREAM_CLEAR) {
           pInfo->scanMode = STREAM_SCAN_FROM_UPDATERES;
         } else if (pInfo->pUpdateDataRes->info.type == STREAM_INVERT) {
           pInfo->scanMode = STREAM_SCAN_FROM_RES;
           return pInfo->pUpdateDataRes;
+        } else if (pInfo->pUpdateDataRes->info.type == STREAM_DELETE_DATA) {
+          pInfo->scanMode = STREAM_SCAN_FROM_DELETERES;
         }
       }
     }
 
     qDebug("scan rows: %d", pBlockInfo->rows);
     return (pBlockInfo->rows == 0) ? NULL : pInfo->pRes;
-
   } else {
     ASSERT(0);
     return NULL;
@@ -1831,12 +1916,7 @@ SOperatorInfo* createStreamScanOperatorInfo(SReadHandle* pHandle, STableScanPhys
       pInfo->tqReader = pHandle->tqReader;
     }
 
-    if (pTSInfo->pdInfo.interval.interval > 0) {
-      pInfo->pUpdateInfo = updateInfoInitP(&pTSInfo->pdInfo.interval, pInfo->twAggSup.waterMark);
-    } else {
     pInfo->pUpdateInfo = NULL;
-    }
-
     pInfo->pTableScanOp = pTableScanOp;
     pInfo->interval = pTSInfo->pdInfo.interval;
 
@@ -1867,8 +1947,7 @@ SOperatorInfo* createStreamScanOperatorInfo(SReadHandle* pHandle, STableScanPhys
   pInfo->pUpdateRes = createSpecialDataBlock(STREAM_CLEAR);
   pInfo->pCondition = pScanPhyNode->node.pConditions;
   pInfo->scanMode = STREAM_SCAN_FROM_READERHANDLE;
-  pInfo->sessionSup =
+  pInfo->windowSup = (SWindowSupporter){.pStreamAggSup = NULL, .gap = -1, .parentType = QUERY_NODE_PHYSICAL_PLAN};
-      (SessionWindowSupporter){.pStreamAggSup = NULL, .gap = -1, .parentType = QUERY_NODE_PHYSICAL_PLAN};
   pInfo->groupId = 0;
   pInfo->pPullDataRes = createSpecialDataBlock(STREAM_RETRIEVE);
   pInfo->pStreamScanOp = pOperator;
@@ -1877,6 +1956,7 @@ SOperatorInfo* createStreamScanOperatorInfo(SReadHandle* pHandle, STableScanPhys
   pInfo->updateWin = (STimeWindow){.skey = INT64_MAX, .ekey = INT64_MAX};
   pInfo->pUpdateDataRes = createSpecialDataBlock(STREAM_CLEAR);
   pInfo->assignBlockUid = pTableScanNode->assignBlockUid;
+  pInfo->partitionSup.needCalc = false;
 
   pOperator->name = "StreamScanOperator";
   pOperator->operatorType = QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN;

source/libs/executor/src/timewindowoperator.c

@@ -42,7 +42,8 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator);
 
 static int64_t* extractTsCol(SSDataBlock* pBlock, const SIntervalAggOperatorInfo* pInfo);
 
-static SResultRowPosition addToOpenWindowList(SResultRowInfo* pResultRowInfo, const SResultRow* pResult, uint64_t groupId);
+static SResultRowPosition addToOpenWindowList(SResultRowInfo* pResultRowInfo, const SResultRow* pResult,
+                                              uint64_t groupId);
 static void doCloseWindow(SResultRowInfo* pResultRowInfo, const SIntervalAggOperatorInfo* pInfo, SResultRow* pResult);
 
 ///*
@@ -897,7 +898,7 @@ int64_t getWinReskey(void* data, int32_t index) {
 
 int32_t compareWinRes(void* pKey, void* data, int32_t index) {
   SArray*     res = (SArray*)data;
-  SWinKey*    pos = taosArrayGetP(res, index);
+  SWinKey*    pos = taosArrayGet(res, index);
   SResKeyPos* pData = (SResKeyPos*)pKey;
   if (*(int64_t*)pData->key == pos->ts) {
     if (pData->groupId > pos->groupId) {
@@ -919,10 +920,11 @@ static void removeDeleteResults(SHashObj* pUpdatedMap, SArray* pDelWins) {
   }
   void* pIte = NULL;
   while ((pIte = taosHashIterate(pUpdatedMap, pIte)) != NULL) {
-    SResKeyPos* pResKey = (SResKeyPos*)pIte;
+    SResKeyPos* pResKey = *(SResKeyPos**)pIte;
     int32_t     index = binarySearchCom(pDelWins, delSize, pResKey, TSDB_ORDER_DESC, compareWinRes);
     if (index >= 0 && 0 == compareWinRes(pResKey, pDelWins, index)) {
       taosArrayRemove(pDelWins, index);
+      delSize = taosArrayGetSize(pDelWins);
     }
   }
 }
@@ -1423,7 +1425,8 @@ bool doDeleteIntervalWindow(SAggSupporter* pAggSup, TSKEY ts, uint64_t groupId)
   return true;
 }
 
-void doDeleteSpecifyIntervalWindow(SAggSupporter* pAggSup, SSDataBlock* pBlock, SArray* pUpWins, SInterval* pInterval) {
+void doDeleteSpecifyIntervalWindow(SAggSupporter* pAggSup, SSDataBlock* pBlock, SArray* pDelWins, SInterval* pInterval,
+                                   SHashObj* pUpdatedMap) {
   SColumnInfoData* pStartCol = taosArrayGet(pBlock->pDataBlock, START_TS_COLUMN_INDEX);
   TSKEY*           tsStarts = (TSKEY*)pStartCol->pData;
   SColumnInfoData* pGroupCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
@@ -1433,9 +1436,12 @@ void doDeleteSpecifyIntervalWindow(SAggSupporter* pAggSup, SSDataBlock* pBlock,
     dumyInfo.cur.pageId = -1;
     STimeWindow win = getActiveTimeWindow(NULL, &dumyInfo, tsStarts[i], pInterval, TSDB_ORDER_ASC);
     doDeleteIntervalWindow(pAggSup, win.skey, groupIds[i]);
-    if (pUpWins) {
     SWinKey winRes = {.ts = win.skey, .groupId = groupIds[i]};
-      taosArrayPush(pUpWins, &winRes);
+    if (pDelWins) {
+      taosArrayPush(pDelWins, &winRes);
+    }
+    if (pUpdatedMap) {
+      taosHashRemove(pUpdatedMap, &winRes, sizeof(SWinKey));
     }
   }
 }
@@ -1446,19 +1452,14 @@ static void doClearWindows(SAggSupporter* pAggSup, SExprSupp* pSup1, SInterval*
   TSKEY*           startTsCols = (TSKEY*)pStartTsCol->pData;
   SColumnInfoData* pEndTsCol = taosArrayGet(pBlock->pDataBlock, END_TS_COLUMN_INDEX);
   TSKEY*           endTsCols = (TSKEY*)pEndTsCol->pData;
-  uint64_t*        pGpDatas = NULL;
-  if (pBlock->info.type == STREAM_RETRIEVE) {
   SColumnInfoData* pGpCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
-    pGpDatas = (uint64_t*)pGpCol->pData;
+  uint64_t*        pGpDatas = (uint64_t*)pGpCol->pData;
-  }
-  int32_t step = 0;
-  int32_t startPos = 0;
   for (int32_t i = 0; i < pBlock->info.rows; i++) {
     SResultRowInfo dumyInfo;
     dumyInfo.cur.pageId = -1;
     STimeWindow win = getActiveTimeWindow(NULL, &dumyInfo, startTsCols[i], pInterval, TSDB_ORDER_ASC);
     while (win.ekey <= endTsCols[i]) {
-      uint64_t winGpId = pGpDatas ? pGpDatas[startPos] : pBlock->info.groupId;
+      uint64_t winGpId = pGpDatas[i];
       bool     res = doClearWindow(pAggSup, pSup1, (char*)&win.skey, sizeof(TSKEY), winGpId, numOfOutput);
       if (pUpWins && res) {
         SWinKey winRes = {.ts = win.skey, .groupId = winGpId};
@@ -1470,7 +1471,6 @@ static void doClearWindows(SAggSupporter* pAggSup, SExprSupp* pSup1, SInterval*
 }
 
 static int32_t getAllIntervalWindow(SSHashObj* pHashMap, SHashObj* resWins) {
-
   void*   pIte = NULL;
   size_t  keyLen = 0;
   int32_t iter = 0;
@@ -1571,13 +1571,10 @@ static void doBuildDeleteResult(SArray* pWins, int32_t* index, SSDataBlock* pBlo
     return;
   }
   blockDataEnsureCapacity(pBlock, size - *index);
-  SColumnInfoData* pTsCol = taosArrayGet(pBlock->pDataBlock, START_TS_COLUMN_INDEX);
+  uint64_t uid = 0;
-  SColumnInfoData* pGroupCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
   for (int32_t i = *index; i < size; i++) {
     SWinKey* pWin = taosArrayGet(pWins, i);
-    colDataAppend(pTsCol, pBlock->info.rows, (const char*)&pWin->ts, false);
+    appendOneRow(pBlock, &pWin->ts, &pWin->ts, &uid, &pWin->groupId);
-    colDataAppend(pGroupCol, pBlock->info.rows, (const char*)&pWin->groupId, false);
-    pBlock->info.rows++;
     (*index)++;
   }
 }
@@ -1596,6 +1593,7 @@ static SSDataBlock* doStreamIntervalAgg(SOperatorInfo* pOperator) {
   if (pOperator->status == OP_RES_TO_RETURN) {
     doBuildDeleteResult(pInfo->pDelWins, &pInfo->delIndex, pInfo->pDelRes);
     if (pInfo->pDelRes->info.rows > 0) {
+      printDataBlock(pInfo->pDelRes, "single interval");
       return pInfo->pDelRes;
     }
 
@@ -1632,7 +1630,7 @@ static SSDataBlock* doStreamIntervalAgg(SOperatorInfo* pOperator) {
       continue;
     }
     if (pBlock->info.type == STREAM_DELETE_DATA) {
-      doDeleteSpecifyIntervalWindow(&pInfo->aggSup, pBlock, pInfo->pDelWins, &pInfo->interval);
+      doDeleteSpecifyIntervalWindow(&pInfo->aggSup, pBlock, pInfo->pDelWins, &pInfo->interval, pUpdatedMap);
       continue;
     } else if (pBlock->info.type == STREAM_GET_ALL) {
       getAllIntervalWindow(pInfo->aggSup.pResultRowHashTable, pUpdatedMap);
@@ -1707,6 +1705,7 @@ static SSDataBlock* doStreamIntervalAgg(SOperatorInfo* pOperator) {
   taosHashCleanup(pUpdatedMap);
   doBuildDeleteResult(pInfo->pDelWins, &pInfo->delIndex, pInfo->pDelRes);
   if (pInfo->pDelRes->info.rows > 0) {
+    printDataBlock(pInfo->pDelRes, "single interval");
     return pInfo->pDelRes;
   }
 
@@ -1828,14 +1827,17 @@ static bool timeWindowinterpNeeded(SqlFunctionCtx* pCtx, int32_t numOfCols, SInt
   return needed;
 }
 
-void initIntervalDownStream(SOperatorInfo* downstream, uint16_t type, SAggSupporter* pSup) {
+void initIntervalDownStream(SOperatorInfo* downstream, uint16_t type, SAggSupporter* pSup, SInterval* pInterval,
+                            int64_t waterMark) {
   if (downstream->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) {
-    // Todo(liuyao) support partition by column
+    initIntervalDownStream(downstream->pDownstream[0], type, pSup, pInterval, waterMark);
     return;
   }
   SStreamScanInfo* pScanInfo = downstream->info;
-  pScanInfo->sessionSup.parentType = type;
+  pScanInfo->windowSup.parentType = type;
-  pScanInfo->sessionSup.pIntervalAggSup = pSup;
+  pScanInfo->windowSup.pIntervalAggSup = pSup;
+  pScanInfo->pUpdateInfo = updateInfoInitP(pInterval, waterMark);
+  pScanInfo->interval = *pInterval;
 }
 
 void initStreamFunciton(SqlFunctionCtx* pCtx, int32_t numOfExpr) {
@@ -1921,7 +1923,8 @@ SOperatorInfo* createIntervalOperatorInfo(SOperatorInfo* downstream, SExprInfo*
                                          destroyIntervalOperatorInfo, aggEncodeResultRow, aggDecodeResultRow, NULL);
 
   if (nodeType(pPhyNode) == QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL) {
-    initIntervalDownStream(downstream, QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL, &pInfo->aggSup);
+    initIntervalDownStream(downstream, QUERY_NODE_PHYSICAL_PLAN_STREAM_INTERVAL, &pInfo->aggSup, &pInfo->interval,
+                           pInfo->twAggSup.waterMark);
   }
 
   code = appendDownstream(pOperator, &downstream, 1);
@@ -2248,7 +2251,6 @@ static void genInterpolationResult(STimeSliceOperatorInfo* pSliceInfo, SExprSupp
   if (hasInterp) {
     pResBlock->info.rows += 1;
   }
-
 }
 
 static int32_t initPrevRowsKeeper(STimeSliceOperatorInfo* pInfo, SSDataBlock* pBlock) {
@@ -2748,7 +2750,9 @@ SOperatorInfo* createStatewindowOperatorInfo(SOperatorInfo* downstream, SStateWi
   initBasicInfo(&pInfo->binfo, pResBlock);
   initResultRowInfo(&pInfo->binfo.resultRowInfo);
 
-  pInfo->twAggSup = (STimeWindowAggSupp){.waterMark = pStateNode->window.watermark, .calTrigger = pStateNode->window.triggerType};;
+  pInfo->twAggSup =
+      (STimeWindowAggSupp){.waterMark = pStateNode->window.watermark, .calTrigger = pStateNode->window.triggerType};
+  ;
   initExecTimeWindowInfo(&pInfo->twAggSup.timeWindowData, &pTaskInfo->window);
 
   pInfo->tsSlotId = tsSlotId;
@@ -2849,14 +2853,26 @@ _error:
 }
 
 void compactFunctions(SqlFunctionCtx* pDestCtx, SqlFunctionCtx* pSourceCtx, int32_t numOfOutput,
-                      SExecTaskInfo* pTaskInfo) {
+                      SExecTaskInfo* pTaskInfo, SColumnInfoData* pTimeWindowData) {
   for (int32_t k = 0; k < numOfOutput; ++k) {
     if (fmIsWindowPseudoColumnFunc(pDestCtx[k].functionId)) {
+      if (!pTimeWindowData) {
         continue;
       }
-    int32_t code = TSDB_CODE_SUCCESS;
+
-    if (functionNeedToExecute(&pDestCtx[k]) && pDestCtx[k].fpSet.combine != NULL) {
+      SResultRowEntryInfo* pEntryInfo = GET_RES_INFO(&pDestCtx[k]);
-      code = pDestCtx[k].fpSet.combine(&pDestCtx[k], &pSourceCtx[k]);
+      char*                p = GET_ROWCELL_INTERBUF(pEntryInfo);
+      SColumnInfoData      idata = {0};
+      idata.info.type = TSDB_DATA_TYPE_BIGINT;
+      idata.info.bytes = tDataTypes[TSDB_DATA_TYPE_BIGINT].bytes;
+      idata.pData = p;
+
+      SScalarParam out = {.columnData = &idata};
+      SScalarParam tw = {.numOfRows = 5, .columnData = pTimeWindowData};
+      pDestCtx[k].sfp.process(&tw, 1, &out);
+      pEntryInfo->numOfRes = 1;
+    } else if (functionNeedToExecute(&pDestCtx[k]) && pDestCtx[k].fpSet.combine != NULL) {
+      int32_t code = pDestCtx[k].fpSet.combine(&pDestCtx[k], &pSourceCtx[k]);
       if (code != TSDB_CODE_SUCCESS) {
         qError("%s apply functions error, code: %s", GET_TASKID(pTaskInfo), tstrerror(code));
         pTaskInfo->code = code;
@@ -2874,8 +2890,15 @@ bool hasIntervalWindow(SAggSupporter* pSup, TSKEY ts, uint64_t groupId) {
   return p1 != NULL;
 }
 
+STimeWindow getFinalTimeWindow(int64_t ts, SInterval* pInterval) {
+  STimeWindow w = {.skey = ts, .ekey = INT64_MAX};
+  w.ekey = taosTimeAdd(w.skey, pInterval->interval, pInterval->intervalUnit, pInterval->precision) - 1;
+  return w;
+}
+
 static void rebuildIntervalWindow(SStreamFinalIntervalOperatorInfo* pInfo, SExprSupp* pSup, SArray* pWinArray,
-                                  int32_t groupId, int32_t numOfOutput, SExecTaskInfo* pTaskInfo, SArray* pUpdated) {
+                                  int32_t groupId, int32_t numOfOutput, SExecTaskInfo* pTaskInfo,
+                                  SHashObj* pUpdatedMap) {
   int32_t size = taosArrayGetSize(pWinArray);
   if (!pInfo->pChildren) {
     return;
@@ -2883,11 +2906,14 @@ static void rebuildIntervalWindow(SStreamFinalIntervalOperatorInfo* pInfo, SExpr
   for (int32_t i = 0; i < size; i++) {
     SWinKey*    pWinRes = taosArrayGet(pWinArray, i);
     SResultRow* pCurResult = NULL;
-    STimeWindow ParentWin = {.skey = pWinRes->ts, .ekey = pWinRes->ts + 1};
+    STimeWindow parentWin = getFinalTimeWindow(pWinRes->ts, &pInfo->interval);
-    setTimeWindowOutputBuf(&pInfo->binfo.resultRowInfo, &ParentWin, true, &pCurResult, pWinRes->groupId, pSup->pCtx,
+    if (isDeletedWindow(&parentWin, pWinRes->groupId, &pInfo->aggSup) && isCloseWindow(&parentWin, &pInfo->twAggSup)) {
+      continue;
+    }
+    setTimeWindowOutputBuf(&pInfo->binfo.resultRowInfo, &parentWin, true, &pCurResult, pWinRes->groupId, pSup->pCtx,
                            numOfOutput, pSup->rowEntryInfoOffset, &pInfo->aggSup, pTaskInfo);
     int32_t numOfChildren = taosArrayGetSize(pInfo->pChildren);
-    bool    find = true;
+    int32_t num = 0;
     for (int32_t j = 0; j < numOfChildren; j++) {
       SOperatorInfo*            pChildOp = taosArrayGetP(pInfo->pChildren, j);
       SIntervalAggOperatorInfo* pChInfo = pChildOp->info;
@@ -2895,15 +2921,16 @@ static void rebuildIntervalWindow(SStreamFinalIntervalOperatorInfo* pInfo, SExpr
       if (!hasIntervalWindow(&pChInfo->aggSup, pWinRes->ts, pWinRes->groupId)) {
         continue;
       }
-      find = true;
+      num++;
       SResultRow* pChResult = NULL;
-      setTimeWindowOutputBuf(&pChInfo->binfo.resultRowInfo, &ParentWin, true, &pChResult, pWinRes->groupId,
+      setTimeWindowOutputBuf(&pChInfo->binfo.resultRowInfo, &parentWin, true, &pChResult, pWinRes->groupId,
                              pChildSup->pCtx, pChildSup->numOfExprs, pChildSup->rowEntryInfoOffset, &pChInfo->aggSup,
                              pTaskInfo);
-      compactFunctions(pSup->pCtx, pChildSup->pCtx, numOfOutput, pTaskInfo);
+      updateTimeWindowInfo(&pInfo->twAggSup.timeWindowData, &parentWin, true);
+      compactFunctions(pSup->pCtx, pChildSup->pCtx, numOfOutput, pTaskInfo, &pInfo->twAggSup.timeWindowData);
     }
-    if (find && pUpdated) {
+    if (num > 1 && pUpdatedMap) {
-      saveResultRow(pCurResult, pWinRes->groupId, pUpdated);
+      saveWinResultRow(pCurResult, pWinRes->groupId, pUpdatedMap);
       setResultBufPageDirty(pInfo->aggSup.pResultBuf, &pInfo->binfo.resultRowInfo.cur);
     }
   }
@@ -2934,12 +2961,6 @@ void addPullWindow(SHashObj* pMap, SWinKey* pWinRes, int32_t size) {
 
 static int32_t getChildIndex(SSDataBlock* pBlock) { return pBlock->info.childId; }
 
-STimeWindow getFinalTimeWindow(int64_t ts, SInterval* pInterval) {
-  STimeWindow w = {.skey = ts, .ekey = INT64_MAX};
-  w.ekey = taosTimeAdd(w.skey, pInterval->interval, pInterval->intervalUnit, pInterval->precision) - 1;
-  return w;
-}
-
 static void doHashInterval(SOperatorInfo* pOperatorInfo, SSDataBlock* pSDataBlock, uint64_t tableGroupId,
                            SHashObj* pUpdatedMap) {
   SStreamFinalIntervalOperatorInfo* pInfo = (SStreamFinalIntervalOperatorInfo*)pOperatorInfo->info;
@@ -3126,6 +3147,31 @@ void processPullOver(SSDataBlock* pBlock, SHashObj* pMap) {
   }
 }
 
+static void addRetriveWindow(SArray* wins, SStreamFinalIntervalOperatorInfo* pInfo) {
+  int32_t size = taosArrayGetSize(wins);
+  for (int32_t i = 0; i < size; i++) {
+    SWinKey*    winKey = taosArrayGet(wins, i);
+    STimeWindow nextWin = getFinalTimeWindow(winKey->ts, &pInfo->interval);
+    if (isCloseWindow(&nextWin, &pInfo->twAggSup) && !pInfo->ignoreExpiredData) {
+      void* chIds = taosHashGet(pInfo->pPullDataMap, winKey, sizeof(SWinKey));
+      if (!chIds) {
+        SPullWindowInfo pull = {.window = nextWin, .groupId = winKey->groupId};
+        // add pull data request
+        savePullWindow(&pull, pInfo->pPullWins);
+        int32_t size = taosArrayGetSize(pInfo->pChildren);
+        addPullWindow(pInfo->pPullDataMap, winKey, size);
+        qDebug("===stream===prepare retrive for delete %" PRId64 ", size:%d", winKey->ts, size);
+      }
+    }
+  }
+}
+
+static void clearFunctionContext(SExprSupp* pSup) {
+  for (int32_t i = 0; i < pSup->numOfExprs; i++) {
+    pSup->pCtx[i].saveHandle.currentPage = -1;
+  }
+}
+
 static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
   SStreamFinalIntervalOperatorInfo* pInfo = pOperator->info;
 
@@ -3150,12 +3196,21 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
       return pInfo->pPullDataRes;
     }
 
+    doBuildDeleteResult(pInfo->pDelWins, &pInfo->delIndex, pInfo->pDelRes);
+    if (pInfo->pDelRes->info.rows != 0) {
+      // process the rest of the data
+      printDataBlock(pInfo->pDelRes, IS_FINAL_OP(pInfo) ? "interval final" : "interval semi");
+      return pInfo->pDelRes;
+    }
+
     doBuildResultDatablock(pOperator, &pInfo->binfo, &pInfo->groupResInfo, pInfo->aggSup.pResultBuf);
     if (pInfo->binfo.pRes->info.rows == 0) {
       pOperator->status = OP_EXEC_DONE;
       if (!IS_FINAL_OP(pInfo)) {
+        clearFunctionContext(&pOperator->exprSupp);
         // semi interval operator clear disk buffer
         clearStreamIntervalOperator(pInfo);
+        qDebug("===stream===clear semi operator");
       } else {
         freeAllPages(pInfo->pRecycledPages, pInfo->aggSup.pResultBuf);
       }
@@ -3219,23 +3274,28 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
       }
       removeResults(pUpWins, pUpdatedMap);
       copyDataBlock(pInfo->pUpdateRes, pBlock);
-      // copyUpdateDataBlock(pInfo->pUpdateRes, pBlock, pInfo->primaryTsIndex);
       pInfo->returnUpdate = true;
       taosArrayDestroy(pUpWins);
       break;
     } else if (pBlock->info.type == STREAM_DELETE_DATA || pBlock->info.type == STREAM_DELETE_RESULT) {
-      doDeleteSpecifyIntervalWindow(&pInfo->aggSup, pBlock, pInfo->pDelWins, &pInfo->interval);
+      SArray* delWins = taosArrayInit(8, sizeof(SWinKey));
+      doDeleteSpecifyIntervalWindow(&pInfo->aggSup, pBlock, delWins, &pInfo->interval, pUpdatedMap);
       if (IS_FINAL_OP(pInfo)) {
         int32_t                           childIndex = getChildIndex(pBlock);
         SOperatorInfo*                    pChildOp = taosArrayGetP(pInfo->pChildren, childIndex);
         SStreamFinalIntervalOperatorInfo* pChildInfo = pChildOp->info;
         SExprSupp*                        pChildSup = &pChildOp->exprSupp;
-        doDeleteSpecifyIntervalWindow(&pChildInfo->aggSup, pBlock, NULL, &pChildInfo->interval);
+        doDeleteSpecifyIntervalWindow(&pChildInfo->aggSup, pBlock, NULL, &pChildInfo->interval, NULL);
-        rebuildIntervalWindow(pInfo, pSup, pInfo->pDelWins, pInfo->binfo.pRes->info.groupId,
+        rebuildIntervalWindow(pInfo, pSup, delWins, pInfo->binfo.pRes->info.groupId, pOperator->exprSupp.numOfExprs,
-                              pOperator->exprSupp.numOfExprs, pOperator->pTaskInfo, pUpdated);
+                              pOperator->pTaskInfo, pUpdatedMap);
+        addRetriveWindow(delWins, pInfo);
+        taosArrayAddAll(pInfo->pDelWins, delWins);
+        taosArrayDestroy(delWins);
         continue;
       }
-      removeResults(pInfo->pDelWins, pUpdatedMap);
+      removeResults(delWins, pUpdatedMap);
+      taosArrayAddAll(pInfo->pDelWins, delWins);
+      taosArrayDestroy(delWins);
       break;
     } else if (pBlock->info.type == STREAM_GET_ALL && IS_FINAL_OP(pInfo)) {
       getAllIntervalWindow(pInfo->aggSup.pResultRowHashTable, pUpdatedMap);
@@ -3309,6 +3369,7 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
     return pInfo->pPullDataRes;
   }
 
+  // we should send result first.
   doBuildResultDatablock(pOperator, &pInfo->binfo, &pInfo->groupResInfo, pInfo->aggSup.pResultBuf);
   if (pInfo->binfo.pRes->info.rows != 0) {
     printDataBlock(pInfo->binfo.pRes, IS_FINAL_OP(pInfo) ? "interval final" : "interval semi");
@@ -3332,40 +3393,6 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
   return NULL;
 }
 
-SSDataBlock* createSpecialDataBlock(EStreamType type) {
-  SSDataBlock* pBlock = taosMemoryCalloc(1, sizeof(SSDataBlock));
-  pBlock->info.hasVarCol = false;
-  pBlock->info.groupId = 0;
-  pBlock->info.rows = 0;
-  pBlock->info.type = type;
-  pBlock->info.rowSize =
-      sizeof(TSKEY) + sizeof(TSKEY) + sizeof(uint64_t) + sizeof(uint64_t) + sizeof(TSKEY) + sizeof(TSKEY);
-  pBlock->info.watermark = INT64_MIN;
-
-  pBlock->pDataBlock = taosArrayInit(6, sizeof(SColumnInfoData));
-  SColumnInfoData infoData = {0};
-  infoData.info.type = TSDB_DATA_TYPE_TIMESTAMP;
-  infoData.info.bytes = sizeof(TSKEY);
-  // window start ts
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-  // window end ts
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-
-  infoData.info.type = TSDB_DATA_TYPE_UBIGINT;
-  infoData.info.bytes = sizeof(uint64_t);
-  // uid
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-  // group id
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-
-  // calculate start ts
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-  // calculate end ts
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-
-  return pBlock;
-}
-
 SOperatorInfo* createStreamFinalIntervalOperatorInfo(SOperatorInfo* downstream, SPhysiNode* pPhyNode,
                                                      SExecTaskInfo* pTaskInfo, int32_t numOfChild) {
   SIntervalPhysiNode*               pIntervalPhyNode = (SIntervalPhysiNode*)pPhyNode;
@@ -3470,7 +3497,7 @@ SOperatorInfo* createStreamFinalIntervalOperatorInfo(SOperatorInfo* downstream,
       createOperatorFpSet(NULL, doStreamFinalIntervalAgg, NULL, NULL, destroyStreamFinalIntervalOperatorInfo,
                           aggEncodeResultRow, aggDecodeResultRow, NULL);
   if (pPhyNode->type == QUERY_NODE_PHYSICAL_PLAN_STREAM_SEMI_INTERVAL) {
-    initIntervalDownStream(downstream, pPhyNode->type, &pInfo->aggSup);
+    initIntervalDownStream(downstream, pPhyNode->type, &pInfo->aggSup, &pInfo->interval, pInfo->twAggSup.waterMark);
   }
   code = appendDownstream(pOperator, &downstream, 1);
   if (code != TSDB_CODE_SUCCESS) {
@@ -3567,10 +3594,18 @@ void initDummyFunction(SqlFunctionCtx* pDummy, SqlFunctionCtx* pCtx, int32_t num
 }
 
 void initDownStream(SOperatorInfo* downstream, SStreamAggSupporter* pAggSup, int64_t gap, int64_t waterMark,
-                    uint16_t type) {
+                    uint16_t type, int32_t tsColIndex) {
-  ASSERT(downstream->operatorType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN);
+  if (downstream->operatorType == QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION) {
+    SStreamPartitionOperatorInfo* pScanInfo = downstream->info;
+    pScanInfo->tsColIndex = tsColIndex;
+  }
+
+  if (downstream->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) {
+    initDownStream(downstream->pDownstream[0], pAggSup, gap, waterMark, type, tsColIndex);
+    return;
+  }
   SStreamScanInfo* pScanInfo = downstream->info;
-  pScanInfo->sessionSup = (SessionWindowSupporter){.pStreamAggSup = pAggSup, .gap = gap, .parentType = type};
+  pScanInfo->windowSup = (SWindowSupporter){.pStreamAggSup = pAggSup, .gap = gap, .parentType = type};
   pScanInfo->pUpdateInfo = updateInfoInit(60000, TSDB_TIME_PRECISION_MILLI, waterMark);
 }
 
@@ -3642,7 +3677,6 @@ SOperatorInfo* createStreamSessionAggOperatorInfo(SOperatorInfo* downstream, SPh
   pInfo->isFinal = false;
   pInfo->pPhyNode = pPhyNode;
   pInfo->ignoreExpiredData = pSessionNode->window.igExpired;
-  pInfo->returnDelete = false;
 
   pOperator->name = "StreamSessionWindowAggOperator";
   pOperator->operatorType = QUERY_NODE_PHYSICAL_PLAN_STREAM_SESSION;
@@ -3653,7 +3687,8 @@ SOperatorInfo* createStreamSessionAggOperatorInfo(SOperatorInfo* downstream, SPh
       createOperatorFpSet(operatorDummyOpenFn, doStreamSessionAgg, NULL, NULL, destroyStreamSessionAggOperatorInfo,
                           aggEncodeResultRow, aggDecodeResultRow, NULL);
   if (downstream) {
-    initDownStream(downstream, &pInfo->streamAggSup, pInfo->gap, pInfo->twAggSup.waterMark, pOperator->operatorType);
+    initDownStream(downstream, &pInfo->streamAggSup, pInfo->gap, pInfo->twAggSup.waterMark, pOperator->operatorType,
+                   pInfo->primaryTsIndex);
     code = appendDownstream(pOperator, &downstream, 1);
   }
   return pOperator;
@@ -3683,13 +3718,15 @@ bool isInTimeWindow(STimeWindow* pWin, TSKEY ts, int64_t gap) {
 
 bool isInWindow(SResultWindowInfo* pWinInfo, TSKEY ts, int64_t gap) { return isInTimeWindow(&pWinInfo->win, ts, gap); }
 
-static SResultWindowInfo* insertNewSessionWindow(SArray* pWinInfos, TSKEY ts, int32_t index) {
+static SResultWindowInfo* insertNewSessionWindow(SArray* pWinInfos, TSKEY startTs, TSKEY endTs, int32_t index) {
-  SResultWindowInfo win = {.pos.offset = -1, .pos.pageId = -1, .win.skey = ts, .win.ekey = ts, .isOutput = false};
+  SResultWindowInfo win = {
+      .pos.offset = -1, .pos.pageId = -1, .win.skey = startTs, .win.ekey = endTs, .isOutput = false};
   return taosArrayInsert(pWinInfos, index, &win);
 }
 
-static SResultWindowInfo* addNewSessionWindow(SArray* pWinInfos, TSKEY ts) {
+static SResultWindowInfo* addNewSessionWindow(SArray* pWinInfos, TSKEY startTs, TSKEY endTs) {
-  SResultWindowInfo win = {.pos.offset = -1, .pos.pageId = -1, .win.skey = ts, .win.ekey = ts, .isOutput = false};
+  SResultWindowInfo win = {
+      .pos.offset = -1, .pos.pageId = -1, .win.skey = startTs, .win.ekey = endTs, .isOutput = false};
   return taosArrayPush(pWinInfos, &win);
 }
 
@@ -3748,7 +3785,7 @@ SResultWindowInfo* getSessionTimeWindow(SStreamAggSupporter* pAggSup, TSKEY star
   int32_t size = taosArrayGetSize(pWinInfos);
   if (size == 0) {
     *pIndex = 0;
-    return addNewSessionWindow(pWinInfos, startTs);
+    return addNewSessionWindow(pWinInfos, startTs, endTs);
   }
   // find the first position which is smaller than the key
   int32_t            index = binarySearch(pWinInfos, size, startTs, TSDB_ORDER_DESC, getSessionWindowEndkey);
@@ -3774,10 +3811,10 @@ SResultWindowInfo* getSessionTimeWindow(SStreamAggSupporter* pAggSup, TSKEY star
 
   if (index == size - 1) {
     *pIndex = taosArrayGetSize(pWinInfos);
-    return addNewSessionWindow(pWinInfos, startTs);
+    return addNewSessionWindow(pWinInfos, startTs, endTs);
   }
   *pIndex = index + 1;
-  return insertNewSessionWindow(pWinInfos, startTs, index + 1);
+  return insertNewSessionWindow(pWinInfos, startTs, endTs, index + 1);
 }
 
 int32_t updateSessionWindowInfo(SResultWindowInfo* pWinInfo, TSKEY* pStartTs, TSKEY* pEndTs, uint64_t groupId,
@@ -3789,7 +3826,7 @@ int32_t updateSessionWindowInfo(SResultWindowInfo* pWinInfo, TSKEY* pStartTs, TS
     if (pWinInfo->win.skey > pStartTs[i]) {
       if (pStDeleted && pWinInfo->isOutput) {
         SWinKey res = {.ts = pWinInfo->win.skey, .groupId = groupId};
-        taosHashPut(pStDeleted, &pWinInfo->pos, sizeof(SResultRowPosition), &res, sizeof(SWinKey));
+        taosHashPut(pStDeleted, &res, sizeof(SWinKey), &res, sizeof(SWinKey));
         pWinInfo->isOutput = false;
       }
       pWinInfo->win.skey = pStartTs[i];
@@ -3817,7 +3854,6 @@ static int32_t setWindowOutputBuf(SResultWindowInfo* pWinInfo, SResultRow** pRes
     if (*pResult == NULL) {
       return TSDB_CODE_OUT_OF_MEMORY;
     }
-    initResultRow(*pResult);
 
     // add a new result set for a new group
     pWinInfo->pos.pageId = (*pResult)->pageId;
@@ -3904,11 +3940,12 @@ void compactTimeWindow(SStreamSessionAggOperatorInfo* pInfo, int32_t startIndex,
     setWindowOutputBuf(pWinInfo, &pWinResult, pInfo->pDummyCtx, groupId, numOfOutput, pSup->rowEntryInfoOffset,
                        &pInfo->streamAggSup, pTaskInfo);
     pCurWin->win.ekey = TMAX(pCurWin->win.ekey, pWinInfo->win.ekey);
-    compactFunctions(pSup->pCtx, pInfo->pDummyCtx, numOfOutput, pTaskInfo);
+    updateTimeWindowInfo(&pInfo->twAggSup.timeWindowData, &pCurWin->win, true);
+    compactFunctions(pSup->pCtx, pInfo->pDummyCtx, numOfOutput, pTaskInfo, &pInfo->twAggSup.timeWindowData);
     taosHashRemove(pStUpdated, &pWinInfo->pos, sizeof(SResultRowPosition));
     if (pWinInfo->isOutput) {
       SWinKey res = {.ts = pWinInfo->win.skey, .groupId = groupId};
-      taosHashPut(pStDeleted, &pWinInfo->pos, sizeof(SResultRowPosition), &res, sizeof(SWinKey));
+      taosHashPut(pStDeleted, &res, sizeof(SWinKey), &res, sizeof(SWinKey));
       pWinInfo->isOutput = false;
     }
     taosArrayRemove(pInfo->streamAggSup.pCurWins, i);
@@ -4005,10 +4042,11 @@ static void doDeleteTimeWindows(SStreamAggSupporter* pAggSup, SSDataBlock* pBloc
       if (!pCurWin) {
         break;
       }
+      SResultWindowInfo delWin = *pCurWin;
       deleteWindow(pAggSup->pCurWins, winIndex, fp);
       if (result) {
-        pCurWin->groupId = gpDatas[i];
+        delWin.groupId = gpDatas[i];
-        taosArrayPush(result, pCurWin);
+        taosArrayPush(result, &delWin);
       }
     }
   }
@@ -4033,6 +4071,7 @@ static void doClearSessionWindows(SStreamAggSupporter* pAggSup, SExprSupp* pSup,
     ASSERT(isInWindow(pCurWin, tsCols[i], gap));
     doClearWindowImpl(&pCurWin->pos, pAggSup->pResultBuf, pSup, numOfOutput);
     if (result) {
+      pCurWin->groupId = gpCols[i];
       taosArrayPush(result, pCurWin);
     }
   }
@@ -4067,10 +4106,18 @@ void doBuildDeleteDataBlock(SHashObj* pStDeleted, SSDataBlock* pBlock, void** It
   size_t keyLen = 0;
   while (((*Ite) = taosHashIterate(pStDeleted, *Ite)) != NULL) {
     SWinKey*         res = *Ite;
-    SColumnInfoData* pTsCol = taosArrayGet(pBlock->pDataBlock, START_TS_COLUMN_INDEX);
+    SColumnInfoData* pStartTsCol = taosArrayGet(pBlock->pDataBlock, START_TS_COLUMN_INDEX);
-    colDataAppend(pTsCol, pBlock->info.rows, (const char*)&res->ts, false);
+    colDataAppend(pStartTsCol, pBlock->info.rows, (const char*)&res->ts, false);
+    SColumnInfoData* pEndTsCol = taosArrayGet(pBlock->pDataBlock, END_TS_COLUMN_INDEX);
+    colDataAppend(pEndTsCol, pBlock->info.rows, (const char*)&res->ts, false);
+    SColumnInfoData* pUidCol = taosArrayGet(pBlock->pDataBlock, UID_COLUMN_INDEX);
+    colDataAppendNULL(pUidCol, pBlock->info.rows);
     SColumnInfoData* pGpCol = taosArrayGet(pBlock->pDataBlock, GROUPID_COLUMN_INDEX);
     colDataAppend(pGpCol, pBlock->info.rows, (const char*)&res->groupId, false);
+    SColumnInfoData* pCalStCol = taosArrayGet(pBlock->pDataBlock, CALCULATE_START_TS_COLUMN_INDEX);
+    colDataAppendNULL(pCalStCol, pBlock->info.rows);
+    SColumnInfoData* pCalEdCol = taosArrayGet(pBlock->pDataBlock, CALCULATE_END_TS_COLUMN_INDEX);
+    colDataAppendNULL(pCalEdCol, pBlock->info.rows);
     pBlock->info.rows += 1;
     if (pBlock->info.rows + 1 >= pBlock->info.capacity) {
       break;
@@ -4081,8 +4128,8 @@ void doBuildDeleteDataBlock(SHashObj* pStDeleted, SSDataBlock* pBlock, void** It
   }
 }
 
-static void rebuildTimeWindow(SStreamSessionAggOperatorInfo* pInfo, SArray* pWinArray, int32_t groupId,
+static void rebuildTimeWindow(SStreamSessionAggOperatorInfo* pInfo, SArray* pWinArray, int32_t numOfOutput,
-                              int32_t numOfOutput, SOperatorInfo* pOperator) {
+                              SOperatorInfo* pOperator, SHashObj* pStUpdated, bool needCreate) {
   SExprSupp*     pSup = &pOperator->exprSupp;
   SExecTaskInfo* pTaskInfo = pOperator->pTaskInfo;
 
@@ -4092,9 +4139,16 @@ static void rebuildTimeWindow(SStreamSessionAggOperatorInfo* pInfo, SArray* pWin
   for (int32_t i = 0; i < size; i++) {
     SResultWindowInfo* pParentWin = taosArrayGet(pWinArray, i);
     SResultRow*        pCurResult = NULL;
+    uint64_t           groupId = pParentWin->groupId;
+    int32_t            winIndex = 0;
+    if (needCreate) {
+      pParentWin =
+          getSessionTimeWindow(&pInfo->streamAggSup, pParentWin->win.skey, pParentWin->win.ekey, groupId, 0, &winIndex);
+    }
     setWindowOutputBuf(pParentWin, &pCurResult, pSup->pCtx, groupId, numOfOutput, pSup->rowEntryInfoOffset,
                        &pInfo->streamAggSup, pTaskInfo);
     int32_t numOfChildren = taosArrayGetSize(pInfo->pChildren);
+    int32_t num = 0;
     for (int32_t j = 0; j < numOfChildren; j++) {
       SOperatorInfo*                 pChild = taosArrayGetP(pInfo->pChildren, j);
       SStreamSessionAggOperatorInfo* pChInfo = pChild->info;
@@ -4110,15 +4164,24 @@ static void rebuildTimeWindow(SStreamSessionAggOperatorInfo* pInfo, SArray* pWin
           SResultRow* pChResult = NULL;
           setWindowOutputBuf(pChWin, &pChResult, pChild->exprSupp.pCtx, groupId, numOfOutput,
                              pChild->exprSupp.rowEntryInfoOffset, &pChInfo->streamAggSup, pTaskInfo);
-          compactFunctions(pSup->pCtx, pChild->exprSupp.pCtx, numOfOutput, pTaskInfo);
+          updateTimeWindowInfo(&pInfo->twAggSup.timeWindowData, &pChWin->win, true);
+          compactFunctions(pSup->pCtx, pChild->exprSupp.pCtx, numOfOutput, pTaskInfo, &pInfo->twAggSup.timeWindowData);
           SFilePage* bufPage = getBufPage(pChInfo->streamAggSup.pResultBuf, pChWin->pos.pageId);
           releaseBufPage(pChInfo->streamAggSup.pResultBuf, bufPage);
+          num++;
           continue;
         } else if (!pChWin->isClosed) {
           break;
         }
       }
     }
+    if (num == 0 && needCreate) {
+      deleteWindow(pInfo->streamAggSup.pCurWins, winIndex, NULL);
+    }
+    if (pStUpdated && num > 0) {
+      SWinKey value = {.ts = pParentWin->win.skey, .groupId = groupId};
+      taosHashPut(pStUpdated, &pParentWin->pos, sizeof(SResultRowPosition), &value, sizeof(SWinKey));
+    }
     SFilePage* bufPage = getBufPage(pInfo->streamAggSup.pResultBuf, pParentWin->pos.pageId);
     ASSERT(size > 0);
     setBufPageDirty(bufPage, true);
@@ -4198,7 +4261,46 @@ static void copyDeleteWindowInfo(SArray* pResWins, SHashObj* pStDeleted) {
   for (int32_t i = 0; i < size; i++) {
     SResultWindowInfo* pWinInfo = taosArrayGet(pResWins, i);
     SWinKey            res = {.ts = pWinInfo->win.skey, .groupId = pWinInfo->groupId};
-    taosHashPut(pStDeleted, &pWinInfo->pos, sizeof(SResultRowPosition), &res, sizeof(SWinKey));
+    taosHashPut(pStDeleted, &res, sizeof(SWinKey), &res, sizeof(SWinKey));
+  }
+}
+
+static void removeSessionResults(SHashObj* pHashMap, SArray* pWins) {
+  int32_t size = taosArrayGetSize(pWins);
+  for (int32_t i = 0; i < size; i++) {
+    SResultWindowInfo* pWin = taosArrayGet(pWins, i);
+    taosHashRemove(pHashMap, &pWin->pos, sizeof(SResultRowPosition));
+  }
+}
+
+int32_t compareWinKey(void* pKey, void* data, int32_t index) {
+  SArray*     res = (SArray*)data;
+  SResKeyPos* pos = taosArrayGetP(res, index);
+  SWinKey*    pData = (SWinKey*)pKey;
+  if (pData->ts == *(int64_t*)pos->key) {
+    if (pData->groupId > pos->groupId) {
+      return 1;
+    } else if (pData->groupId < pos->groupId) {
+      return -1;
+    }
+    return 0;
+  } else if (pData->ts > *(int64_t*)pos->key) {
+    return 1;
+  }
+  return -1;
+}
+
+static void removeSessionDeleteResults(SArray* update, SHashObj* pStDeleted) {
+  int32_t size = taosHashGetSize(pStDeleted);
+  if (size == 0) {
+    return;
+  }
+
+  int32_t num = taosArrayGetSize(update);
+  for (int32_t i = 0; i < num; i++) {
+    SResKeyPos* pos = taosArrayGetP(update, i);
+    SWinKey     winKey = {.ts = *(int64_t*)pos->key, .groupId = pos->groupId};
+    taosHashRemove(pStDeleted, &winKey, sizeof(SWinKey));
   }
 }
 
@@ -4226,7 +4328,7 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator) {
   _hash_fn_t     hashFn = taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY);
   SHashObj*      pStUpdated = taosHashInit(64, hashFn, true, HASH_NO_LOCK);
   SOperatorInfo* downstream = pOperator->pDownstream[0];
-  SArray*        pUpdated = taosArrayInit(16, POINTER_BYTES);
+  SArray*        pUpdated = taosArrayInit(16, POINTER_BYTES);  // SResKeyPos
   while (1) {
     SSDataBlock* pBlock = downstream->fpSet.getNextFn(downstream);
     if (pBlock == NULL) {
@@ -4244,7 +4346,7 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator) {
         SStreamSessionAggOperatorInfo* pChildInfo = pChildOp->info;
         doClearSessionWindows(&pChildInfo->streamAggSup, &pChildOp->exprSupp, pBlock, START_TS_COLUMN_INDEX,
                               pChildOp->exprSupp.numOfExprs, 0, NULL);
-        rebuildTimeWindow(pInfo, pWins, pBlock->info.groupId, pOperator->exprSupp.numOfExprs, pOperator);
+        rebuildTimeWindow(pInfo, pWins, pOperator->exprSupp.numOfExprs, pOperator, NULL, false);
       }
       taosArrayDestroy(pWins);
       continue;
@@ -4258,9 +4360,10 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator) {
         SStreamSessionAggOperatorInfo* pChildInfo = pChildOp->info;
         // gap must be 0
         doDeleteTimeWindows(&pChildInfo->streamAggSup, pBlock, 0, NULL, NULL);
-        rebuildTimeWindow(pInfo, pWins, pBlock->info.groupId, pOperator->exprSupp.numOfExprs, pOperator);
+        rebuildTimeWindow(pInfo, pWins, pOperator->exprSupp.numOfExprs, pOperator, pStUpdated, true);
       }
       copyDeleteWindowInfo(pWins, pInfo->pStDeleted);
+      removeSessionResults(pStUpdated, pWins);
       taosArrayDestroy(pWins);
       continue;
     } else if (pBlock->info.type == STREAM_GET_ALL) {
@@ -4303,6 +4406,7 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator) {
                      pInfo->ignoreExpiredData, NULL);
   closeChildSessionWindow(pInfo->pChildren, pInfo->twAggSup.maxTs, pInfo->ignoreExpiredData, NULL);
   copyUpdateResult(pStUpdated, pUpdated);
+  removeSessionDeleteResults(pUpdated, pInfo->pStDeleted);
   taosHashCleanup(pStUpdated);
 
   finalizeUpdatedResult(pSup->numOfExprs, pInfo->streamAggSup.pResultBuf, pUpdated, pSup->rowEntryInfoOffset);
@@ -4333,14 +4437,6 @@ static void clearStreamSessionOperator(SStreamSessionAggOperatorInfo* pInfo) {
   pInfo->streamAggSup.currentPageId = -1;
 }
 
-static void removeSessionResults(SHashObj* pHashMap, SArray* pWins) {
-  int32_t size = taosArrayGetSize(pWins);
-  for (int32_t i = 0; i < size; i++) {
-    SResultWindowInfo* pWin = taosArrayGet(pWins, i);
-    taosHashRemove(pHashMap, &pWin->pos, sizeof(SResultRowPosition));
-  }
-}
-
 static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
   SStreamSessionAggOperatorInfo* pInfo = pOperator->info;
   SOptrBasicInfo*                pBInfo = &pInfo->binfo;
@@ -4349,31 +4445,36 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
 
   if (pOperator->status == OP_EXEC_DONE) {
     return NULL;
-  } else if (pOperator->status == OP_RES_TO_RETURN) {
+  }
+
+  {
     doBuildResultDatablock(pOperator, pBInfo, &pInfo->groupResInfo, pInfo->streamAggSup.pResultBuf);
     if (pBInfo->pRes->info.rows > 0) {
       printDataBlock(pBInfo->pRes, "semi session");
       return pBInfo->pRes;
     }
 
-    // doBuildDeleteDataBlock(pInfo->pStDeleted, pInfo->pDelRes, &pInfo->pDelIterator);
+    doBuildDeleteDataBlock(pInfo->pStDeleted, pInfo->pDelRes, &pInfo->pDelIterator);
-    if (pInfo->pDelRes->info.rows > 0 && !pInfo->returnDelete) {
+    if (pInfo->pDelRes->info.rows > 0) {
-      pInfo->returnDelete = true;
       printDataBlock(pInfo->pDelRes, "semi session");
       return pInfo->pDelRes;
     }
 
-    if (pInfo->pUpdateRes->info.rows > 0) {
+    if (pInfo->pUpdateRes->info.rows > 0 && pInfo->returnUpdate) {
+      pInfo->returnUpdate = false;
       // process the rest of the data
-      pOperator->status = OP_OPENED;
       printDataBlock(pInfo->pUpdateRes, "semi session");
       return pInfo->pUpdateRes;
     }
+
+    if (pOperator->status == OP_RES_TO_RETURN) {
+      clearFunctionContext(&pOperator->exprSupp);
       // semi interval operator clear disk buffer
       clearStreamSessionOperator(pInfo);
       pOperator->status = OP_EXEC_DONE;
       return NULL;
     }
+  }
 
   _hash_fn_t     hashFn = taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY);
   SHashObj*      pStUpdated = taosHashInit(64, hashFn, true, HASH_NO_LOCK);
@@ -4383,6 +4484,7 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
     SSDataBlock* pBlock = downstream->fpSet.getNextFn(downstream);
     if (pBlock == NULL) {
       clearSpecialDataBlock(pInfo->pUpdateRes);
+      pOperator->status = OP_RES_TO_RETURN;
       break;
     }
     printDataBlock(pBlock, "semi session recv");
@@ -4393,12 +4495,15 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
       removeSessionResults(pStUpdated, pWins);
       taosArrayDestroy(pWins);
       copyDataBlock(pInfo->pUpdateRes, pBlock);
+      pInfo->returnUpdate = true;
       break;
     } else if (pBlock->info.type == STREAM_DELETE_DATA || pBlock->info.type == STREAM_DELETE_RESULT) {
       // gap must be 0
-      doDeleteTimeWindows(&pInfo->streamAggSup, pBlock, 0, NULL, NULL);
+      SArray* pWins = taosArrayInit(16, sizeof(SResultWindowInfo));
-      copyDataBlock(pInfo->pDelRes, pBlock);
+      doDeleteTimeWindows(&pInfo->streamAggSup, pBlock, 0, pWins, NULL);
-      pInfo->pDelRes->info.type = STREAM_DELETE_RESULT;
+      copyDeleteWindowInfo(pWins, pInfo->pStDeleted);
+      removeSessionResults(pStUpdated, pWins);
+      taosArrayDestroy(pWins);
       break;
     } else if (pBlock->info.type == STREAM_GET_ALL) {
       getAllSessionWindow(pInfo->streamAggSup.pResultRows, pUpdated, getResWinForSession);
@@ -4411,18 +4516,15 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
     }
     // the pDataBlock are always the same one, no need to call this again
     setInputDataBlock(pOperator, pSup->pCtx, pBlock, TSDB_ORDER_ASC, MAIN_SCAN, true);
-    doStreamSessionAggImpl(pOperator, pBlock, pStUpdated, pInfo->pStDeleted, false);
+    doStreamSessionAggImpl(pOperator, pBlock, pStUpdated, NULL, false);
     maxTs = TMAX(pInfo->twAggSup.maxTs, pBlock->info.window.ekey);
   }
 
   pInfo->twAggSup.maxTs = TMAX(pInfo->twAggSup.maxTs, maxTs);
   pBInfo->pRes->info.watermark = pInfo->twAggSup.maxTs;
-  // restore the value
+
-  pOperator->status = OP_RES_TO_RETURN;
-  // semi operator
-  // closeSessionWindow(pInfo->streamAggSup.pResultRows, &pInfo->twAggSup, pUpdated,
-  //                    getResWinForSession);
   copyUpdateResult(pStUpdated, pUpdated);
+  removeSessionDeleteResults(pUpdated, pInfo->pStDeleted);
   taosHashCleanup(pStUpdated);
 
   finalizeUpdatedResult(pOperator->exprSupp.numOfExprs, pInfo->streamAggSup.pResultBuf, pUpdated,
@@ -4436,16 +4538,15 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
     return pBInfo->pRes;
   }
 
-  // doBuildDeleteDataBlock(pInfo->pStDeleted, pInfo->pDelRes, &pInfo->pDelIterator);
+  doBuildDeleteDataBlock(pInfo->pStDeleted, pInfo->pDelRes, &pInfo->pDelIterator);
-  if (pInfo->pDelRes->info.rows > 0 && !pInfo->returnDelete) {
+  if (pInfo->pDelRes->info.rows > 0) {
-    pInfo->returnDelete = true;
     printDataBlock(pInfo->pDelRes, "semi session");
     return pInfo->pDelRes;
   }
 
-  if (pInfo->pUpdateRes->info.rows > 0) {
+  if (pInfo->pUpdateRes->info.rows > 0 && pInfo->returnUpdate) {
+    pInfo->returnUpdate = false;
     // process the rest of the data
-    pOperator->status = OP_OPENED;
     printDataBlock(pInfo->pUpdateRes, "semi session");
     return pInfo->pUpdateRes;
   }
@@ -4669,7 +4770,7 @@ int32_t updateStateWindowInfo(SArray* pWinInfos, int32_t winIndex, TSKEY* pTs, u
     if (pWinInfo->winInfo.win.skey > pTs[i]) {
       if (pSeDeleted && pWinInfo->winInfo.isOutput) {
         SWinKey res = {.ts = pWinInfo->winInfo.win.skey, .groupId = groupId};
-        taosHashPut(pSeDeleted, &pWinInfo->winInfo.pos, sizeof(SResultRowPosition), &res, sizeof(SWinKey));
+        taosHashPut(pSeDeleted, &res, sizeof(SWinKey), &res, sizeof(SWinKey));
         pWinInfo->winInfo.isOutput = false;
       }
       pWinInfo->winInfo.win.skey = pTs[i];
@@ -4737,8 +4838,8 @@ static void doStreamStateAggImpl(SOperatorInfo* pOperator, SSDataBlock* pSDataBl
     winRows = updateStateWindowInfo(pAggSup->pCurWins, winIndex, tsCols, groupId, pKeyColInfo, pSDataBlock->info.rows,
                                     i, &allEqual, pStDeleted);
     if (!allEqual) {
-      appendOneRow(pAggSup->pScanBlock, &pCurWin->winInfo.win.skey, &pCurWin->winInfo.win.ekey, GROUPID_COLUMN_INDEX,
+      uint64_t uid = 0;
-                   &groupId);
+      appendOneRow(pAggSup->pScanBlock, &pCurWin->winInfo.win.skey, &pCurWin->winInfo.win.ekey, &uid, &groupId);
       taosHashRemove(pSeUpdated, &pCurWin->winInfo.pos, sizeof(SResultRowPosition));
       deleteWindow(pAggSup->pCurWins, winIndex, destroyStateWinInfo);
       continue;
@@ -4767,6 +4868,7 @@ static SSDataBlock* doStreamStateAgg(SOperatorInfo* pOperator) {
   SExprSupp*                   pSup = &pOperator->exprSupp;
   SStreamStateAggOperatorInfo* pInfo = pOperator->info;
   SOptrBasicInfo*              pBInfo = &pInfo->binfo;
+  int64_t                      maxTs = INT64_MIN;
   if (pOperator->status == OP_RES_TO_RETURN) {
     doBuildDeleteDataBlock(pInfo->pSeDeleted, pInfo->pDelRes, &pInfo->pDelIterator);
     if (pInfo->pDelRes->info.rows > 0) {
@@ -4799,6 +4901,7 @@ static SSDataBlock* doStreamStateAgg(SOperatorInfo* pOperator) {
       SArray* pWins = taosArrayInit(16, sizeof(SResultWindowInfo));
       doDeleteTimeWindows(&pInfo->streamAggSup, pBlock, 0, pWins, destroyStateWinInfo);
       copyDeleteWindowInfo(pWins, pInfo->pSeDeleted);
+      removeSessionResults(pSeUpdated, pWins);
       taosArrayDestroy(pWins);
       continue;
     } else if (pBlock->info.type == STREAM_GET_ALL) {
@@ -4813,8 +4916,9 @@ static SSDataBlock* doStreamStateAgg(SOperatorInfo* pOperator) {
     // the pDataBlock are always the same one, no need to call this again
     setInputDataBlock(pOperator, pSup->pCtx, pBlock, TSDB_ORDER_ASC, MAIN_SCAN, true);
     doStreamStateAggImpl(pOperator, pBlock, pSeUpdated, pInfo->pSeDeleted);
-    pInfo->twAggSup.maxTs = TMAX(pInfo->twAggSup.maxTs, pBlock->info.window.ekey);
+    maxTs = TMAX(maxTs, pBlock->info.window.ekey);
   }
+  pInfo->twAggSup.maxTs = TMAX(pInfo->twAggSup.maxTs, maxTs);
   // restore the value
   pOperator->status = OP_RES_TO_RETURN;
 
@@ -4913,7 +5017,8 @@ SOperatorInfo* createStreamStateAggOperatorInfo(SOperatorInfo* downstream, SPhys
   pOperator->info = pInfo;
   pOperator->fpSet = createOperatorFpSet(operatorDummyOpenFn, doStreamStateAgg, NULL, NULL,
                                          destroyStreamStateOperatorInfo, aggEncodeResultRow, aggDecodeResultRow, NULL);
-  initDownStream(downstream, &pInfo->streamAggSup, 0, pInfo->twAggSup.waterMark, pOperator->operatorType);
+  initDownStream(downstream, &pInfo->streamAggSup, 0, pInfo->twAggSup.waterMark, pOperator->operatorType,
+                 pInfo->primaryTsIndex);
   code = appendDownstream(pOperator, &downstream, 1);
   if (code != TSDB_CODE_SUCCESS) {
     goto _error;
@@ -4942,8 +5047,8 @@ static int32_t outputMergeAlignedIntervalResult(SOperatorInfo* pOperatorInfo, ui
   SExprSupp*                pSup = &pOperatorInfo->exprSupp;
 
   SET_RES_WINDOW_KEY(iaInfo->aggSup.keyBuf, &wstartTs, TSDB_KEYSIZE, tableGroupId);
-  SResultRowPosition* p1 = (SResultRowPosition*)tSimpleHashGet(iaInfo->aggSup.pResultRowHashTable, iaInfo->aggSup.keyBuf,
+  SResultRowPosition* p1 = (SResultRowPosition*)tSimpleHashGet(
-                                                            GET_RES_WINDOW_KEY_LEN(TSDB_KEYSIZE));
+      iaInfo->aggSup.pResultRowHashTable, iaInfo->aggSup.keyBuf, GET_RES_WINDOW_KEY_LEN(TSDB_KEYSIZE));
   ASSERT(p1 != NULL);
 
   finalizeResultRowIntoResultDataBlock(iaInfo->aggSup.pResultBuf, p1, pSup->pCtx, pSup->pExprInfo, pSup->numOfExprs,
@@ -5237,8 +5342,8 @@ static int32_t finalizeWindowResult(SOperatorInfo* pOperatorInfo, uint64_t table
   SExprSupp*                     pExprSup = &pOperatorInfo->exprSupp;
 
   SET_RES_WINDOW_KEY(iaInfo->aggSup.keyBuf, &win->skey, TSDB_KEYSIZE, tableGroupId);
-  SResultRowPosition* p1 = (SResultRowPosition*)tSimpleHashGet(iaInfo->aggSup.pResultRowHashTable, iaInfo->aggSup.keyBuf,
+  SResultRowPosition* p1 = (SResultRowPosition*)tSimpleHashGet(
-                                                            GET_RES_WINDOW_KEY_LEN(TSDB_KEYSIZE));
+      iaInfo->aggSup.pResultRowHashTable, iaInfo->aggSup.keyBuf, GET_RES_WINDOW_KEY_LEN(TSDB_KEYSIZE));
   ASSERT(p1 != NULL);
   finalizeResultRowIntoResultDataBlock(iaInfo->aggSup.pResultBuf, p1, pExprSup->pCtx, pExprSup->pExprInfo,
                                        pExprSup->numOfExprs, pExprSup->rowEntryInfoOffset, pResultBlock, pTaskInfo);
@@ -5483,7 +5588,6 @@ SOperatorInfo* createMergeIntervalOperatorInfo(SOperatorInfo* downstream, SMerge
   initBasicInfo(&pIntervalInfo->binfo, pResBlock);
   initExecTimeWindowInfo(&pIntervalInfo->twAggSup.timeWindowData, &pIntervalInfo->win);
 
-
   pIntervalInfo->timeWindowInterpo = timeWindowinterpNeeded(pExprSupp->pCtx, num, pIntervalInfo);
   if (pIntervalInfo->timeWindowInterpo) {
     pIntervalInfo->binfo.resultRowInfo.openWindow = tdListNew(sizeof(SOpenWindowInfo));

source/libs/function/src/builtinsimpl.c

@@ -1357,8 +1357,7 @@ int32_t doMinMaxHelper(SqlFunctionCtx* pCtx, int32_t isMinFunc) {
 
         numOfElems += 1;
       }
-    } else if (type == TSDB_DATA_TYPE_BIGINT ||
+    } else if (type == TSDB_DATA_TYPE_BIGINT || type == TSDB_DATA_TYPE_TIMESTAMP) {
-               type == TSDB_DATA_TYPE_TIMESTAMP) {
       int64_t* pData = (int64_t*)pCol->pData;
       int64_t* val = (int64_t*)&pBuf->v;
 
@@ -1601,7 +1600,8 @@ int32_t maxFunction(SqlFunctionCtx* pCtx) {
 }
 
 static void setNullSelectivityValue(SqlFunctionCtx* pCtx, SSDataBlock* pBlock, int32_t rowIndex);
-static void setSelectivityValue(SqlFunctionCtx* pCtx, SSDataBlock* pBlock, const STuplePos* pTuplePos, int32_t rowIndex);
+static void setSelectivityValue(SqlFunctionCtx* pCtx, SSDataBlock* pBlock, const STuplePos* pTuplePos,
+                                int32_t rowIndex);
 
 int32_t minmaxFunctionFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) {
   SResultRowEntryInfo* pEntryInfo = GET_RES_INFO(pCtx);
@@ -1712,7 +1712,6 @@ void appendSelectivityValue(SqlFunctionCtx* pCtx, int32_t rowIndex, int32_t pos)
       colDataAppend(pDstCol, pos, pData, false);
     }
   }
-
 }
 
 void replaceTupleData(STuplePos* pDestPos, STuplePos* pSourcePos) {
@@ -2590,8 +2589,8 @@ static void apercentileTransferInfo(SAPercentileInfo* pInput, SAPercentileInfo*
       memcpy(pHisto, pInput->pHisto, sizeof(SHistogramInfo) + sizeof(SHistBin) * (MAX_HISTOGRAM_BIN + 1));
       pHisto->elems = (SHistBin*)((char*)pHisto + sizeof(SHistogramInfo));
 
-      qDebug("%s merge histo, total:%" PRId64 ", entry:%d, %p", __FUNCTION__, pHisto->numOfElems,
+      qDebug("%s merge histo, total:%" PRId64 ", entry:%d, %p", __FUNCTION__, pHisto->numOfElems, pHisto->numOfEntries,
-              pHisto->numOfEntries, pHisto);
+             pHisto);
     } else {
       pHisto->elems = (SHistBin*)((char*)pHisto + sizeof(SHistogramInfo));
       qDebug("%s input histogram, elem:%" PRId64 ", entry:%d, %p", __FUNCTION__, pHisto->numOfElems,
@@ -2601,8 +2600,8 @@ static void apercentileTransferInfo(SAPercentileInfo* pInput, SAPercentileInfo*
       memcpy(pHisto, pRes, sizeof(SHistogramInfo) + sizeof(SHistBin) * MAX_HISTOGRAM_BIN);
       pHisto->elems = (SHistBin*)((char*)pHisto + sizeof(SHistogramInfo));
 
-      qDebug("%s merge histo, total:%" PRId64 ", entry:%d, %p", __FUNCTION__, pHisto->numOfElems,
+      qDebug("%s merge histo, total:%" PRId64 ", entry:%d, %p", __FUNCTION__, pHisto->numOfElems, pHisto->numOfEntries,
-             pHisto->numOfEntries, pHisto);
+             pHisto);
       tHistogramDestroy(&pRes);
     }
   }
@@ -2629,8 +2628,8 @@ int32_t apercentileFunctionMerge(SqlFunctionCtx* pCtx) {
   }
 
   if (pInfo->algo != APERCT_ALGO_TDIGEST) {
-    qDebug("%s after merge, total:%d, numOfEntry:%d, %p", __FUNCTION__, pInfo->pHisto->numOfElems, pInfo->pHisto->numOfEntries,
+    qDebug("%s after merge, total:%d, numOfEntry:%d, %p", __FUNCTION__, pInfo->pHisto->numOfElems,
-           pInfo->pHisto);
+           pInfo->pHisto->numOfEntries, pInfo->pHisto);
   }
 
   SET_VAL(pResInfo, 1, 1);
@@ -2752,7 +2751,8 @@ static FORCE_INLINE TSKEY getRowPTs(SColumnInfoData* pTsColInfo, int32_t rowInde
   return *(TSKEY*)colDataGetData(pTsColInfo, rowIndex);
 }
 
-static void firstlastSaveTupleData(const SSDataBlock* pSrcBlock, int32_t rowIndex, SqlFunctionCtx* pCtx, SFirstLastRes* pInfo) {
+static void firstlastSaveTupleData(const SSDataBlock* pSrcBlock, int32_t rowIndex, SqlFunctionCtx* pCtx,
+                                   SFirstLastRes* pInfo) {
   if (pCtx->subsidiaries.num <= 0) {
     return;
   }
@@ -3537,7 +3537,8 @@ void doAddIntoResult(SqlFunctionCtx* pCtx, void* pData, int32_t rowIndex, SSData
  * |(n columns, one bit for each column)| src column #1| src column #2|
  * +------------------------------------+--------------+--------------+
  */
-void* serializeTupleData(const SSDataBlock* pSrcBlock, int32_t rowIndex, SSubsidiaryResInfo* pSubsidiaryies, char* buf) {
+void* serializeTupleData(const SSDataBlock* pSrcBlock, int32_t rowIndex, SSubsidiaryResInfo* pSubsidiaryies,
+                         char* buf) {
   char* nullList = buf;
   char* pStart = (char*)(nullList + sizeof(bool) * pSubsidiaryies->num);
 
@@ -3621,7 +3622,6 @@ static int32_t doUpdateTupleData(SSerializeDataHandle* pHandle, const void* pBuf
     setBufPageDirty(pPage, true);
     releaseBufPage(pHandle->pBuf, pPage);
   } else {
-
   }
 
   return TSDB_CODE_SUCCESS;
@@ -3980,8 +3980,8 @@ int32_t elapsedFunction(SqlFunctionCtx* pCtx) {
       }
 
       if (pCtx->end.key == INT64_MIN) {
-        pInfo->min = (pInfo->min > ptsList[start + pInput->numOfRows - 1]) ?
+        pInfo->min =
-                     ptsList[start + pInput->numOfRows - 1] : pInfo->min;
+            (pInfo->min > ptsList[start + pInput->numOfRows - 1]) ? ptsList[start + pInput->numOfRows - 1] : pInfo->min;
       } else {
         pInfo->min = pCtx->end.key;
       }
@@ -3993,8 +3993,8 @@ int32_t elapsedFunction(SqlFunctionCtx* pCtx) {
       }
 
       if (pCtx->end.key == INT64_MIN) {
-        pInfo->max = (pInfo->max < ptsList[start + pInput->numOfRows - 1]) ?
+        pInfo->max =
-                     ptsList[start + pInput->numOfRows - 1] : pInfo->max;
+            (pInfo->max < ptsList[start + pInput->numOfRows - 1]) ? ptsList[start + pInput->numOfRows - 1] : pInfo->max;
       } else {
         pInfo->max = pCtx->end.key + 1;
       }

source/libs/index/inc/indexInt.h

@@ -40,26 +40,31 @@ extern "C" {
 #define indexTrace(...) do { if (idxDebugFlag & DEBUG_TRACE) { taosPrintLog("IDX", DEBUG_TRACE, idxDebugFlag, __VA_ARGS__);} } while (0)
 // clang-format on
 
+extern void* indexQhandle;
+
 typedef enum { LT, LE, GT, GE, CONTAINS, EQ } RangeType;
 typedef enum { kTypeValue, kTypeDeletion } STermValueType;
 typedef enum { kRebuild, kFinished } SIdxStatus;
 
 typedef struct SIndexStat {
-  int32_t totalAdded;    //
+  int32_t total;
-  int32_t totalDeled;    //
+  int32_t add;      //
-  int32_t totalUpdated;  //
+  int32_t del;      //
-  int32_t totalTerms;    //
+  int32_t update;   //
-  int32_t distinctCol;   // distinct column
+  int32_t terms;    //
+  int32_t distCol;  // distinct column
 } SIndexStat;
 
 struct SIndex {
+  SIndexOpts opts;
+
   int64_t   refId;
   void*     cache;
   void*     tindex;
   SHashObj* colObj;  // < field name, field id>
 
   int64_t    suid;     // current super table id, -1 is normal table
-  int32_t    cVersion;  // current version allocated to cache
+  int32_t    version;  // current version allocated to cache
   SLRUCache* lru;
   char*      path;
 
@@ -68,7 +73,6 @@ struct SIndex {
   TdThreadMutex mtx;
   tsem_t        sem;
   bool          quit;
-  SIndexOpts    opts;
 };
 
 struct SIndexMultiTermQuery {
@@ -111,14 +115,15 @@ typedef struct Iterate {
 
 void iterateValueDestroy(IterateValue* iv, bool destroy);
 
-extern void* indexQhandle;
-
 typedef struct TFileCacheKey {
   uint64_t suid;
   uint8_t  colType;
   char*    colName;
   int32_t  nColName;
 } ICacheKey;
+
+int32_t idxSerialCacheKey(ICacheKey* key, char* buf);
+
 int idxFlushCacheToTFile(SIndex* sIdx, void*, bool quit);
 
 int64_t idxAddRef(void* p);
@@ -126,10 +131,6 @@ int32_t idxRemoveRef(int64_t ref);
 void    idxAcquireRef(int64_t ref);
 void    idxReleaseRef(int64_t ref);
 
-int32_t idxSerialCacheKey(ICacheKey* key, char* buf);
-// int32_t indexSerialKey(ICacheKey* key, char* buf);
-// int32_t indexSerialTermKey(SIndexTerm* itm, char* buf);
-
 #define IDX_TYPE_CONTAIN_EXTERN_TYPE(ty, exTy) (((ty >> 4) & (exTy)) != 0)
 
 #define IDX_TYPE_GET_TYPE(ty) (ty & 0x0F)

source/libs/index/src/index.c

@@ -25,10 +25,6 @@
 #include "tref.h"
 #include "tsched.h"
 
-#ifdef USE_LUCENE
-#include "lucene++/Lucene_c.h"
-#endif
-
 #define INDEX_NUM_OF_THREADS 5
 #define INDEX_QUEUE_SIZE     200
 
@@ -74,7 +70,7 @@ void indexCleanup() {
 
 typedef struct SIdxColInfo {
   int colId;  // generated by index internal
-  int cVersion;
+  int version;
 } SIdxColInfo;
 
 static TdThreadOnce isInit = PTHREAD_ONCE_INIT;
@@ -123,7 +119,7 @@ int indexOpen(SIndexOpts* opts, const char* path, SIndex** index) {
   }
 
   idx->colObj = taosHashInit(8, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY), true, HASH_ENTRY_LOCK);
-  idx->cVersion = 1;
+  idx->version = 1;
   idx->path = tstrdup(path);
   taosThreadMutexInit(&idx->mtx, NULL);
   tsem_init(&idx->sem, 0, 0);

source/libs/index/src/indexCache.c

@@ -566,7 +566,6 @@ int idxCachePut(void* cache, SIndexTerm* term, uint64_t uid) {
   taosThreadMutexUnlock(&pCache->mtx);
   idxCacheUnRef(pCache);
   return 0;
-  // encode end
 }
 void idxCacheForceToMerge(void* cache) {
   IndexCache* pCache = cache;
@@ -602,10 +601,10 @@ static int32_t idxQueryMem(MemTable* mem, SIndexTermQuery* query, SIdxTRslt* tr,
   }
 }
 int idxCacheSearch(void* cache, SIndexTermQuery* query, SIdxTRslt* result, STermValueType* s) {
-  int64_t st = taosGetTimestampUs();
   if (cache == NULL) {
     return 0;
   }
+
   IndexCache* pCache = cache;
 
   MemTable *mem = NULL, *imm = NULL;
@@ -616,6 +615,8 @@ int idxCacheSearch(void* cache, SIndexTermQuery* query, SIdxTRslt* result, STerm
   idxMemRef(imm);
   taosThreadMutexUnlock(&pCache->mtx);
 
+  int64_t st = taosGetTimestampUs();
+
   int ret = (mem && mem->mem) ? idxQueryMem(mem, query, result, s) : 0;
   if (ret == 0 && *s != kTypeDeletion) {
     // continue search in imm

source/libs/index/src/indexComm.c

@@ -178,9 +178,9 @@ TExeCond tDoCompare(__compar_fn_t func, int8_t comparType, void* a, void* b) {
   // optime later
   int32_t ret = func(a, b);
   switch (comparType) {
-    case QUERY_LESS_THAN: {
+    case QUERY_LESS_THAN:
       if (ret < 0) return MATCH;
-    } break;
+      break;
     case QUERY_LESS_EQUAL: {
       if (ret <= 0) return MATCH;
       break;

source/libs/index/src/indexFilter.c

@@ -27,6 +27,44 @@
 #define SIF_RET(c) do { int32_t _code = c; if (_code != TSDB_CODE_SUCCESS) { terrno = _code; } return _code; } while (0)
 #define SIF_ERR_JRET(c) do { code = c; if (code != TSDB_CODE_SUCCESS) { terrno = code; goto _return; } } while (0)
 // clang-format on
+
+typedef union {
+  uint8_t  u8;
+  uint16_t u16;
+  uint32_t u32;
+  uint64_t u64;
+
+  int8_t  i8;
+  int16_t i16;
+  int32_t i32;
+  int64_t i64;
+
+  double d;
+  float  f;
+} SDataTypeBuf;
+
+#define SIF_DATA_CONVERT(type, val, dst)       \
+  do {                                         \
+    if (type == TSDB_DATA_TYPE_DOUBLE)         \
+      dst = GET_DOUBLE_VAL(val);               \
+    else if (type == TSDB_DATA_TYPE_BIGINT)    \
+      dst = *(int64_t *)val;                   \
+    else if (type == TSDB_DATA_TYPE_INT)       \
+      dst = *(int32_t *)val;                   \
+    else if (type == TSDB_DATA_TYPE_SMALLINT)  \
+      dst = *(int16_t *)val;                   \
+    else if (type == TSDB_DATA_TYPE_TINYINT)   \
+      dst = *(int8_t *)val;                    \
+    else if (type == TSDB_DATA_TYPE_UTINYINT)  \
+      dst = *(uint8_t *)val;                   \
+    else if (type == TSDB_DATA_TYPE_USMALLINT) \
+      dst = *(uint16_t *)val;                  \
+    else if (type == TSDB_DATA_TYPE_UINT)      \
+      dst = *(uint32_t *)val;                  \
+    else if (type == TSDB_DATA_TYPE_UBIGINT)   \
+      dst = *(uint64_t *)val;                  \
+  } while (0);
+
 typedef struct SIFParam {
   SHashObj *pFilter;
 
@@ -48,7 +86,6 @@ typedef struct SIFCtx {
   SHashObj     *pRes;    /* element is SIFParam */
   bool          noExec;  // true: just iterate condition tree, and add hint to executor plan
   SIndexMetaArg arg;
-  // SIdxFltStatus st;
 } SIFCtx;
 
 static int32_t sifGetFuncFromSql(EOperatorType src, EIndexQueryType *dst) {
@@ -75,11 +112,6 @@ static int32_t sifGetFuncFromSql(EOperatorType src, EIndexQueryType *dst) {
 typedef int32_t (*sif_func_t)(SIFParam *left, SIFParam *rigth, SIFParam *output);
 
 static sif_func_t sifNullFunc = NULL;
-// typedef struct SIFWalkParm
-// construct tag filter operator later
-// static void destroyTagFilterOperatorInfo(void *param) {
-//  STagFilterOperatorInfo *pInfo = (STagFilterOperatorInfo *)param;
-//}
 
 static void sifFreeParam(SIFParam *param) {
   if (param == NULL) return;
@@ -365,42 +397,6 @@ static Filter sifGetFilterFunc(EIndexQueryType type, bool *reverse) {
   }
   return NULL;
 }
-typedef union {
-  uint8_t  u8;
-  uint16_t u16;
-  uint32_t u32;
-  uint64_t u64;
-
-  int8_t  i8;
-  int16_t i16;
-  int32_t i32;
-  int64_t i64;
-
-  double d;
-  float  f;
-} SDataTypeBuf;
-
-#define SIF_DATA_CONVERT(type, val, dst)       \
-  do {                                         \
-    if (type == TSDB_DATA_TYPE_DOUBLE)         \
-      dst = GET_DOUBLE_VAL(val);               \
-    else if (type == TSDB_DATA_TYPE_BIGINT)    \
-      dst = *(int64_t *)val;                   \
-    else if (type == TSDB_DATA_TYPE_INT)       \
-      dst = *(int32_t *)val;                   \
-    else if (type == TSDB_DATA_TYPE_SMALLINT)  \
-      dst = *(int16_t *)val;                   \
-    else if (type == TSDB_DATA_TYPE_TINYINT)   \
-      dst = *(int8_t *)val;                    \
-    else if (type == TSDB_DATA_TYPE_UTINYINT)  \
-      dst = *(uint8_t *)val;                   \
-    else if (type == TSDB_DATA_TYPE_USMALLINT) \
-      dst = *(uint16_t *)val;                  \
-    else if (type == TSDB_DATA_TYPE_UINT)      \
-      dst = *(uint32_t *)val;                  \
-    else if (type == TSDB_DATA_TYPE_UBIGINT)   \
-      dst = *(uint64_t *)val;                  \
-  } while (0);
 
 static void sifSetFltParam(SIFParam *left, SIFParam *right, SDataTypeBuf *typedata, SMetaFltParam *param) {
   int8_t ltype = left->colValType, rtype = right->colValType;
@@ -693,11 +689,8 @@ static int32_t sifExecLogic(SLogicConditionNode *node, SIFCtx *ctx, SIFParam *ou
     for (int32_t m = 0; m < node->pParameterList->length; m++) {
       if (node->condType == LOGIC_COND_TYPE_AND) {
         taosArrayAddAll(output->result, params[m].result);
-        // taosArrayDestroy(params[m].result);
-        //  params[m].result = NULL;
       } else if (node->condType == LOGIC_COND_TYPE_OR) {
         taosArrayAddAll(output->result, params[m].result);
-        // params[m].result = NULL;
       } else if (node->condType == LOGIC_COND_TYPE_NOT) {
         // taosArrayAddAll(output->result, params[m].result);
       }

source/libs/index/src/indexFstFile.c

@@ -211,9 +211,7 @@ IdxFstFile* idxFileCreate(void* wrt) {
   return cw;
 }
 void idxFileDestroy(IdxFstFile* cw) {
-  // free wrt object: close fd or free mem
   idxFileFlush(cw);
-  // idxFileCtxDestroy((IFileCtx *)(cw->wrt));
   taosMemoryFree(cw);
 }
 
@@ -222,9 +220,7 @@ int idxFileWrite(IdxFstFile* write, uint8_t* buf, uint32_t len) {
     return 0;
   }
   // update checksum
-  // write data to file/socket or mem
   IFileCtx* ctx = write->wrt;
-
   int       nWrite = ctx->write(ctx, buf, len);
   assert(nWrite == len);
   write->count += len;

source/libs/index/src/indexTfile.c

@@ -183,13 +183,14 @@ TFileReader* tfileReaderCreate(IFileCtx* ctx) {
     return NULL;
   }
   reader->ctx = ctx;
+  reader->remove = false;
 
   if (0 != tfileReaderVerify(reader)) {
     indexError("invalid tfile, suid:%" PRIu64 ", colName:%s", reader->header.suid, reader->header.colName);
     tfileReaderDestroy(reader);
     return NULL;
   }
-  // T_REF_INC(reader);
+
   if (0 != tfileReaderLoadHeader(reader)) {
     indexError("failed to load index header, suid:%" PRIu64 ", colName:%s", reader->header.suid,
                reader->header.colName);
@@ -203,7 +204,6 @@ TFileReader* tfileReaderCreate(IFileCtx* ctx) {
     tfileReaderDestroy(reader);
     return NULL;
   }
-  reader->remove = false;
 
   return reader;
 }
@@ -211,7 +211,6 @@ void tfileReaderDestroy(TFileReader* reader) {
   if (reader == NULL) {
     return;
   }
-  // T_REF_INC(reader);
   fstDestroy(reader->fst);
   if (reader->remove) {
     indexInfo("%s is removed", reader->ctx->file.buf);
@@ -222,6 +221,7 @@ void tfileReaderDestroy(TFileReader* reader) {
 
   taosMemoryFree(reader);
 }
+
 static int32_t tfSearchTerm(void* reader, SIndexTerm* tem, SIdxTRslt* tr) {
   int      ret = 0;
   char*    p = tem->colVal;
@@ -494,7 +494,6 @@ int tfileReaderSearch(TFileReader* reader, SIndexTermQuery* query, SIdxTRslt* tr
 TFileWriter* tfileWriterOpen(char* path, uint64_t suid, int64_t version, const char* colName, uint8_t colType) {
   char fullname[256] = {0};
   tfileGenFileFullName(fullname, path, suid, colName, version);
-  // indexInfo("open write file name %s", fullname);
   IFileCtx* wcx = idxFileCtxCreate(TFILE, fullname, false, 1024 * 1024 * 64);
   if (wcx == NULL) {
     return NULL;
@@ -503,8 +502,8 @@ TFileWriter* tfileWriterOpen(char* path, uint64_t suid, int64_t version, const c
   TFileHeader tfh = {0};
   tfh.suid = suid;
   tfh.version = version;
-  memcpy(tfh.colName, colName, strlen(colName));
   tfh.colType = colType;
+  memcpy(tfh.colName, colName, strlen(colName));
 
   return tfileWriterCreate(wcx, &tfh);
 }
@@ -706,7 +705,6 @@ static bool tfileIteratorNext(Iterate* iiter) {
   iv->type = ADD_VALUE;  // value in tfile always ADD_VALUE
   iv->colVal = colVal;
   return true;
-  // std::string key(ch, sz);
 }
 
 static IterateValue* tifileIterateGetValue(Iterate* iter) { return &iter->val; }

source/libs/nodes/src/nodesCloneFuncs.c

@@ -777,6 +777,7 @@ SNode* nodesCloneNode(const SNode* pNode) {
       code = physiSessionCopy((const SSessionWinodwPhysiNode*)pNode, (SSessionWinodwPhysiNode*)pDst);
       break;
     case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION:
       code = physiPartitionCopy((const SPartitionPhysiNode*)pNode, (SPartitionPhysiNode*)pDst);
       break;
     default:

source/libs/nodes/src/nodesCodeFuncs.c

@@ -265,6 +265,8 @@ const char* nodesNodeName(ENodeType type) {
       return "PhysiStreamStateWindow";
     case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
       return "PhysiPartition";
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION:
+      return "PhysiStreamPartition";
     case QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC:
       return "PhysiIndefRowsFunc";
     case QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC:
@@ -4485,6 +4487,7 @@ static int32_t specificNodeToJson(const void* pObj, SJson* pJson) {
     case QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE:
       return physiStateWindowNodeToJson(pObj, pJson);
     case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION:
       return physiPartitionNodeToJson(pObj, pJson);
     case QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC:
       return physiIndefRowsFuncNodeToJson(pObj, pJson);
@@ -4632,6 +4635,7 @@ static int32_t jsonToSpecificNode(const SJson* pJson, void* pObj) {
     case QUERY_NODE_PHYSICAL_PLAN_STREAM_STATE:
       return jsonToPhysiStateWindowNode(pJson, pObj);
     case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION:
       return jsonToPhysiPartitionNode(pJson, pObj);
     case QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC:
       return jsonToPhysiIndefRowsFuncNode(pJson, pObj);

source/libs/nodes/src/nodesTraverseFuncs.c

@@ -537,7 +537,8 @@ static EDealRes dispatchPhysiPlan(SNode* pNode, ETraversalOrder order, FNodeWalk
       }
       break;
     }
-    case QUERY_NODE_PHYSICAL_PLAN_PARTITION: {
+    case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION: {
       SPartitionPhysiNode* pPart = (SPartitionPhysiNode*)pNode;
       res = walkPhysiNode((SPhysiNode*)pNode, order, walker, pContext);
       if (DEAL_RES_ERROR != res && DEAL_RES_END != res) {

source/libs/nodes/src/nodesUtilFuncs.c

@@ -322,6 +322,8 @@ SNode* nodesMakeNode(ENodeType type) {
       return makeNode(type, sizeof(SStreamStateWinodwPhysiNode));
     case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
       return makeNode(type, sizeof(SPartitionPhysiNode));
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION:
+      return makeNode(type, sizeof(SStreamPartitionPhysiNode));
     case QUERY_NODE_PHYSICAL_PLAN_INDEF_ROWS_FUNC:
       return makeNode(type, sizeof(SIndefRowsFuncPhysiNode));
     case QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC:
@@ -951,7 +953,8 @@ void nodesDestroyNode(SNode* pNode) {
       nodesDestroyNode(pPhyNode->pStateKey);
       break;
     }
-    case QUERY_NODE_PHYSICAL_PLAN_PARTITION: {
+    case QUERY_NODE_PHYSICAL_PLAN_PARTITION:
+    case QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION: {
       SPartitionPhysiNode* pPhyNode = (SPartitionPhysiNode*)pNode;
       destroyPhysiNode((SPhysiNode*)pPhyNode);
       nodesDestroyList(pPhyNode->pExprs);

source/libs/parser/inc/parAst.h

@@ -59,7 +59,10 @@ typedef enum EDatabaseOptionType {
   DB_OPTION_WAL_RETENTION_PERIOD,
   DB_OPTION_WAL_RETENTION_SIZE,
   DB_OPTION_WAL_ROLL_PERIOD,
-  DB_OPTION_WAL_SEGMENT_SIZE
+  DB_OPTION_WAL_SEGMENT_SIZE,
+  DB_OPTION_SST_TRIGGER,
+  DB_OPTION_TABLE_PREFIX,
+  DB_OPTION_TABLE_SUFFIX
 } EDatabaseOptionType;
 
 typedef enum ETableOptionType {
@@ -170,6 +173,7 @@ SNode* createShowCreateDatabaseStmt(SAstCreateContext* pCxt, SToken* pDbName);
 SNode* createShowCreateTableStmt(SAstCreateContext* pCxt, ENodeType type, SNode* pRealTable);
 SNode* createShowTableDistributedStmt(SAstCreateContext* pCxt, SNode* pRealTable);
 SNode* createShowDnodeVariablesStmt(SAstCreateContext* pCxt, SNode* pDnodeId);
+SNode* createShowVnodesStmt(SAstCreateContext* pCxt, SNode* pDnodeId, SNode* pDnodeEndpoint);
 SNode* createCreateUserStmt(SAstCreateContext* pCxt, SToken* pUserName, const SToken* pPassword, int8_t sysinfo);
 SNode* createAlterUserStmt(SAstCreateContext* pCxt, SToken* pUserName, int8_t alterType, const SToken* pVal);
 SNode* createDropUserStmt(SAstCreateContext* pCxt, SToken* pUserName);

source/libs/parser/inc/sql.y

@@ -207,6 +207,9 @@ db_options(A) ::= db_options(B) WAL_RETENTION_SIZE NK_MINUS(D) NK_INTEGER(C).
                                                                                   }
 db_options(A) ::= db_options(B) WAL_ROLL_PERIOD NK_INTEGER(C).                    { A = setDatabaseOption(pCxt, B, DB_OPTION_WAL_ROLL_PERIOD, &C); }
 db_options(A) ::= db_options(B) WAL_SEGMENT_SIZE NK_INTEGER(C).                   { A = setDatabaseOption(pCxt, B, DB_OPTION_WAL_SEGMENT_SIZE, &C); }
+db_options(A) ::= db_options(B) SST_TRIGGER NK_INTEGER(C).                        { A = setDatabaseOption(pCxt, B, DB_OPTION_SST_TRIGGER, &C); }
+db_options(A) ::= db_options(B) TABLE_PREFIX NK_INTEGER(C).                       { A = setDatabaseOption(pCxt, B, DB_OPTION_TABLE_PREFIX, &C); }
+db_options(A) ::= db_options(B) TABLE_SUFFIX NK_INTEGER(C).                       { A = setDatabaseOption(pCxt, B, DB_OPTION_TABLE_SUFFIX, &C); }
 
 alter_db_options(A) ::= alter_db_option(B).                                       { A = createAlterDatabaseOptions(pCxt); A = setAlterDatabaseOption(pCxt, A, &B); }
 alter_db_options(A) ::= alter_db_options(B) alter_db_option(C).                   { A = setAlterDatabaseOption(pCxt, B, &C); }
@@ -223,6 +226,7 @@ alter_db_option(A) ::= KEEP variable_list(B).
 //alter_db_option(A) ::= REPLICA NK_INTEGER(B).                                     { A.type = DB_OPTION_REPLICA; A.val = B; }
 //alter_db_option(A) ::= STRICT NK_STRING(B).                                       { A.type = DB_OPTION_STRICT; A.val = B; }
 alter_db_option(A) ::= WAL_LEVEL NK_INTEGER(B).                                   { A.type = DB_OPTION_WAL; A.val = B; }
+alter_db_option(A) ::= SST_TRIGGER NK_INTEGER(B).                                 { A.type = DB_OPTION_SST_TRIGGER; A.val = B; }
 
 %type integer_list                                                                { SNodeList* }
 %destructor integer_list                                                          { nodesDestroyList($$); }
@@ -410,6 +414,8 @@ cmd ::= SHOW TABLE DISTRIBUTED full_table_name(A).
 cmd ::= SHOW CONSUMERS.                                                           { pCxt->pRootNode = createShowStmt(pCxt, QUERY_NODE_SHOW_CONSUMERS_STMT); }
 cmd ::= SHOW SUBSCRIPTIONS.                                                       { pCxt->pRootNode = createShowStmt(pCxt, QUERY_NODE_SHOW_SUBSCRIPTIONS_STMT); }
 cmd ::= SHOW TAGS FROM table_name_cond(A) from_db_opt(B).                         { pCxt->pRootNode = createShowStmtWithCond(pCxt, QUERY_NODE_SHOW_TAGS_STMT, B, A, OP_TYPE_EQUAL); }
+cmd ::= SHOW VNODES NK_INTEGER(A).                                                { pCxt->pRootNode = createShowVnodesStmt(pCxt, createValueNode(pCxt, TSDB_DATA_TYPE_BIGINT, &A), NULL); }
+cmd ::= SHOW VNODES NK_STRING(A).                                                 { pCxt->pRootNode = createShowVnodesStmt(pCxt, NULL, createValueNode(pCxt, TSDB_DATA_TYPE_VARCHAR, &A)); }
 
 db_name_cond_opt(A) ::= .                                                         { A = createDefaultDatabaseCondValue(pCxt); }
 db_name_cond_opt(A) ::= db_name(B) NK_DOT.                                        { A = createValueNode(pCxt, TSDB_DATA_TYPE_BINARY, &B); }
@@ -997,4 +1003,4 @@ null_ordering_opt(A) ::= NULLS LAST.
 
 %fallback ABORT AFTER ATTACH BEFORE BEGIN BITAND BITNOT BITOR BLOCKS CHANGE COMMA COMPACT CONCAT CONFLICT COPY DEFERRED DELIMITERS DETACH DIVIDE DOT EACH END FAIL 
   FILE FOR GLOB ID IMMEDIATE IMPORT INITIALLY INSTEAD ISNULL KEY NK_BITNOT NK_SEMI NOTNULL OF PLUS PRIVILEGE RAISE REPLACE RESTRICT ROW SEMI STAR STATEMENT STRING 
-  TIMES UPDATE VALUES VARIABLE VIEW VNODES WAL.
+  TIMES UPDATE VALUES VARIABLE VIEW WAL.

source/libs/parser/src/parAstCreater.c

@@ -795,6 +795,20 @@ SNode* createSetOperator(SAstCreateContext* pCxt, ESetOperatorType type, SNode*
   return (SNode*)setOp;
 }
 
+static void updateWalOptionsDefault(SDatabaseOptions* pOptions) {
+  if (!pOptions->walRetentionPeriodIsSet) {
+    pOptions->walRetentionPeriod =
+        pOptions->replica > 1 ? TSDB_REPS_DEF_DB_WAL_RET_PERIOD : TSDB_REP_DEF_DB_WAL_RET_PERIOD;
+  }
+  if (!pOptions->walRetentionSizeIsSet) {
+    pOptions->walRetentionSize = pOptions->replica > 1 ? TSDB_REPS_DEF_DB_WAL_RET_SIZE : TSDB_REP_DEF_DB_WAL_RET_SIZE;
+  }
+  if (!pOptions->walRollPeriodIsSet) {
+    pOptions->walRollPeriod =
+        pOptions->replica > 1 ? TSDB_REPS_DEF_DB_WAL_ROLL_PERIOD : TSDB_REP_DEF_DB_WAL_ROLL_PERIOD;
+  }
+}
+
 SNode* createDefaultDatabaseOptions(SAstCreateContext* pCxt) {
   CHECK_PARSER_STATUS(pCxt);
   SDatabaseOptions* pOptions = (SDatabaseOptions*)nodesMakeNode(QUERY_NODE_DATABASE_OPTIONS);
@@ -819,10 +833,11 @@ SNode* createDefaultDatabaseOptions(SAstCreateContext* pCxt) {
   pOptions->numOfVgroups = TSDB_DEFAULT_VN_PER_DB;
   pOptions->singleStable = TSDB_DEFAULT_DB_SINGLE_STABLE;
   pOptions->schemaless = TSDB_DEFAULT_DB_SCHEMALESS;
-  pOptions->walRetentionPeriod = TSDB_DEFAULT_DB_WAL_RETENTION_PERIOD;
+  updateWalOptionsDefault(pOptions);
-  pOptions->walRetentionSize = TSDB_DEFAULT_DB_WAL_RETENTION_SIZE;
-  pOptions->walRollPeriod = TSDB_DEFAULT_DB_WAL_ROLL_PERIOD;
   pOptions->walSegmentSize = TSDB_DEFAULT_DB_WAL_SEGMENT_SIZE;
+  pOptions->sstTrigger = TSDB_DEFAULT_SST_TRIGGER;
+  pOptions->tablePrefix = TSDB_DEFAULT_HASH_PREFIX;
+  pOptions->tableSuffix = TSDB_DEFAULT_HASH_SUFFIX;
   return (SNode*)pOptions;
 }
 
@@ -854,83 +869,100 @@ SNode* createAlterDatabaseOptions(SAstCreateContext* pCxt) {
   pOptions->walRetentionSize = -1;
   pOptions->walRollPeriod = -1;
   pOptions->walSegmentSize = -1;
+  pOptions->sstTrigger = -1;
+  pOptions->tablePrefix = -1;
+  pOptions->tableSuffix = -1;
   return (SNode*)pOptions;
 }
 
 SNode* setDatabaseOption(SAstCreateContext* pCxt, SNode* pOptions, EDatabaseOptionType type, void* pVal) {
   CHECK_PARSER_STATUS(pCxt);
+  SDatabaseOptions* pDbOptions = (SDatabaseOptions*)pOptions;
   switch (type) {
     case DB_OPTION_BUFFER:
-      ((SDatabaseOptions*)pOptions)->buffer = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->buffer = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_CACHEMODEL:
-      COPY_STRING_FORM_STR_TOKEN(((SDatabaseOptions*)pOptions)->cacheModelStr, (SToken*)pVal);
+      COPY_STRING_FORM_STR_TOKEN(pDbOptions->cacheModelStr, (SToken*)pVal);
       break;
     case DB_OPTION_CACHESIZE:
-      ((SDatabaseOptions*)pOptions)->cacheLastSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->cacheLastSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_COMP:
-      ((SDatabaseOptions*)pOptions)->compressionLevel = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->compressionLevel = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_DAYS: {
       SToken* pToken = pVal;
       if (TK_NK_INTEGER == pToken->type) {
-        ((SDatabaseOptions*)pOptions)->daysPerFile = taosStr2Int32(pToken->z, NULL, 10) * 1440;
+        pDbOptions->daysPerFile = taosStr2Int32(pToken->z, NULL, 10) * 1440;
       } else {
-        ((SDatabaseOptions*)pOptions)->pDaysPerFile = (SValueNode*)createDurationValueNode(pCxt, pToken);
+        pDbOptions->pDaysPerFile = (SValueNode*)createDurationValueNode(pCxt, pToken);
       }
       break;
     }
     case DB_OPTION_FSYNC:
-      ((SDatabaseOptions*)pOptions)->fsyncPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->fsyncPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_MAXROWS:
-      ((SDatabaseOptions*)pOptions)->maxRowsPerBlock = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->maxRowsPerBlock = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_MINROWS:
-      ((SDatabaseOptions*)pOptions)->minRowsPerBlock = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->minRowsPerBlock = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_KEEP:
-      ((SDatabaseOptions*)pOptions)->pKeep = pVal;
+      pDbOptions->pKeep = pVal;
       break;
     case DB_OPTION_PAGES:
-      ((SDatabaseOptions*)pOptions)->pages = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->pages = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_PAGESIZE:
-      ((SDatabaseOptions*)pOptions)->pagesize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->pagesize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_PRECISION:
-      COPY_STRING_FORM_STR_TOKEN(((SDatabaseOptions*)pOptions)->precisionStr, (SToken*)pVal);
+      COPY_STRING_FORM_STR_TOKEN(pDbOptions->precisionStr, (SToken*)pVal);
       break;
     case DB_OPTION_REPLICA:
-      ((SDatabaseOptions*)pOptions)->replica = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->replica = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
+      updateWalOptionsDefault(pDbOptions);
       break;
     case DB_OPTION_STRICT:
-      COPY_STRING_FORM_STR_TOKEN(((SDatabaseOptions*)pOptions)->strictStr, (SToken*)pVal);
+      COPY_STRING_FORM_STR_TOKEN(pDbOptions->strictStr, (SToken*)pVal);
       break;
     case DB_OPTION_WAL:
-      ((SDatabaseOptions*)pOptions)->walLevel = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walLevel = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_VGROUPS:
-      ((SDatabaseOptions*)pOptions)->numOfVgroups = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->numOfVgroups = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_SINGLE_STABLE:
-      ((SDatabaseOptions*)pOptions)->singleStable = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->singleStable = taosStr2Int8(((SToken*)pVal)->z, NULL, 10);
       break;
     case DB_OPTION_RETENTIONS:
-      ((SDatabaseOptions*)pOptions)->pRetentions = pVal;
+      pDbOptions->pRetentions = pVal;
       break;
     case DB_OPTION_WAL_RETENTION_PERIOD:
-      ((SDatabaseOptions*)pOptions)->walRetentionPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRetentionPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRetentionPeriodIsSet = true;
       break;
     case DB_OPTION_WAL_RETENTION_SIZE:
-      ((SDatabaseOptions*)pOptions)->walRetentionSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRetentionSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRetentionSizeIsSet = true;
       break;
     case DB_OPTION_WAL_ROLL_PERIOD:
-      ((SDatabaseOptions*)pOptions)->walRollPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRollPeriod = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walRollPeriodIsSet = true;
       break;
     case DB_OPTION_WAL_SEGMENT_SIZE:
-      ((SDatabaseOptions*)pOptions)->walSegmentSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      pDbOptions->walSegmentSize = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      break;
+    case DB_OPTION_SST_TRIGGER:
+      pDbOptions->sstTrigger = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      break;
+    case DB_OPTION_TABLE_PREFIX:
+      pDbOptions->tablePrefix = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
+      break;
+    case DB_OPTION_TABLE_SUFFIX:
+      pDbOptions->tableSuffix = taosStr2Int32(((SToken*)pVal)->z, NULL, 10);
       break;
     default:
       break;
@@ -1251,7 +1283,8 @@ SNode* createUseDatabaseStmt(SAstCreateContext* pCxt, SToken* pDbName) {
 
 static bool needDbShowStmt(ENodeType type) {
   return QUERY_NODE_SHOW_TABLES_STMT == type || QUERY_NODE_SHOW_STABLES_STMT == type ||
-         QUERY_NODE_SHOW_VGROUPS_STMT == type;
+         QUERY_NODE_SHOW_VGROUPS_STMT == type || QUERY_NODE_SHOW_INDEXES_STMT == type ||
+         QUERY_NODE_SHOW_TAGS_STMT == type;
 }
 
 SNode* createShowStmt(SAstCreateContext* pCxt, ENodeType type) {
@@ -1264,7 +1297,7 @@ SNode* createShowStmt(SAstCreateContext* pCxt, ENodeType type) {
 SNode* createShowStmtWithCond(SAstCreateContext* pCxt, ENodeType type, SNode* pDbName, SNode* pTbName,
                               EOperatorType tableCondType) {
   CHECK_PARSER_STATUS(pCxt);
-  if (needDbShowStmt(type) && NULL == pDbName && NULL == pCxt->pQueryCxt->db) {
+  if (needDbShowStmt(type) && NULL == pDbName) {
     snprintf(pCxt->pQueryCxt->pMsg, pCxt->pQueryCxt->msgLen, "db not specified");
     pCxt->errCode = TSDB_CODE_PAR_SYNTAX_ERROR;
     return NULL;
@@ -1316,6 +1349,15 @@ SNode* createShowDnodeVariablesStmt(SAstCreateContext* pCxt, SNode* pDnodeId) {
   return (SNode*)pStmt;
 }
 
+SNode* createShowVnodesStmt(SAstCreateContext* pCxt, SNode* pDnodeId, SNode* pDnodeEndpoint) {
+  CHECK_PARSER_STATUS(pCxt);
+  SShowVnodesStmt* pStmt = (SShowVnodesStmt*)nodesMakeNode(QUERY_NODE_SHOW_VNODES_STMT);
+  CHECK_OUT_OF_MEM(pStmt);
+  pStmt->pDnodeId = pDnodeId;
+  pStmt->pDnodeEndpoint = pDnodeEndpoint;
+  return (SNode*)pStmt;
+}
+
 SNode* createCreateUserStmt(SAstCreateContext* pCxt, SToken* pUserName, const SToken* pPassword, int8_t sysinfo) {
   CHECK_PARSER_STATUS(pCxt);
   char password[TSDB_USET_PASSWORD_LEN] = {0};

source/libs/parser/src/parAstParser.c

@@ -374,8 +374,13 @@ static int32_t collectMetaKeyFromShowIndexes(SCollectMetaKeyCxt* pCxt, SShowStmt
 }
 
 static int32_t collectMetaKeyFromShowStables(SCollectMetaKeyCxt* pCxt, SShowStmt* pStmt) {
-  return reserveTableMetaInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_STABLES,
+  int32_t code = reserveTableMetaInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_STABLES,
                                          pCxt->pMetaCache);
+  if (TSDB_CODE_SUCCESS == code) {
+    code = reserveUserAuthInCache(pCxt->pParseCxt->acctId, pCxt->pParseCxt->pUser,
+                                  ((SValueNode*)pStmt->pDbName)->literal, AUTH_TYPE_READ, pCxt->pMetaCache);
+  }
+  return code;
 }
 
 static int32_t collectMetaKeyFromShowStreams(SCollectMetaKeyCxt* pCxt, SShowStmt* pStmt) {
@@ -387,11 +392,11 @@ static int32_t collectMetaKeyFromShowTables(SCollectMetaKeyCxt* pCxt, SShowStmt*
   int32_t code = reserveTableMetaInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_TABLES,
                                          pCxt->pMetaCache);
   if (TSDB_CODE_SUCCESS == code) {
-    if (NULL != pStmt->pDbName) {
     code = reserveDbVgInfoInCache(pCxt->pParseCxt->acctId, ((SValueNode*)pStmt->pDbName)->literal, pCxt->pMetaCache);
-    } else {
-      code = reserveDbVgInfoInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, pCxt->pMetaCache);
   }
+  if (TSDB_CODE_SUCCESS == code) {
+    code = reserveUserAuthInCache(pCxt->pParseCxt->acctId, pCxt->pParseCxt->pUser,
+                                  ((SValueNode*)pStmt->pDbName)->literal, AUTH_TYPE_READ, pCxt->pMetaCache);
   }
   return code;
 }
@@ -400,11 +405,7 @@ static int32_t collectMetaKeyFromShowTags(SCollectMetaKeyCxt* pCxt, SShowStmt* p
   int32_t code = reserveTableMetaInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_TAGS,
                                          pCxt->pMetaCache);
   if (TSDB_CODE_SUCCESS == code) {
-    if (NULL != pStmt->pDbName) {
     code = reserveDbVgInfoInCache(pCxt->pParseCxt->acctId, ((SValueNode*)pStmt->pDbName)->literal, pCxt->pMetaCache);
-    } else {
-      code = reserveDbVgInfoInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, pCxt->pMetaCache);
-    }
   }
   return code;
 }
@@ -463,6 +464,11 @@ static int32_t collectMetaKeyFromShowDnodeVariables(SCollectMetaKeyCxt* pCxt, SS
   return code;
 }
 
+static int32_t collectMetaKeyFromShowVnodes(SCollectMetaKeyCxt* pCxt, SShowVnodesStmt* pStmt) {
+  return reserveTableMetaInCache(pCxt->pParseCxt->acctId, TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_VNODES,
+                                 pCxt->pMetaCache);
+}
+
 static int32_t collectMetaKeyFromShowCreateDatabase(SCollectMetaKeyCxt* pCxt, SShowCreateDatabaseStmt* pStmt) {
   return reserveDbCfgInCache(pCxt->pParseCxt->acctId, pStmt->dbName, pCxt->pMetaCache);
 }
@@ -603,6 +609,8 @@ static int32_t collectMetaKeyFromQuery(SCollectMetaKeyCxt* pCxt, SNode* pStmt) {
       return collectMetaKeyFromShowVariables(pCxt, (SShowStmt*)pStmt);
     case QUERY_NODE_SHOW_DNODE_VARIABLES_STMT:
       return collectMetaKeyFromShowDnodeVariables(pCxt, (SShowDnodeVariablesStmt*)pStmt);
+    case QUERY_NODE_SHOW_VNODES_STMT:
+      return collectMetaKeyFromShowVnodes(pCxt, (SShowVnodesStmt*)pStmt);
     case QUERY_NODE_SHOW_CREATE_DATABASE_STMT:
       return collectMetaKeyFromShowCreateDatabase(pCxt, (SShowCreateDatabaseStmt*)pStmt);
     case QUERY_NODE_SHOW_CREATE_TABLE_STMT:

source/libs/parser/src/parAuthenticator.c

@@ -96,6 +96,10 @@ static int32_t authInsert(SAuthCxt* pCxt, SInsertStmt* pInsert) {
   return code;
 }
 
+static int32_t authShowTables(SAuthCxt* pCxt, SShowStmt* pStmt) {
+  return checkAuth(pCxt, ((SValueNode*)pStmt->pDbName)->literal, AUTH_TYPE_READ);
+}
+
 static int32_t authShowCreateTable(SAuthCxt* pCxt, SShowCreateTableStmt* pStmt) {
   return checkAuth(pCxt, pStmt->dbName, AUTH_TYPE_READ);
 }
@@ -127,6 +131,9 @@ static int32_t authQuery(SAuthCxt* pCxt, SNode* pStmt) {
     case QUERY_NODE_SHOW_VNODES_STMT:
     case QUERY_NODE_SHOW_SCORES_STMT:
       return !pCxt->pParseCxt->enableSysInfo ? TSDB_CODE_PAR_PERMISSION_DENIED : TSDB_CODE_SUCCESS;
+    case QUERY_NODE_SHOW_TABLES_STMT:
+    case QUERY_NODE_SHOW_STABLES_STMT:
+      return authShowTables(pCxt, (SShowStmt*)pStmt);
     case QUERY_NODE_SHOW_CREATE_TABLE_STMT:
     case QUERY_NODE_SHOW_CREATE_STABLE_STMT:
       return authShowCreateTable(pCxt, (SShowCreateTableStmt*)pStmt);

source/libs/parser/src/parInsert.c

@@ -786,6 +786,7 @@ static void buildCreateTbReq(SVCreateTbReq* pTbReq, const char* tname, STag* pTa
   if (sname) pTbReq->ctb.name = strdup(sname);
   pTbReq->ctb.pTag = (uint8_t*)pTag;
   pTbReq->ctb.tagName = taosArrayDup(tagName);
+  pTbReq->ttl = TSDB_DEFAULT_TABLE_TTL;
   pTbReq->commentLen = -1;
 
   return;
@@ -1121,6 +1122,40 @@ static int32_t ignoreAutoCreateTableClause(SInsertParseContext* pCxt) {
   return code;
 }
 
+static int32_t parseTableOptions(SInsertParseContext* pCxt) {
+  do {
+    int32_t index = 0;
+    SToken  sToken;
+    NEXT_TOKEN_KEEP_SQL(pCxt->pSql, sToken, index);
+    if (TK_TTL == sToken.type) {
+      pCxt->pSql += index;
+      NEXT_TOKEN(pCxt->pSql, sToken);
+      if (TK_NK_INTEGER != sToken.type) {
+        return buildSyntaxErrMsg(&pCxt->msg, "Invalid option ttl", sToken.z);
+      }
+      pCxt->createTblReq.ttl = taosStr2Int32(sToken.z, NULL, 10);
+    } else if (TK_COMMENT == sToken.type) {
+      pCxt->pSql += index;
+      NEXT_TOKEN(pCxt->pSql, sToken);
+      if (TK_NK_STRING != sToken.type) {
+        return buildSyntaxErrMsg(&pCxt->msg, "Invalid option comment", sToken.z);
+      }
+      if (sToken.n >= TSDB_TB_COMMENT_LEN) {
+        return buildSyntaxErrMsg(&pCxt->msg, "comment too long", sToken.z);
+      }
+      int32_t len = trimString(sToken.z, sToken.n, pCxt->tmpTokenBuf, TSDB_TB_COMMENT_LEN);
+      pCxt->createTblReq.comment = strndup(pCxt->tmpTokenBuf, len);
+      if (NULL == pCxt->createTblReq.comment) {
+        return TSDB_CODE_OUT_OF_MEMORY;
+      }
+      pCxt->createTblReq.commentLen = len;
+    } else {
+      break;
+    }
+  } while (1);
+  return TSDB_CODE_SUCCESS;
+}
+
 // pSql -> stb_name [(tag1_name, ...)] TAGS (tag1_value, ...)
 static int32_t parseUsingClause(SInsertParseContext* pCxt, int32_t tbNo, SName* name, char* tbFName) {
   int32_t      len = strlen(tbFName);
@@ -1172,7 +1207,7 @@ static int32_t parseUsingClause(SInsertParseContext* pCxt, int32_t tbNo, SName*
     return buildSyntaxErrMsg(&pCxt->msg, ") is expected", sToken.z);
   }
 
-  return TSDB_CODE_SUCCESS;
+  return parseTableOptions(pCxt);
 }
 
 static int parseOneRow(SInsertParseContext* pCxt, STableDataBlocks* pDataBlocks, int16_t timePrec, bool* gotRow,
@@ -1700,6 +1735,21 @@ static int32_t skipValuesClause(SInsertParseSyntaxCxt* pCxt) {
 
 static int32_t skipTagsClause(SInsertParseSyntaxCxt* pCxt) { return skipParentheses(pCxt); }
 
+static int32_t skipTableOptions(SInsertParseSyntaxCxt* pCxt) {
+  do {
+    int32_t index = 0;
+    SToken  sToken;
+    NEXT_TOKEN_KEEP_SQL(pCxt->pSql, sToken, index);
+    if (TK_TTL == sToken.type || TK_COMMENT == sToken.type) {
+      pCxt->pSql += index;
+      NEXT_TOKEN(pCxt->pSql, sToken);
+    } else {
+      break;
+    }
+  } while (1);
+  return TSDB_CODE_SUCCESS;
+}
+
 // pSql -> [(tag1_name, ...)] TAGS (tag1_value, ...)
 static int32_t skipUsingClause(SInsertParseSyntaxCxt* pCxt) {
   SToken sToken;
@@ -1718,6 +1768,7 @@ static int32_t skipUsingClause(SInsertParseSyntaxCxt* pCxt) {
     return buildSyntaxErrMsg(&pCxt->msg, "( is expected", sToken.z);
   }
   CHECK_CODE(skipTagsClause(pCxt));
+  CHECK_CODE(skipTableOptions(pCxt));
 
   return TSDB_CODE_SUCCESS;
 }

source/libs/parser/src/parTokenizer.c

@@ -187,6 +187,7 @@ static SKeyword keywordTable[] = {
     {"SNODES",               TK_SNODES},
     {"SOFFSET",              TK_SOFFSET},
     {"SPLIT",                TK_SPLIT},
+    {"SST_TRIGGER",          TK_SST_TRIGGER},
     {"STABLE",               TK_STABLE},
     {"STABLES",              TK_STABLES},
     {"STATE",                TK_STATE},
@@ -199,6 +200,8 @@ static SKeyword keywordTable[] = {
     {"SYSINFO",              TK_SYSINFO},
     {"TABLE",                TK_TABLE},
     {"TABLES",               TK_TABLES},
+    {"TABLE_PREFIX",         TK_TABLE_PREFIX},
+    {"TABLE_SUFFIX",         TK_TABLE_SUFFIX},
     {"TAG",                  TK_TAG},
     {"TAGS",                 TK_TAGS},
     {"TBNAME",               TK_TBNAME},
@@ -228,6 +231,7 @@ static SKeyword keywordTable[] = {
     {"VERBOSE",              TK_VERBOSE},
     {"VGROUP",               TK_VGROUP},
     {"VGROUPS",              TK_VGROUPS},
+    {"VNODES",               TK_VNODES},
     {"WAL_FSYNC_PERIOD",     TK_WAL_FSYNC_PERIOD},
     {"WAL_LEVEL",            TK_WAL_LEVEL},
     {"WAL_RETENTION_PERIOD", TK_WAL_RETENTION_PERIOD},

source/libs/parser/src/parTranslater.c

@@ -245,6 +245,12 @@ static const SSysTableShowAdapter sysTableShowAdapter[] = {
     .numOfShowCols = 1,
     .pShowCols = {"*"}
   },
+  { .showType = QUERY_NODE_SHOW_VNODES_STMT,
+    .pDbName = TSDB_INFORMATION_SCHEMA_DB,
+    .pTableName = TSDB_INS_TABLE_VNODES,
+    .numOfShowCols = 1,
+    .pShowCols = {"*"}
+  },
 };
 // clang-format on
 
@@ -3476,6 +3482,9 @@ static int32_t buildCreateDbReq(STranslateContext* pCxt, SCreateDatabaseStmt* pS
   pReq->walRetentionSize = pStmt->pOptions->walRetentionSize;
   pReq->walRollPeriod = pStmt->pOptions->walRollPeriod;
   pReq->walSegmentSize = pStmt->pOptions->walSegmentSize;
+  pReq->sstTrigger = pStmt->pOptions->sstTrigger;
+  pReq->hashPrefix = pStmt->pOptions->tablePrefix;
+  pReq->hashSuffix = pStmt->pOptions->tableSuffix;
   pReq->ignoreExist = pStmt->ignoreExists;
   return buildCreateDbRetentions(pStmt->pOptions->pRetentions, pReq);
 }
@@ -3760,6 +3769,15 @@ static int32_t checkDatabaseOptions(STranslateContext* pCxt, const char* pDbName
     code =
         checkDbRangeOption(pCxt, "walSegmentSize", pOptions->walSegmentSize, TSDB_DB_MIN_WAL_SEGMENT_SIZE, INT32_MAX);
   }
+  if (TSDB_CODE_SUCCESS == code) {
+    code = checkDbRangeOption(pCxt, "sstTrigger", pOptions->sstTrigger, TSDB_MIN_SST_TRIGGER, TSDB_MAX_SST_TRIGGER);
+  }
+  if (TSDB_CODE_SUCCESS == code) {
+    code = checkDbRangeOption(pCxt, "tablePrefix", pOptions->tablePrefix, TSDB_MIN_HASH_PREFIX, TSDB_MAX_HASH_PREFIX);
+  }
+  if (TSDB_CODE_SUCCESS == code) {
+    code = checkDbRangeOption(pCxt, "tableSuffix", pOptions->tableSuffix, TSDB_MIN_HASH_SUFFIX, TSDB_MAX_HASH_SUFFIX);
+  }
   if (TSDB_CODE_SUCCESS == code) {
     code = checkOptionsDependency(pCxt, pDbName, pOptions);
   }
@@ -3833,6 +3851,7 @@ static void buildAlterDbReq(STranslateContext* pCxt, SAlterDatabaseStmt* pStmt,
   pReq->cacheLast = pStmt->pOptions->cacheModel;
   pReq->cacheLastSize = pStmt->pOptions->cacheLastSize;
   pReq->replications = pStmt->pOptions->replica;
+  pReq->sstTrigger = pStmt->pOptions->sstTrigger;
   return;
 }
 
@@ -5837,6 +5856,25 @@ static int32_t rewriteShowDnodeVariables(STranslateContext* pCxt, SQuery* pQuery
   return code;
 }
 
+static int32_t rewriteShowVnodes(STranslateContext* pCxt, SQuery* pQuery) {
+  SShowVnodesStmt* pShow = (SShowVnodesStmt*)(pQuery->pRoot);
+  SSelectStmt*     pStmt = NULL;
+  int32_t          code = createSelectStmtForShow(QUERY_NODE_SHOW_VNODES_STMT, &pStmt);
+  if (TSDB_CODE_SUCCESS == code) {
+    if (NULL != pShow->pDnodeId) {
+      code = createOperatorNode(OP_TYPE_EQUAL, "dnode_id", pShow->pDnodeId, &pStmt->pWhere);
+    } else {
+      code = createOperatorNode(OP_TYPE_EQUAL, "dnode_ep", pShow->pDnodeEndpoint, &pStmt->pWhere);
+    }
+  }
+  if (TSDB_CODE_SUCCESS == code) {
+    pQuery->showRewrite = true;
+    nodesDestroyNode(pQuery->pRoot);
+    pQuery->pRoot = (SNode*)pStmt;
+  }
+  return code;
+}
+
 static SNode* createBlockDistInfoFunc() {
   SFunctionNode* pFunc = (SFunctionNode*)nodesMakeNode(QUERY_NODE_FUNCTION);
   if (NULL == pFunc) {
@@ -6408,7 +6446,8 @@ typedef struct SVgroupDropTableBatch {
   char             dbName[TSDB_DB_NAME_LEN];
 } SVgroupDropTableBatch;
 
-static void addDropTbReqIntoVgroup(SHashObj* pVgroupHashmap, SDropTableClause* pClause, SVgroupInfo* pVgInfo, uint64_t suid) {
+static void addDropTbReqIntoVgroup(SHashObj* pVgroupHashmap, SDropTableClause* pClause, SVgroupInfo* pVgInfo,
+                                   uint64_t suid) {
   SVDropTbReq            req = {.name = pClause->tableName, .suid = suid, .igNotExists = pClause->ignoreNotExists};
   SVgroupDropTableBatch* pTableBatch = taosHashGet(pVgroupHashmap, &pVgInfo->vgId, sizeof(pVgInfo->vgId));
   if (NULL == pTableBatch) {
@@ -6946,6 +6985,9 @@ static int32_t rewriteQuery(STranslateContext* pCxt, SQuery* pQuery) {
     case QUERY_NODE_SHOW_DNODE_VARIABLES_STMT:
       code = rewriteShowDnodeVariables(pCxt, pQuery);
       break;
+    case QUERY_NODE_SHOW_VNODES_STMT:
+      code = rewriteShowVnodes(pCxt, pQuery);
+      break;
     case QUERY_NODE_SHOW_TABLE_DISTRIBUTED_STMT:
       code = rewriteShowTableDist(pCxt, pQuery);
       break;

source/libs/parser/test/mockCatalog.cpp

@@ -32,100 +32,59 @@
 namespace {
 
 void generateInformationSchema(MockCatalogService* mcs) {
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_DNODES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN);
+      .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_MNODES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN);
+      .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_MODULES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN);
+      .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_QNODES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN);
+      .addColumn("endpoint", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_DATABASES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN);
+      .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_FUNCTIONS, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_FUNC_NAME_LEN);
+      .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_FUNC_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_INDEXES, TSDB_SYSTEM_TABLE, 3)
       .addColumn("index_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN)
       .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-            .addColumn("table_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN);
+      .addColumn("table_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_STABLES, TSDB_SYSTEM_TABLE, 2)
       .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-            .addColumn("stable_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN);
+      .addColumn("stable_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_TABLES, TSDB_SYSTEM_TABLE, 2)
       .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-            .addColumn("table_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN);
+      .addColumn("table_name", TSDB_DATA_TYPE_BINARY, TSDB_TABLE_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_TABLE_DISTRIBUTED, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN);
+      .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_USERS, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_USER_LEN);
+      .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_USER_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_VGROUPS, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN);
+      .addColumn("db_name", TSDB_DATA_TYPE_BINARY, TSDB_DB_NAME_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_CONFIGS, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_CONFIG_OPTION_LEN);
+      .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_CONFIG_OPTION_LEN)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_DNODE_VARIABLES, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("dnode_id", TSDB_DATA_TYPE_INT);
+      .addColumn("dnode_id", TSDB_DATA_TYPE_INT)
-    builder.done();
+      .done();
-  }
-  {
-    ITableBuilder& builder =
   mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_CLUSTER, TSDB_SYSTEM_TABLE, 1)
-            .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_CLUSTER_ID_LEN);
+      .addColumn("name", TSDB_DATA_TYPE_BINARY, TSDB_CLUSTER_ID_LEN)
-    builder.done();
+      .done();
-  }
+  mcs->createTableBuilder(TSDB_INFORMATION_SCHEMA_DB, TSDB_INS_TABLE_VNODES, TSDB_SYSTEM_TABLE, 2)
+      .addColumn("dnode_id", TSDB_DATA_TYPE_INT)
+      .addColumn("dnode_ep", TSDB_DATA_TYPE_BINARY, TSDB_EP_LEN)
+      .done();
 }
 
 void generatePerformanceSchema(MockCatalogService* mcs) {

source/libs/parser/test/parAlterToBalanceTest.cpp

@@ -88,6 +88,7 @@ TEST_F(ParserInitialATest, alterDnode) {
  *   | REPLICA int_value                                         -- todo: enum 1, 3, default 1, unit replica
  *   | STRICT {'off' | 'on'}                                     -- todo: default 'off'
  *   | WAL_LEVEL int_value                                       -- enum 1, 2, default 1
+ *   | SST_TRIGGER int_value                                     -- rang [1, 128], default 8
  * }
  */
 TEST_F(ParserInitialATest, alterDatabase) {
@@ -112,6 +113,7 @@ TEST_F(ParserInitialATest, alterDatabase) {
     expect.cacheLast = -1;
     expect.cacheLastSize = -1;
     expect.replications = -1;
+    expect.sstTrigger = -1;
   };
   auto setAlterDbBuffer = [&](int32_t buffer) { expect.buffer = buffer; };
   auto setAlterDbPageSize = [&](int32_t pageSize) { expect.pageSize = pageSize; };
@@ -128,6 +130,7 @@ TEST_F(ParserInitialATest, alterDatabase) {
   auto setAlterDbStrict = [&](int8_t strict) { expect.strict = strict; };
   auto setAlterDbCacheModel = [&](int8_t cacheModel) { expect.cacheLast = cacheModel; };
   auto setAlterDbReplica = [&](int8_t replications) { expect.replications = replications; };
+  auto setAlterDbSstTrigger = [&](int8_t sstTrigger) { expect.sstTrigger = sstTrigger; };
 
   setCheckDdlFunc([&](const SQuery* pQuery, ParserStage stage) {
     ASSERT_EQ(nodeType(pQuery->pRoot), QUERY_NODE_ALTER_DATABASE_STMT);
@@ -146,6 +149,7 @@ TEST_F(ParserInitialATest, alterDatabase) {
     ASSERT_EQ(req.strict, expect.strict);
     ASSERT_EQ(req.cacheLast, expect.cacheLast);
     ASSERT_EQ(req.replications, expect.replications);
+    ASSERT_EQ(req.sstTrigger, expect.sstTrigger);
   });
 
   const int32_t MINUTE_PER_DAY = MILLISECOND_PER_DAY / MILLISECOND_PER_MINUTE;
@@ -157,7 +161,8 @@ TEST_F(ParserInitialATest, alterDatabase) {
   setAlterDbFsync(200);
   setAlterDbWal(1);
   setAlterDbCacheModel(TSDB_CACHE_MODEL_LAST_ROW);
-  run("ALTER DATABASE test CACHEMODEL 'last_row' CACHESIZE 32 WAL_FSYNC_PERIOD 200 KEEP 10 WAL_LEVEL 1");
+  setAlterDbSstTrigger(20);
+  run("ALTER DATABASE test CACHEMODEL 'last_row' CACHESIZE 32 WAL_FSYNC_PERIOD 200 KEEP 10 WAL_LEVEL 1 SST_TRIGGER 20");
   clearAlterDbReq();
 
   initAlterDb("test");
@@ -231,6 +236,8 @@ TEST_F(ParserInitialATest, alterDatabaseSemanticCheck) {
   run("ALTER DATABASE test KEEP 1w", TSDB_CODE_PAR_INVALID_DB_OPTION);
   run("ALTER DATABASE test WAL_LEVEL 0", TSDB_CODE_PAR_INVALID_DB_OPTION);
   run("ALTER DATABASE test WAL_LEVEL 3", TSDB_CODE_PAR_INVALID_DB_OPTION);
+  run("ALTER DATABASE test SST_TRIGGER 0", TSDB_CODE_PAR_INVALID_DB_OPTION);
+  run("ALTER DATABASE test SST_TRIGGER 129", TSDB_CODE_PAR_INVALID_DB_OPTION);
   // Regardless of the specific sentence
   run("ALTER DATABASE db WAL_LEVEL 0     # td-14436", TSDB_CODE_PAR_SYNTAX_ERROR, PARSER_STAGE_PARSE);
 }

source/libs/parser/test/parInitialCTest.cpp

@@ -111,10 +111,13 @@ TEST_F(ParserInitialCTest, createDatabase) {
     expect.numOfVgroups = TSDB_DEFAULT_VN_PER_DB;
     expect.numOfStables = TSDB_DEFAULT_DB_SINGLE_STABLE;
     expect.schemaless = TSDB_DEFAULT_DB_SCHEMALESS;
-    expect.walRetentionPeriod = TSDB_DEFAULT_DB_WAL_RETENTION_PERIOD;
+    expect.walRetentionPeriod = TSDB_REP_DEF_DB_WAL_RET_PERIOD;
-    expect.walRetentionSize = TSDB_DEFAULT_DB_WAL_RETENTION_SIZE;
+    expect.walRetentionSize = TSDB_REP_DEF_DB_WAL_RET_SIZE;
-    expect.walRollPeriod = TSDB_DEFAULT_DB_WAL_ROLL_PERIOD;
+    expect.walRollPeriod = TSDB_REP_DEF_DB_WAL_ROLL_PERIOD;
     expect.walSegmentSize = TSDB_DEFAULT_DB_WAL_SEGMENT_SIZE;
+    expect.sstTrigger = TSDB_DEFAULT_SST_TRIGGER;
+    expect.hashPrefix = TSDB_DEFAULT_HASH_PREFIX;
+    expect.hashSuffix = TSDB_DEFAULT_HASH_SUFFIX;
   };
 
   auto setDbBufferFunc = [&](int32_t buffer) { expect.buffer = buffer; };
@@ -155,6 +158,9 @@ TEST_F(ParserInitialCTest, createDatabase) {
   auto setDbWalRetentionSize = [&](int32_t walRetentionSize) { expect.walRetentionSize = walRetentionSize; };
   auto setDbWalRollPeriod = [&](int32_t walRollPeriod) { expect.walRollPeriod = walRollPeriod; };
   auto setDbWalSegmentSize = [&](int32_t walSegmentSize) { expect.walSegmentSize = walSegmentSize; };
+  auto setDbSstTrigger = [&](int32_t sstTrigger) { expect.sstTrigger = sstTrigger; };
+  auto setDbHashPrefix = [&](int32_t hashPrefix) { expect.hashPrefix = hashPrefix; };
+  auto setDbHashSuffix = [&](int32_t hashSuffix) { expect.hashSuffix = hashSuffix; };
 
   setCheckDdlFunc([&](const SQuery* pQuery, ParserStage stage) {
     ASSERT_EQ(nodeType(pQuery->pRoot), QUERY_NODE_CREATE_DATABASE_STMT);
@@ -185,7 +191,9 @@ TEST_F(ParserInitialCTest, createDatabase) {
     ASSERT_EQ(req.walRetentionSize, expect.walRetentionSize);
     ASSERT_EQ(req.walRollPeriod, expect.walRollPeriod);
     ASSERT_EQ(req.walSegmentSize, expect.walSegmentSize);
-    // ASSERT_EQ(req.schemaless, expect.schemaless);
+    ASSERT_EQ(req.sstTrigger, expect.sstTrigger);
+    ASSERT_EQ(req.hashPrefix, expect.hashPrefix);
+    ASSERT_EQ(req.hashSuffix, expect.hashSuffix);
     ASSERT_EQ(req.ignoreExist, expect.ignoreExist);
     ASSERT_EQ(req.numOfRetensions, expect.numOfRetensions);
     if (expect.numOfRetensions > 0) {
@@ -233,6 +241,9 @@ TEST_F(ParserInitialCTest, createDatabase) {
   setDbWalRetentionSize(-1);
   setDbWalRollPeriod(10);
   setDbWalSegmentSize(20);
+  setDbSstTrigger(16);
+  setDbHashPrefix(3);
+  setDbHashSuffix(4);
   run("CREATE DATABASE IF NOT EXISTS wxy_db "
       "BUFFER 64 "
       "CACHEMODEL 'last_value' "
@@ -256,7 +267,10 @@ TEST_F(ParserInitialCTest, createDatabase) {
       "WAL_RETENTION_PERIOD -1 "
       "WAL_RETENTION_SIZE -1 "
       "WAL_ROLL_PERIOD 10 "
-      "WAL_SEGMENT_SIZE 20");
+      "WAL_SEGMENT_SIZE 20 "
+      "SST_TRIGGER 16 "
+      "TABLE_PREFIX 3"
+      "TABLE_SUFFIX 4");
   clearCreateDbReq();
 
   setCreateDbReqFunc("wxy_db", 1);
@@ -266,6 +280,14 @@ TEST_F(ParserInitialCTest, createDatabase) {
       "DURATION 100m "
       "KEEP 1440m,300h,400d ");
   clearCreateDbReq();
+
+  setCreateDbReqFunc("wxy_db", 1);
+  setDbReplicaFunc(3);
+  setDbWalRetentionPeriod(TSDB_REPS_DEF_DB_WAL_RET_PERIOD);
+  setDbWalRetentionSize(TSDB_REPS_DEF_DB_WAL_RET_SIZE);
+  setDbWalRollPeriod(TSDB_REPS_DEF_DB_WAL_ROLL_PERIOD);
+  run("CREATE DATABASE IF NOT EXISTS wxy_db REPLICA 3");
+  clearCreateDbReq();
 }
 
 TEST_F(ParserInitialCTest, createDatabaseSemanticCheck) {

source/libs/parser/test/parShowToUse.cpp

@@ -218,7 +218,13 @@ TEST_F(ParserShowToUseTest, showVgroups) {
   run("SHOW test.vgroups");
 }
 
-// todo SHOW vnodes
+TEST_F(ParserShowToUseTest, showVnodes) {
+  useDb("root", "test");
+
+  run("SHOW VNODES 1");
+
+  run("SHOW VNODES 'node1:7030'");
+}
 
 TEST_F(ParserShowToUseTest, splitVgroup) {
   useDb("root", "test");

source/libs/planner/src/planOptimizer.c

@@ -16,6 +16,7 @@
 #include "filter.h"
 #include "functionMgt.h"
 #include "planInt.h"
+#include "tglobal.h"
 #include "ttime.h"
 
 #define OPTIMIZE_FLAG_MASK(n) (1 << n)
@@ -2410,7 +2411,7 @@ static const SOptimizeRule optimizeRuleSet[] = {
 static const int32_t optimizeRuleNum = (sizeof(optimizeRuleSet) / sizeof(SOptimizeRule));
 
 static void dumpLogicSubplan(const char* pRuleName, SLogicSubplan* pSubplan) {
-  if (0 == (qDebugFlag & DEBUG_DEBUG)) {
+  if (!tsQueryPlannerTrace) {
     return;
   }
   char* pStr = NULL;

source/libs/planner/src/planPhysiCreater.c

@@ -1324,7 +1324,8 @@ static int32_t createSortPhysiNode(SPhysiPlanContext* pCxt, SNodeList* pChildren
 static int32_t createPartitionPhysiNode(SPhysiPlanContext* pCxt, SNodeList* pChildren,
                                         SPartitionLogicNode* pPartLogicNode, SPhysiNode** pPhyNode) {
   SPartitionPhysiNode* pPart =
-      (SPartitionPhysiNode*)makePhysiNode(pCxt, (SLogicNode*)pPartLogicNode, QUERY_NODE_PHYSICAL_PLAN_PARTITION);
+      (SPartitionPhysiNode*)makePhysiNode(pCxt, (SLogicNode*)pPartLogicNode,
+          pCxt->pPlanCxt->streamQuery ? QUERY_NODE_PHYSICAL_PLAN_STREAM_PARTITION : QUERY_NODE_PHYSICAL_PLAN_PARTITION);
   if (NULL == pPart) {
     return TSDB_CODE_OUT_OF_MEMORY;
   }

source/libs/planner/src/planSpliter.c

@@ -1427,7 +1427,7 @@ static const SSplitRule splitRuleSet[] = {
 static const int32_t splitRuleNum = (sizeof(splitRuleSet) / sizeof(SSplitRule));
 
 static void dumpLogicSubplan(const char* pRuleName, SLogicSubplan* pSubplan) {
-  if (0 == (qDebugFlag & DEBUG_DEBUG)) {
+  if (!tsQueryPlannerTrace) {
     return;
   }
   char* pStr = NULL;

source/libs/planner/src/planner.c

@@ -17,9 +17,10 @@
 
 #include "planInt.h"
 #include "scalar.h"
+#include "tglobal.h"
 
 static void dumpQueryPlan(SQueryPlan* pPlan) {
-  if (0 == (qDebugFlag & DEBUG_DEBUG)) {
+  if (!tsQueryPlannerTrace) {
     return;
   }
   char* pStr = NULL;