Merge branch '3.0' into feature/TD-14761

.gitignore

@@ -118,3 +118,4 @@ contrib/*
 !contrib/test
 sql
 debug*/
+.env
\ No newline at end of file

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 8a5e336
+        GIT_TAG 3c7dafe
         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/taosws-rs.git
-        GIT_TAG 648cc62
+        GIT_TAG 29424d5
         SOURCE_DIR "${TD_SOURCE_DIR}/tools/taosws-rs"
         BINARY_DIR ""
         #BUILD_IN_SOURCE TRUE

docs/examples/go/connect/afconn/main.go

@@ -2,6 +2,7 @@ package main
 
 import (
 	"fmt"
+	"log"
 
 	"github.com/taosdata/driver-go/v3/af"
 )
@@ -10,7 +11,7 @@ func main() {
 	conn, err := af.Open("localhost", "root", "taosdata", "", 6030)
 	defer conn.Close()
 	if err != nil {
-		fmt.Println("failed to connect, err:", err)
+		log.Fatalln("failed to connect, err:", err)
 	} else {
 		fmt.Println("connected")
 	}

docs/examples/go/connect/cgoexample/main.go

@@ -3,6 +3,7 @@ package main
 import (
 	"database/sql"
 	"fmt"
+	"log"
 
 	_ "github.com/taosdata/driver-go/v3/taosSql"
 )
@@ -11,7 +12,7 @@ func main() {
 	var taosDSN = "root:taosdata@tcp(localhost:6030)/"
 	taos, err := sql.Open("taosSql", taosDSN)
 	if err != nil {
-		fmt.Println("failed to connect TDengine, err:", err)
+		log.Fatalln("failed to connect TDengine, err:", err)
 		return
 	}
 	fmt.Println("Connected")

docs/examples/go/connect/restexample/main.go

@@ -3,6 +3,7 @@ package main
 import (
 	"database/sql"
 	"fmt"
+	"log"
 
 	_ "github.com/taosdata/driver-go/v3/taosRestful"
 )
@@ -11,7 +12,7 @@ func main() {
 	var taosDSN = "root:taosdata@http(localhost:6041)/"
 	taos, err := sql.Open("taosRestful", taosDSN)
 	if err != nil {
-		fmt.Println("failed to connect TDengine, err:", err)
+		log.Fatalln("failed to connect TDengine, err:", err)
 		return
 	}
 	fmt.Println("Connected")

docs/examples/go/connect/wrapper/main.go

-package main
-
-import (
-	"fmt"
-
-	"github.com/taosdata/driver-go/v3/wrapper"
-)
-
-func main() {
-	conn, err := wrapper.TaosConnect("localhost", "root", "taosdata", "", 6030)
-	defer wrapper.TaosClose(conn)
-	if err != nil {
-		fmt.Println("fail to connect, err:", err)
-	} else {
-		fmt.Println("connected")
-	}
-}

docs/examples/go/insert/json/main.go

 package main
 
 import (
-	"fmt"
+	"log"
 
 	"github.com/taosdata/driver-go/v3/af"
 )
@@ -20,7 +20,7 @@ func prepareDatabase(conn *af.Connector) {
 func main() {
 	conn, err := af.Open("localhost", "root", "taosdata", "", 6030)
 	if err != nil {
-		fmt.Println("fail to connect, err:", err)
+		log.Fatalln("fail to connect, err:", err)
 	}
 	defer conn.Close()
 	prepareDatabase(conn)
@@ -32,6 +32,6 @@ func main() {
 
 	err = conn.OpenTSDBInsertJsonPayload(payload)
 	if err != nil {
-		fmt.Println("insert error:", err)
+		log.Fatalln("insert error:", err)
 	}
 }

docs/examples/go/insert/line/main.go

@@ -2,6 +2,7 @@ package main
 
 import (
 	"fmt"
+	"log"
 
 	"github.com/taosdata/driver-go/v3/af"
 )
@@ -33,6 +34,6 @@ func main() {
 
 	err = conn.InfluxDBInsertLines(lines, "ms")
 	if err != nil {
-		fmt.Println("insert error:", err)
+		log.Fatalln("insert error:", err)
 	}
 }

docs/examples/go/insert/sql/main.go

@@ -3,6 +3,7 @@ package main
 import (
 	"database/sql"
 	"fmt"
+	"log"
 
 	_ "github.com/taosdata/driver-go/v3/taosRestful"
 )
@@ -10,28 +11,26 @@ import (
 func createStable(taos *sql.DB) {
 	_, err := taos.Exec("CREATE DATABASE power")
 	if err != nil {
-		fmt.Println("failed to create database, err:", err)
+		log.Fatalln("failed to create database, err:", err)
 	}
 	_, err = taos.Exec("CREATE STABLE power.meters (ts TIMESTAMP, current FLOAT, voltage INT, phase FLOAT) TAGS (location BINARY(64), groupId INT)")
 	if err != nil {
-		fmt.Println("failed to create stable, err:", err)
+		log.Fatalln("failed to create stable, err:", err)
 	}
 }
 
 func insertData(taos *sql.DB) {
-	sql := `INSERT INTO power.d1001 USING power.meters TAGS(California.SanFrancisco, 2) VALUES ('2018-10-03 14:38:05.000', 10.30000, 219, 0.31000) ('2018-10-03 14:38:15.000', 12.60000, 218, 0.33000) ('2018-10-03 14:38:16.800', 12.30000, 221, 0.31000)
-	power.d1002 USING power.meters TAGS(California.SanFrancisco, 3) VALUES ('2018-10-03 14:38:16.650', 10.30000, 218, 0.25000)
-	power.d1003 USING power.meters TAGS(California.LosAngeles, 2) VALUES ('2018-10-03 14:38:05.500', 11.80000, 221, 0.28000) ('2018-10-03 14:38:16.600', 13.40000, 223, 0.29000)
-	power.d1004 USING power.meters TAGS(California.LosAngeles, 3) VALUES ('2018-10-03 14:38:05.000', 10.80000, 223, 0.29000) ('2018-10-03 14:38:06.500', 11.50000, 221, 0.35000)`
+	sql := `INSERT INTO power.d1001 USING power.meters TAGS('California.SanFrancisco', 2) VALUES ('2018-10-03 14:38:05.000', 10.30000, 219, 0.31000) ('2018-10-03 14:38:15.000', 12.60000, 218, 0.33000) ('2018-10-03 14:38:16.800', 12.30000, 221, 0.31000)
+	power.d1002 USING power.meters TAGS('California.SanFrancisco', 3) VALUES ('2018-10-03 14:38:16.650', 10.30000, 218, 0.25000)
+	power.d1003 USING power.meters TAGS('California.LosAngeles', 2) VALUES ('2018-10-03 14:38:05.500', 11.80000, 221, 0.28000) ('2018-10-03 14:38:16.600', 13.40000, 223, 0.29000)
+	power.d1004 USING power.meters TAGS('California.LosAngeles', 3) VALUES ('2018-10-03 14:38:05.000', 10.80000, 223, 0.29000) ('2018-10-03 14:38:06.500', 11.50000, 221, 0.35000)`
 	result, err := taos.Exec(sql)
 	if err != nil {
-		fmt.Println("failed to insert, err:", err)
-		return
+		log.Fatalln("failed to insert, err:", err)
 	}
 	rowsAffected, err := result.RowsAffected()
 	if err != nil {
-		fmt.Println("failed to get affected rows, err:", err)
-		return
+		log.Fatalln("failed to get affected rows, err:", err)
 	}
 	fmt.Println("RowsAffected", rowsAffected)
 }
@@ -40,8 +39,7 @@ func main() {
 	var taosDSN = "root:taosdata@http(localhost:6041)/"
 	taos, err := sql.Open("taosRestful", taosDSN)
 	if err != nil {
-		fmt.Println("failed to connect TDengine, err:", err)
-		return
+		log.Fatalln("failed to connect TDengine, err:", err)
 	}
 	defer taos.Close()
 	createStable(taos)

docs/examples/go/insert/stmt/main.go

@@ -5,8 +5,8 @@ import (
 	"time"
 
 	"github.com/taosdata/driver-go/v3/af"
-	"github.com/taosdata/driver-go/v3/af/param"
 	"github.com/taosdata/driver-go/v3/common"
+	"github.com/taosdata/driver-go/v3/common/param"
 )
 
 func checkErr(err error, prompt string) {

docs/examples/go/insert/telnet/main.go

 package main
 
 import (
-	"fmt"
+	"log"
 
 	"github.com/taosdata/driver-go/v3/af"
 )
@@ -20,7 +20,7 @@ func prepareDatabase(conn *af.Connector) {
 func main() {
 	conn, err := af.Open("localhost", "root", "taosdata", "", 6030)
 	if err != nil {
-		fmt.Println("fail to connect, err:", err)
+		log.Fatalln("fail to connect, err:", err)
 	}
 	defer conn.Close()
 	prepareDatabase(conn)
@@ -37,6 +37,6 @@ func main() {
 
 	err = conn.OpenTSDBInsertTelnetLines(lines)
 	if err != nil {
-		fmt.Println("insert error:", err)
+		log.Fatalln("insert error:", err)
 	}
 }

docs/examples/go/query/sync/main.go

@@ -2,7 +2,7 @@ package main
 
 import (
 	"database/sql"
-	"fmt"
+	"log"
 	"time"
 
 	_ "github.com/taosdata/driver-go/v3/taosRestful"
@@ -12,14 +12,12 @@ func main() {
 	var taosDSN = "root:taosdata@http(localhost:6041)/power"
 	taos, err := sql.Open("taosRestful", taosDSN)
 	if err != nil {
-		fmt.Println("failed to connect TDengine, err:", err)
-		return
+		log.Fatalln("failed to connect TDengine, err:", err)
 	}
 	defer taos.Close()
 	rows, err := taos.Query("SELECT ts, current FROM meters LIMIT 2")
 	if err != nil {
-		fmt.Println("failed to select from table, err:", err)
-		return
+		log.Fatalln("failed to select from table, err:", err)
 	}
 
 	defer rows.Close()
@@ -30,9 +28,9 @@ func main() {
 		}
 		err := rows.Scan(&r.ts, &r.current)
 		if err != nil {
-			fmt.Println("scan error:\n", err)
+			log.Fatalln("scan error:\n", err)
 			return
 		}
-		fmt.Println(r.ts, r.current)
+		log.Fatalln(r.ts, r.current)
 	}
 }

docs/examples/node/nativeexample/async_query_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 const conn = taos.connect({ host: "localhost", database: "power" });
 const cursor = conn.cursor();
 
@@ -18,4 +18,3 @@ try {
     conn.close();
   }, 2000);
 }
-// bug here: jira 14506

docs/examples/node/nativeexample/connect.js

-const taos = require("td2.0-connector");
+const { options, connect } = require("@tdengine/rest");
 
-var conn = taos.connect({
-  host: "localhost",
-  port: 6030,
-  user: "root",
-  password: "taosdata",
-});
-conn.close();
+async function test() {
+  options.path = "/rest/sql";
+  options.host = "localhost";
+  let conn = connect(options);
+  let cursor = conn.cursor();
+  try {
+    let res = await cursor.query("SELECT server_version()");
+    res.toString();
+  } catch (err) {
+    console.log(err);
+  }
+}
+test();
 
-// run with: node connect.js
 // output:
-// Successfully connected to TDengine
+// server_version() |
+// ===================
+// 3.0.0.0          |

docs/examples/node/nativeexample/influxdb_line_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",

docs/examples/node/nativeexample/insert_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",
@@ -11,11 +11,11 @@ try {
   cursor.execute(
     "CREATE STABLE meters (ts TIMESTAMP, current FLOAT, voltage INT, phase FLOAT) TAGS (location BINARY(64), groupId INT)"
   );
-  var sql = `INSERT INTO power.d1001 USING power.meters TAGS(California.SanFrancisco, 2) VALUES ('2018-10-03 14:38:05.000', 10.30000, 219, 0.31000) ('2018-10-03 14:38:15.000', 12.60000, 218, 0.33000) ('2018-10-03 14:38:16.800', 12.30000, 221, 0.31000)
-power.d1002 USING power.meters TAGS(California.SanFrancisco, 3) VALUES ('2018-10-03 14:38:16.650', 10.30000, 218, 0.25000)
-power.d1003 USING power.meters TAGS(California.LosAngeles, 2) VALUES ('2018-10-03 14:38:05.500', 11.80000, 221, 0.28000) ('2018-10-03 14:38:16.600', 13.40000, 223, 0.29000)
-power.d1004 USING power.meters TAGS(California.LosAngeles, 3) VALUES ('2018-10-03 14:38:05.000', 10.80000, 223, 0.29000) ('2018-10-03 14:38:06.500', 11.50000, 221, 0.35000)`;
-  cursor.execute(sql);
+  var sql = `INSERT INTO power.d1001 USING power.meters TAGS('California.SanFrancisco', 2) VALUES ('2018-10-03 14:38:05.000', 10.30000, 219, 0.31000) ('2018-10-03 14:38:15.000', 12.60000, 218, 0.33000) ('2018-10-03 14:38:16.800', 12.30000, 221, 0.31000)
+power.d1002 USING power.meters TAGS('California.SanFrancisco', 3) VALUES ('2018-10-03 14:38:16.650', 10.30000, 218, 0.25000)
+power.d1003 USING power.meters TAGS('California.LosAngeles', 2) VALUES ('2018-10-03 14:38:05.500', 11.80000, 221, 0.28000) ('2018-10-03 14:38:16.600', 13.40000, 223, 0.29000)
+power.d1004 USING power.meters TAGS('California.LosAngeles', 3) VALUES ('2018-10-03 14:38:05.000', 10.80000, 223, 0.29000) ('2018-10-03 14:38:06.500', 11.50000, 221, 0.35000)`;
+  cursor.execute(sql,{'quiet':false});
 } finally {
   cursor.close();
   conn.close();

docs/examples/node/nativeexample/multi_bind_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",
@@ -24,10 +24,10 @@ function insertData() {
   );
 
   // bind table name and tags
-  let tagBind = new taos.TaosBind(2);
-  tagBind.bindBinary("California.SanFrancisco");
-  tagBind.bindInt(2);
-  cursor.stmtSetTbnameTags("d1001", tagBind.getBind());
+  let tagBind = new taos.TaosMultiBindArr(2);
+  tagBind.multiBindBinary(["California.SanFrancisco"]);
+  tagBind.multiBindInt([2]);
+  cursor.stmtSetTbnameTags("d1001", tagBind.getMultiBindArr());
 
   // bind values
   let valueBind = new taos.TaosMultiBindArr(4);

docs/examples/node/nativeexample/opentsdb_json_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",

docs/examples/node/nativeexample/opentsdb_telnet_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",

docs/examples/node/nativeexample/param_bind_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({
   host: "localhost",
@@ -23,25 +23,22 @@ function insertData() {
   );
 
   // bind table name and tags
-  let tagBind = new taos.TaosBind(2);
-  tagBind.bindBinary("California.SanFrancisco");
-  tagBind.bindInt(2);
-  cursor.stmtSetTbnameTags("d1001", tagBind.getBind());
+  let tagBind = new taos.TaosMultiBindArr(2);
+  tagBind.multiBindBinary(["California.SanFrancisco"]);
+  tagBind.multiBindInt([2]);
+  cursor.stmtSetTbnameTags("d1001", tagBind.getMultiBindArr());
 
   // bind values
-  let rows = [
-    [1648432611249, 10.3, 219, 0.31],
-    [1648432611749, 12.6, 218, 0.33],
-  ];
-  for (let row of rows) {
-    let valueBind = new taos.TaosBind(4);
-    valueBind.bindTimestamp(row[0]);
-    valueBind.bindFloat(row[1]);
-    valueBind.bindInt(row[2]);
-    valueBind.bindFloat(row[3]);
-    cursor.stmtBindParam(valueBind.getBind());
+  let rows = [[1648432611249, 1648432611749], [10.3, 12.6], [219, 218], [0.31, 0.33]];
+
+  let valueBind = new taos.TaosMultiBindArr(4);
+  valueBind.multiBindTimestamp(rows[0]);
+  valueBind.multiBindFloat(rows[1]);
+  valueBind.multiBindInt(rows[2]);
+  valueBind.multiBindFloat(rows[3]);
+  cursor.stmtBindParamBatch(valueBind.getMultiBindArr());
   cursor.stmtAddBatch();
-  }
+
 
   // execute
   cursor.stmtExecute();

docs/examples/node/nativeexample/query_example.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({ host: "localhost", database: "power" });
 const cursor = conn.cursor();
@@ -9,8 +9,6 @@ query.execute().then(function (result) {
 
 // output:
 // Successfully connected to TDengine
-// Query OK, 2 row(s) in set (0.00317767s)
-
 //            ts             |         current          |
 // =======================================================
 // 2018-10-03 14:38:05.000   | 10.3                     |

docs/examples/node/nativeexample/subscribe_demo.js

-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 
 const conn = taos.connect({ host: "localhost", database: "power" });
-// 未完成
+var cursor = conn.cursor();
\ No newline at end of file
+
+function runConsumer() {
+
+    // create topic 
+    cursor.execute("create topic topic_name_example as select * from meters");
+
+    let consumer = taos.consumer({
+        'group.id': 'tg2',
+        'td.connect.user': 'root',
+        'td.connect.pass': 'taosdata',
+        'msg.with.table.name': 'true',
+        'enable.auto.commit': 'true'
+    });
+    
+    // subscribe the topic just created.
+    consumer.subscribe("topic_name_example");
+
+    // get subscribe topic list
+    let topicList = consumer.subscription();
+    console.log(topicList);
+
+    for (let i = 0; i < 5; i++) {
+        let msg = consumer.consume(100);
+        console.log(msg.topicPartition);
+        console.log(msg.block);
+        console.log(msg.fields)
+        consumer.commit(msg);
+        console.log(`=======consumer ${i} done`)
+    }
+
+    consumer.unsubscribe();
+    consumer.close();
+
+    // drop topic
+    cursor.execute("drop topic topic_name_example");
+}
+
+
+try {
+    runConsumer();
+} finally {
+
+    setTimeout(() => {
+        cursor.close();
+        conn.close();
+    }, 2000);
+}
\ No newline at end of file

docs/examples/node/package.json

@@ -4,7 +4,7 @@
   "main": "index.js",
   "license": "MIT",
   "dependencies": {
-    "td2.0-connector": "^2.0.12",
-    "td2.0-rest-connector": "^1.0.0"
+    "@tdengine/client": "^3.0.0",
+    "@tdengine/rest": "^3.0.0"
   }
 }

docs/examples/node/restexample/connect.js

-const { options, connect } = require("td2.0-rest-connector");
+const { options, connect } = require("@tdengine/rest");
 
 async function test() {
   options.path = "/rest/sqlt";
@@ -17,4 +17,4 @@ test();
 // output:
 // server_version() |
 // ===================
-// 2.4.0.12         |
+// 3.0.0.0          |

docs/zh/07-develop/03-insert-data/_go_stmt.mdx

@@ -3,6 +3,6 @@
 ```
 
 :::tip
-driver-go 的模块 `github.com/taosdata/driver-go/v2/wrapper` 是 C 接口的底层封装。使用这个模块也可以实现参数绑定写入。
+driver-go 的模块 `github.com/taosdata/driver-go/v3/wrapper` 是 C 接口的底层封装。使用这个模块也可以实现参数绑定写入。
 
 :::

docs/zh/07-develop/07-tmq.md

 ---
-sidebar_label: 消息队列
-description: "数据订阅与推送服务。连续写入到 TDengine 中的时序数据能够被自动推送到订阅客户端。"
-title: 消息队列
+sidebar_label: 数据订阅
+description: "数据订阅与推送服务。写入到 TDengine 中的时序数据能够被自动推送到订阅客户端。"
+title: 数据订阅
 ---
 
-基于数据天然的时间序列特性，TDengine 的数据写入（insert）与消息系统的数据发布（pub）逻辑上一致，均可视为系统中插入一条带时间戳的新记录。同时，TDengine 在内部严格按照数据时间序列单调递增的方式保存数据。本质上来说，TDengine 中每一张表均可视为一个标准的消息队列。
+为了帮助应用实时获取写入 TDengine 的数据，或者以事件到达顺序处理数据，TDengine提供了类似消息队列产品的数据订阅、消费接口。这样在很多场景下，采用 TDengine 的时序数据处理系统不再需要集成消息队列产品，比如 kafka, 从而简化系统设计的复杂度，降低运营维护成本。
 
-TDengine 内嵌支持消息订阅与推送服务（下文都简称TMQ）。使用系统提供的 API，用户可使用普通查询语句订阅数据库中的一张或多张表，或整个库。客户端启动订阅后，定时或按需轮询服务器是否有新的记录到达，有新的记录到达就会将结果反馈到客户。
+与 kafka 一样，你需要定义 topic, 但 TDengine 的 topic 可以是一张超级表，或一张子表。不仅如此，你可以通过标签、表名、列、表达式等多种方法过滤所需数据，并且支持对数据进行函数变换、预处理（包括标量udf计算）。与其他消息队列软件相比，这是 TDengine 数据订阅功能的最大的优势，它提供了更大的灵活性，数据的颗粒度可以由应用随时调整，而且数据的过滤交给 TDengine，而不是应用完成，有效的减少传输的数据量。
 
-TMQ提供了提交机制来保证消息队列的可靠性和正确性。在调用方法上，支持自动提交和手动提交。
+消费者订阅 topic 后，可以实时获得最新的数据。多个消费者可以组成一个消费者组 (consumer group), 一个消费者组里的多个消费者共享消费进度，便于多线程分布式的消费数据，提高数据通吐率。但不同消费者组即使消费同一个topic, 并不共享消费进度。一个消费者组可以订阅多个 topic。如果订阅的是超级表，数据可能会分布在多个不同的vnode上，也就是多个shard上，这样一个消费组里有多个消费者可以提高消费效率。TDengine 的消息队列提供了消息的ACK机制，在宕机、重启等复杂环境下确保at least once消费。
+
+为了实现上述功能，TDengine 采用了灵活的 WAL (Write-Ahead-Log) 文件切换与保留机制：可以按照时间或文件大小来保留WAL文件（详见create database语句）。在消费时，TDengine 从 WAL 中获取数据，并经过过滤、变换等操作，将数据推送给消费者。
+
+本文档不对消息队列本身的基础知识做介绍，如果需要了解，请自行搜索。
+
+## 主要数据结构和API
 
 TMQ 的 API 中，与订阅相关的主要数据结构和API如下：
 
@@ -47,7 +53,9 @@ DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_comm
 
 这些 API 的文档请见 [C/C++ Connector](/reference/connector/cpp)，下面介绍一下它们的具体用法（超级表和子表结构请参考“数据建模”一节），完整的示例代码可以在 [tmq.c](https://github.com/taosdata/TDengine/blob/3.0/examples/c/tmq.c) 看到。
 
-一、首先完成建库、建一张超级表和多张子表，并每个子表插入若干条数据记录：
+## 写入数据
+
+首先完成建库、建一张超级表和多张子表操作，然后就可以写入数据了，比如：
 
 ```sql
 drop database if exists tmqdb;
@@ -63,14 +71,15 @@ insert into tmqdb.ctb2 values(now, 2, 2, 'a1')(now+1s, 22, 22, 'a22');
 insert into tmqdb.ctb3 values(now, 3, 3, 'a1')(now+1s, 33, 33, 'a33');
 ```
 
-二、创建topic：
+## 创建topic：
 
 ```sql
 create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
 ```
 
-注：TMQ支持多种订阅类型：
-1、列订阅
+TMQ支持多种订阅类型：
+
+### 列订阅
 
 语法：CREATE TOPIC topic_name as subquery
 通过select语句订阅（包括select *，或select ts, c1等指定列描述订阅，可以带条件过滤、标量函数计算，但不支持聚合函数、不支持时间窗口聚合）
@@ -79,25 +88,18 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
 - 被订阅或用于计算的column和tag不可被删除、修改
 - 若发生schema变更，新增的column不出现在结果中
 
-2、超级表订阅
+### 超级表订阅
 语法：CREATE TOPIC topic_name AS STABLE stbName
 
-- 订阅某超级表的全部数据，schema变更不受限，schema变更后写入的数据将以最新schema返回
-- 在tmq的返回消息中schema是块级别的，每块的schema可能不一样
-- 列变更后写入的数据若未落盘，将以写入时的schema返回
-- 列变更后写入的数据若已落盘，将以落盘时的schema返回
-
-3、db订阅
-语法：CREATE TOPIC topic_name AS DATABASE db_name
-
-- 订阅某一db的全部数据，schema变更不受限
-- 在tmq的返回消息中schema是块级别的，每块的schema可能不一样
-- 列变更后写入的数据若未落盘，将以写入时的schema返回
-- 列变更后写入的数据若已落盘，将以落盘时的schema返回
+与select * from stbName订阅的区别是：
+- 不会限制用户的schema变更
+- 返回的是非结构化的数据：返回数据的schema会随之超级表的schema变化而变化
+- 用户对于要处理的每一个数据块都可能有不同的schema，因此，必须重新获取schema
+- 返回数据不带有tag
 
-三、创建consumer
+## 创建 consumer 以及consumer group
 
-目前支持的config：
+对于consumer, 目前支持的config包括：
 
 | 参数名称                     | 参数值                         | 备注                                                   |
 | ---------------------------- | ------------------------------ | ------------------------------------------------------ |
@@ -131,7 +133,12 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
   return tmq;
 ```
 
-四、创建订阅主题列表
+上述配置中包括consumer group ID，如果多个 consumer 指定的 consumer group ID一样，则自动形成一个consumer group，共享消费进度。
+
+
+## 创建 topic 列表
+
+单个consumer支持同时订阅多个topic。
 
 ```sql
   tmq_list_t* topicList = tmq_list_new();
@@ -139,9 +146,7 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
   return topicList;
 ```
 
-单个consumer支持同时订阅多个topic。
-
-五、启动订阅并开始消费
+## 启动订阅并开始消费
 
 ```sql
   /* 启动订阅 */
@@ -151,9 +156,9 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
   /* 循环poll消息 */
   int32_t totalRows = 0;
   int32_t msgCnt = 0;
-  int32_t consumeDelay = 5000;
+  int32_t timeOut = 5000;
   while (running) {
-    TAOS_RES* tmqmsg = tmq_consumer_poll(tmq, consumeDelay);
+    TAOS_RES* tmqmsg = tmq_consumer_poll(tmq, timeOut);
     if (tmqmsg) {
       msgCnt++;
       totalRows += msg_process(tmqmsg);
@@ -199,7 +204,7 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
 }
 ```
 
-五、结束消费
+## 结束消费
 
 ```sql
   /* 取消订阅 */
@@ -209,7 +214,7 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
   tmq_consumer_close(tmq);
 ```
 
-六、删除topic
+## 删除topic
 
 如果不再需要，可以删除创建topic，但注意：只有没有被订阅的topic才能别删除。
 
@@ -218,7 +223,7 @@ create topic topicName as select ts, c1, c2, c3 from tmqdb.stb where c1 > 1;
   drop topic topicName;
 ```
 
-七、状态查看
+## 状态查看
 
 1、topics：查询已经创建的topic
 

docs/zh/07-develop/09-udf.md

@@ -124,52 +124,49 @@ gcc -g -O0 -fPIC -shared add_one.c -o add_one.so
 
 用户可以通过 SQL 指令在系统中加载客户端所在主机上的 UDF 函数库（不能通过 RESTful 接口或 HTTP 管理界面来进行这一过程）。一旦创建成功，则当前 TDengine 集群的所有用户都可以在 SQL 指令中使用这些函数。UDF 存储在系统的 MNode 节点上，因此即使重启 TDengine 系统，已经创建的 UDF 也仍然可用。
 
-在创建 UDF 时，需要区分标量函数和聚合函数。如果创建时声明了错误的函数类别，则可能导致通过 SQL 指令调用函数时出错。此外， UDF 支持输入与输出类型不一致，用户需要保证输入数据类型与 UDF 程序匹配，UDF 输出数据类型与 OUTPUTTYPE 匹配。
+在创建 UDF 时，需要区分标量函数和聚合函数。如果创建时声明了错误的函数类别，则可能导致通过 SQL 指令调用函数时出错。此外，用户需要保证输入数据类型与 UDF 程序匹配，UDF 输出数据类型与 OUTPUTTYPE 匹配。
 
 - 创建标量函数
 ```sql
-CREATE FUNCTION ids(X) AS ids(Y) OUTPUTTYPE typename(Z) [ BUFSIZE B ];
+CREATE FUNCTION function_name AS library_path OUTPUTTYPE output_type;
 ```
 
-  - ids(X)：标量函数未来在 SQL 指令中被调用时的函数名，必须与函数实现中 udfNormalFunc 的实际名称一致；
-  - ids(Y)：包含 UDF 函数实现的动态链接库的库文件绝对路径（指的是库文件在当前客户端所在主机上的保存路径，通常是指向一个 .so 文件），这个路径需要用英文单引号或英文双引号括起来；
-  - typename(Z)：此函数计算结果的数据类型，与上文中 udfNormalFunc 的 itype 参数不同，这里不是使用数字表示法，而是直接写类型名称即可；
-  - B：中间计算结果的缓冲区大小，单位是字节，最小 0，最大 512，如果不使用可以不设置。
+  - function_name：标量函数未来在 SQL 中被调用时的函数名，必须与函数实现中 udf 的实际名称一致；
+  - library_path：包含 UDF 函数实现的动态链接库的库文件绝对路径（指的是库文件在当前客户端所在主机上的保存路径，通常是指向一个 .so 文件），这个路径需要用英文单引号或英文双引号括起来；
+  - output_type：此函数计算结果的数据类型名称；
 
-  例如，如下语句可以把 add_one.so 创建为系统中可用的 UDF：
+  例如，如下语句可以把 libbitand.so 创建为系统中可用的 UDF：
 
   ```sql
-  CREATE FUNCTION add_one AS "/home/taos/udf_example/add_one.so" OUTPUTTYPE INT;
+  CREATE FUNCTION bit_and AS "/home/taos/udf_example/libbitand.so" OUTPUTTYPE INT;
   ```
 
 - 创建聚合函数：
 ```sql
-CREATE AGGREGATE FUNCTION ids(X) AS ids(Y) OUTPUTTYPE typename(Z) [ BUFSIZE B ];
+CREATE AGGREGATE FUNCTION function_name AS library_path OUTPUTTYPE output_type [ BUFSIZE buffer_size ];
 ```
 
-  - ids(X)：聚合函数未来在 SQL 指令中被调用时的函数名，必须与函数实现中 udfNormalFunc 的实际名称一致；
-  - ids(Y)：包含 UDF 函数实现的动态链接库的库文件绝对路径（指的是库文件在当前客户端所在主机上的保存路径，通常是指向一个 .so 文件），这个路径需要用英文单引号或英文双引号括起来；
-  - typename(Z)：此函数计算结果的数据类型，与上文中 udfNormalFunc 的 itype 参数不同，这里不是使用数字表示法，而是直接写类型名称即可；
-  - B：中间计算结果的缓冲区大小，单位是字节，最小 0，最大 512，如果不使用可以不设置。
+  - function_name：聚合函数未来在 SQL 中被调用时的函数名，必须与函数实现中 udfNormalFunc 的实际名称一致；
+  - library_path：包含 UDF 函数实现的动态链接库的库文件绝对路径（指的是库文件在当前客户端所在主机上的保存路径，通常是指向一个 .so 文件），这个路径需要用英文单引号或英文双引号括起来；
+  - output_type：此函数计算结果的数据类型，与上文中 udfNormalFunc 的 itype 参数不同，这里不是使用数字表示法，而是直接写类型名称即可；
+  - buffer_size：中间计算结果的缓冲区大小，单位是字节。如果不使用可以不设置。
 
-  关于中间计算结果的使用，可以参考示例程序[demo.c](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/demo.c)
-
-  例如，如下语句可以把 demo.so 创建为系统中可用的 UDF：
+  例如，如下语句可以把 libsqrsum.so 创建为系统中可用的 UDF：
 
   ```sql
-  CREATE AGGREGATE FUNCTION demo AS "/home/taos/udf_example/demo.so" OUTPUTTYPE DOUBLE bufsize 14;
+  CREATE AGGREGATE FUNCTION sqr_sum AS "/home/taos/udf_example/libsqrsum.so" OUTPUTTYPE DOUBLE bufsize 8;
   ```
 
 ### 管理 UDF
 
 - 删除指定名称的用户定义函数：
 ```
-DROP FUNCTION ids(X);
+DROP FUNCTION function_name;
 ```
 
-- ids(X)：此参数的含义与 CREATE 指令中的 ids(X) 参数一致，也即要删除的函数的名字，例如 
+- function_name：此参数的含义与 CREATE 指令中的 function_name 参数一致，也即要删除的函数的名字，例如 
 ```sql
-DROP FUNCTION add_one;
+DROP FUNCTION bit_and;
 ```
 - 显示系统中当前可用的所有 UDF：
 ```sql
@@ -180,53 +177,32 @@ SHOW FUNCTIONS;
 
 在 SQL 指令中，可以直接以在系统中创建 UDF 时赋予的函数名来调用用户定义函数。例如：
 ```sql
-SELECT X(c) FROM table/stable;
+SELECT X(c1,c2) FROM table/stable;
 ```
 
-表示对名为 c 的数据列调用名为 X 的用户定义函数。SQL 指令中用户定义函数可以配合 WHERE 等查询特性来使用。
-
-## UDF 的一些使用限制
-
-在当前版本下，使用 UDF 存在如下这些限制：
+表示对名为 c1, c2 的数据列调用名为 X 的用户定义函数。SQL 指令中用户定义函数可以配合 WHERE 等查询特性来使用。
 
-1. 在创建和调用 UDF 时，服务端和客户端都只支持 Linux 操作系统；
-2. UDF 不能与系统内建的 SQL 函数混合使用，暂不支持在一条 SQL 语句中使用多个不同名的 UDF ；
-3. UDF 只支持以单个数据列作为输入；
-4. UDF 只要创建成功，就会被持久化存储到 MNode 节点中；
-5. 无法通过 RESTful 接口来创建 UDF；
-6. UDF 在 SQL 中定义的函数名，必须与 .so 库文件实现中的接口函数名前缀保持一致，也即必须是 udfNormalFunc 的名称，而且不可与 TDengine 中已有的内建 SQL 函数重名。
 
 ## 示例代码
 
-### 标量函数示例 [add_one](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/add_one.c)
-
-<details>
-<summary>add_one.c</summary>
-
-```c
-{{#include tests/script/sh/add_one.c}}
-```
-
-</details>
-
-### 向量函数示例 [abs_max](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/abs_max.c)
+### 标量函数示例 [bit_and](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/bit_and.c)
 
 <details>
-<summary>abs_max.c</summary>
+<summary>bit_and.c</summary>
 
 ```c
-{{#include tests/script/sh/abs_max.c}}
+{{#include tests/script/sh/bit_and.c}}
 ```
 
 </details>
 
-### 使用中间计算结果示例 [demo](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/demo.c)
+### 聚合函数示例 [sqr_sum](https://github.com/taosdata/TDengine/blob/develop/tests/script/sh/sqr_sum.c)
 
 <details>
-<summary>demo.c</summary>
+<summary>sqr_sum.c</summary>
 
 ```c
-{{#include tests/script/sh/demo.c}}
+{{#include tests/script/sh/sqr_sum.c}}
 ```
 
 </details>

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

@@ -8,21 +8,30 @@ title: 用户自定义函数
 ## 创建函数
 
 ```sql
-CREATE [AGGREGATE] FUNCTION func_name AS library_path OUTPUTTYPE type_name [BUFSIZE value]
+CREATE [AGGREGATE] FUNCTION func_name AS library_path OUTPUTTYPE type_name [BUFSIZE buffer_size]
 ```
 
 语法说明：
 
 AGGREGATE：标识此函数是标量函数还是聚集函数。
-func_name：函数名，必须与函数实现中udfNormalFunc的实际名称一致。
+func_name：函数名，必须与函数实现中 udf 的实际名称一致。
 library_path：包含UDF函数实现的动态链接库的绝对路径，是在客户端侧主机上的绝对路径。
-OUTPUTTYPE：标识此函数的返回类型。
-BUFSIZE：中间结果的缓冲区大小，单位是字节。不设置则默认为0。最大不可超过512字节。
+type_name：标识此函数的返回类型。
+buffer_size：中间结果的缓冲区大小，单位是字节。不设置则默认为0。
 
 关于如何开发自定义函数，请参考 [UDF使用说明](../../develop/udf)。
 
 ## 删除自定义函数
 
+```
+DROP FUNCTION function_name;
+```
+
+- function_name：此参数的含义与 CREATE 指令中的 function_name 参数一致，也即要删除的函数的名字，例如 
+
+
+## 显示 UDF
+
 ```sql
-DROP FUNCTION func_name
+SHOW FUNCTION;
 ```

docs/zh/14-reference/03-connector/node.mdx

@@ -15,11 +15,11 @@ import NodeOpenTSDBTelnet from "../../07-develop/03-insert-data/_js_opts_telnet.
 import NodeOpenTSDBJson from "../../07-develop/03-insert-data/_js_opts_json.mdx";
 import NodeQuery from "../../07-develop/04-query-data/_js.mdx";
 
-`td2.0-connector` 和 `td2.0-rest-connector` 是 TDengine 的官方 Node.js 语言连接器。Node.js 开发人员可以通过它开发可以存取 TDengine 集群数据的应用软件。
+`@tdengine/client` 和 `@tdengine/rest` 是 TDengine 的官方 Node.js 语言连接器。 Node.js 开发人员可以通过它开发可以存取 TDengine 集群数据的应用软件。注意：从 TDengine 3.0 开始 Node.js 原生连接器的包名由 `td2.0-connector` 改名为 `@tdengine/client` 而 rest 连接器的包名由 `td2.0-rest-connector` 改为 `@tdengine/rest`。并且不与 TDengine 2.x 兼容。
 
-`td2.0-connector` 是**原生连接器**，它通过 TDengine 客户端驱动程序（taosc）连接 TDengine 运行实例，支持数据写入、查询、订阅、schemaless 接口和参数绑定接口等功能。`td2.0-rest-connector` 是 **REST 连接器**，它通过 taosAdapter 提供的 REST 接口连接 TDengine 的运行实例。REST 连接器可以在任何平台运行，但性能略为下降，接口实现的功能特性集合和原生接口有少量不同。
+`@tdengine/client` 是**原生连接器**，它通过 TDengine 客户端驱动程序（taosc）连接 TDengine 运行实例，支持数据写入、查询、订阅、schemaless 接口和参数绑定接口等功能。`@tdengine/rest` 是 **REST 连接器**，它通过 taosAdapter 提供的 REST 接口连接 TDengine 的运行实例。REST 连接器可以在任何平台运行，但性能略为下降，接口实现的功能特性集合和原生接口有少量不同。
 
-Node.js 连接器源码托管在 [GitHub](https://github.com/taosdata/taos-connector-node)。
+Node.js 连接器源码托管在 [GitHub](https://github.com/taosdata/taos-connector-node/tree/3.0)。
 
 ## 支持的平台
 
@@ -58,7 +58,7 @@ REST 连接器支持所有能运行 Node.js 的平台。
 <TabItem value="Linux" label="Linux 系统安装依赖工具">
 
 - `python` (建议`v2.7` , `v3.x.x` 目前还不支持)
-- `td2.0-connector` 2.0.6 支持 Node.js LTS v10.9.0 或更高版本, Node.js LTS v12.8.0 或更高版本；2.0.5 及更早版本支持 Node.js LTS v10.x 版本。其他版本可能存在包兼容性的问题
+- `@tdengine/client` 3.0.0 支持 Node.js LTS v10.9.0 或更高版本, Node.js LTS v12.8.0 或更高版本；其他版本可能存在包兼容性的问题
 - `make`
 - C 语言编译器，[GCC](https://gcc.gnu.org) v4.8.5 或更高版本
 
@@ -90,14 +90,14 @@ REST 连接器支持所有能运行 Node.js 的平台。
 <TabItem value="install_native" label="安装原生连接器">
 
 ```bash
-npm install td2.0-connector
+npm install @tdengine/client
 ```
 
 </TabItem>
 <TabItem value="install_rest" label="安装 REST 连接器">
 
 ```bash
-npm i td2.0-rest-connector
+npm install @tdengine/rest
 ```
 
 </TabItem>
@@ -109,13 +109,13 @@ npm i td2.0-rest-connector
 
 验证方法：
 
-- 新建安装验证目录，例如：`~/tdengine-test`，下载 GitHub 上 [nodejsChecker.js 源代码](https://github.com/taosdata/TDengine/tree/develop/examples/nodejs/nodejsChecker.js)到本地。
+- 新建安装验证目录，例如：`~/tdengine-test`，下载 GitHub 上 [nodejsChecker.js 源代码](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/nodejsChecker.js)到本地。
 
 - 在命令行中执行以下命令。
 
 ```bash
 npm init -y
-npm install td2.0-connector
+npm install @tdengine/client
 node nodejsChecker.js host=localhost
 ```
 
@@ -128,11 +128,11 @@ node nodejsChecker.js host=localhost
 <Tabs defaultValue="native">
 <TabItem value="native" label="原生连接">
 
-安装并引用 `td2.0-connector` 包。
+安装并引用 `@tdengine/client` 包。
 
 ```javascript
 //A cursor also needs to be initialized in order to interact with TDengine from Node.js.
-const taos = require("td2.0-connector");
+const taos = require("@tdengine/client");
 var conn = taos.connect({
   host: "127.0.0.1",
   user: "root",
@@ -149,12 +149,12 @@ conn.close();
 </TabItem>
 <TabItem value="rest" label="REST 连接">
 
-安装并引用 `td2.0-rest-connector` 包。
+安装并引用 `@tdengine/rest` 包。
 
 ```javascript
 //A cursor also needs to be initialized in order to interact with TDengine from Node.js.
-import { options, connect } from "td2.0-rest-connector";
-options.path = "/rest/sqlt";
+import { options, connect } from "@tdengine/rest";
+options.path = "/rest/sql";
 // set host
 options.host = "localhost";
 // set other options like user/passwd
@@ -190,26 +190,23 @@ let cursor = conn.cursor();
 
 <NodeQuery />
 
+
 ## 更多示例程序
 
 | 示例程序                                                                                                                                   | 示例程序描述                           |
 | ------------------------------------------------------------------------------------------------------------------------------------------ | -------------------------------------- |
-| [connection](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/cursorClose.js)                                  | 建立连接的示例。                       |
-| [stmtBindBatch](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/stmtBindParamBatchSample.js)                  | 绑定多行参数插入的示例。               |
-| [stmtBind](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/stmtBindParamSample.js)                            | 一行一行绑定参数插入的示例。           |
-| [stmtBindSingleParamBatch](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/stmtBindSingleParamBatchSample.js) | 按列绑定参数插入的示例。               |
-| [stmtUseResult](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/stmtUseResultSample.js)                       | 绑定参数查询的示例。                   |
-| [json tag](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/testJsonTag.js)                                    | Json tag 的使用示例。                  |
-| [Nanosecond](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/testNanoseconds.js)                              | 时间戳为纳秒精度的使用的示例。         |
-| [Microsecond](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/testMicroseconds.js)                            | 时间戳为微秒精度的使用的示例。         |
-| [schemless insert](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/testSchemalessInsert.js)                   | schemless 插入的示例。                 |
-| [subscribe](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/testSubscribe.js)                                 | 订阅的使用示例。                       |
-| [asyncQuery](https://github.com/taosdata/taos-connector-node/tree/develop/nodejs/examples/tset.js)                                         | 异步查询的使用示例。                   |
-| [REST](https://github.com/taosdata/taos-connector-node/blob/develop/typescript-rest/example/example.ts)                                    | 使用 REST 连接的 TypeScript 使用示例。 |
+| [basicUse](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/queryExample.js)                                  | 基本的使用如如建立连接，执行 SQL 等操作。                       |
+| [stmtBindBatch](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/bindParamBatch.js)                  | 绑定多行参数插入的示例。               | |
+| [stmtBindSingleParamBatch](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/bindSingleParamBatch.js) | 按列绑定参数插入的示例。               |
+| [stmtQuery](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/stmtQuery.js)                       | 绑定参数查询的示例。                   |
+| [schemless insert](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/schemaless.js)                   | schemless 插入的示例。                 |
+| [TMQ](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/tmq.js)                                 | 订阅的使用示例。                       |
+| [asyncQuery](https://github.com/taosdata/taos-connector-node/blob/3.0/nodejs/examples/asyncQueryExample.js)                                         | 异步查询的使用示例。                   |
+| [REST](https://github.com/taosdata/taos-connector-node/blob/3.0/typescript-rest/example/example.ts)                                    | 使用 REST 连接的 TypeScript 使用示例。 |
 
 ## 使用限制
 
-Node.js 连接器 >= v2.0.6 目前支持 node 的版本为：支持 >=v12.8.0 <= v12.9.1 || >=v10.20.0 <= v10.9.0 ；2.0.5 及更早版本支持 v10.x 版本，其他版本可能存在包兼容性的问题。
+native 连接器（`@tdengine/client`） >= v3.0.0 目前支持 node 的版本为：支持 >=v12.8.0 <= v12.9.1 || >=v10.20.0 <= v10.9.0 ；2.0.5 及更早版本支持 v10.x 版本，其他版本可能存在包兼容性的问题。
 
 ## 其他说明
 
@@ -225,7 +222,7 @@ Node.js 连接器的使用参见[视频教程](https://www.taosdata.com/blog/202
 
 2. Node.js 版本
 
-  连接器 >v2.0.6 目前兼容的 Node.js 版本为：>=v10.20.0 <= v10.9.0 || >=v12.8.0 <= v12.9.1
+   原生连接器 `@tdengine/client` 目前兼容的 Node.js 版本为：>=v10.20.0 <= v10.9.0 || >=v12.8.0 <= v12.9.1
 
 3. "Unable to establish connection"，"Unable to resolve FQDN"
 
@@ -235,17 +232,21 @@ Node.js 连接器的使用参见[视频教程](https://www.taosdata.com/blog/202
 
 ### 原生连接器
 
-| td2.0-connector 版本 | 说明                                                             |
-| -------------------- | ---------------------------------------------------------------- |
-| 2.0.12               | 修复 cursor.close() 报错的 bug。                                 |
-| 2.0.11               | 支持绑定参数、json tag、schemaless 接口等功能。                  |
-| 2.0.10               | 支持连接管理，普通查询、连续查询、获取系统信息、订阅功能等功能。 |
-
+| package name     | version | TDengine version    | 说明                                                             |
+|------------------|---------|---------------------|------------------------------------------------------------------|
+| @tdengine/client | 3.0.0   | 3.0.0               | 支持TDengine 3.0 且不与2.x 兼容。                                                          |
+| td2.0-connector  | 2.0.12  | 2.4.x；2.5.x；2.6.x | 修复 cursor.close() 报错的 bug。                                 |
+| td2.0-connector  | 2.0.11  | 2.4.x；2.5.x；2.6.x | 支持绑定参数、json tag、schemaless 接口等功能。                  |
+| td2.0-connector  | 2.0.10  | 2.4.x；2.5.x；2.6.x | 支持连接管理，普通查询、连续查询、获取系统信息、订阅功能等功能。 |
 ### REST 连接器
 
-| td2.0-rest-connector 版本 | 说明                                                             |
-| ------------------------- | ---------------------------------------------------------------- |
-| 1.0.3                     | 支持连接管理、普通查询、获取系统信息、错误信息、连续查询等功能。 |
+| package name         | version | TDengine version    | 说明                                                                      |
+|----------------------|---------|---------------------|---------------------------------------------------------------------------|
+| @tdengine/rest       | 3.0.0   | 3.0.0               | 支持 TDegnine 3.0，且不与2.x 兼容。                                               |
+| td2.0-rest-connector | 1.0.7   | 2.4.x；2.5.x；2.6.x | 移除默认端口 6041。                                                       |
+| td2.0-rest-connector | 1.0.6   | 2.4.x；2.5.x；2.6.x | 修复create，insert，update，alter 等SQL 执行返回的 affectRows 错误的bug。 |
+| td2.0-rest-connector | 1.0.5   | 2.4.x；2.5.x；2.6.x | 支持云服务 cloud Token；                                                  |
+| td2.0-rest-connector | 1.0.3   | 2.4.x；2.5.x；2.6.x | 支持连接管理、普通查询、获取系统信息、错误信息、连续查询等功能。          |
 
 ## API 参考
 

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

@@ -233,11 +233,28 @@ taosBenchmark -A INT,DOUBLE,NCHAR,BINARY\(16\)
 
 - **drop** : 插入前是否删除数据库，默认为 true。
 
+#### 流式计算相关配置参数
+
+创建流式计算的相关参数在 json 配置文件中的 `stream` 中配置，具体参数如下。
+
+- **stream_name** : 流式计算的名称，必填项。
+
+- **stream_stb** : 流式计算对应的超级表名称，必填项。
+
+- **stream_sql** : 流式计算的sql语句，必填项。
+
+- **trigger_mode** : 流式计算的触发模式，可选项。
+
+- **watermark** : 流式计算的水印，可选项。
+
+- **drop** : 是否创建流式计算，可选项为 "yes" 或者 "no", 为 "no" 时不创建。
+
 #### 超级表相关配置参数
 
-创建超级表时的相关参数在 json 配置文件中的 `super_tables` 中配置，具体参数如下表。
+创建超级表时的相关参数在 json 配置文件中的 `super_tables` 中配置，具体参数如下。
 
 - **name**: 超级表名，必须配置，没有默认值。
+
 - **child_table_exists** : 子表是否已经存在，默认值为 "no"，可选值为 "yes" 或 "no"。
 
 - **child_table_count** : 子表的数量，默认值为 10。
@@ -288,6 +305,22 @@ taosBenchmark -A INT,DOUBLE,NCHAR,BINARY\(16\)
 
 - **tags_file** : 仅当 insert_mode 为 taosc, rest 的模式下生效。 最终的 tag 的数值与 childtable_count 有关，如果 csv 文件内的 tag 数据行小于给定的子表数量，那么会循环读取 csv 文件数据直到生成 childtable_count 指定的子表数量；否则则只会读取 childtable_count 行 tag 数据。也即最终生成的子表数量为二者取小。
 
+#### tsma配置参数
+
+指定tsma的配置参数在 `super_tables` 中的 `tsmas` 中，具体参数如下。
+
+- **name** : 指定 tsma 的名字，必选项。
+
+- **function** : 指定 tsma 的函数，必选项。
+
+- **interval** : 指定 tsma 的时间间隔，必选项。
+
+- **sliding** : 指定 tsma 的窗口时间位移，必选项。
+
+- **custom** : 指定 tsma 的创建语句结尾追加的自定义配置，可选项。
+
+- **start_when_inserted** : 指定当插入多少行时创建 tsma，可选项，默认为 0。
+
 #### 标签列与数据列配置参数
 
 指定超级表标签列与数据列的配置参数分别在 `super_tables` 中的 `columns` 和 `tag` 中。

docs/zh/14-reference/11-docker/index.md

@@ -147,7 +147,7 @@ import (
     "fmt"
     "time"
 
-    _ "github.com/taosdata/driver-go/v2/taosSql"
+    _ "github.com/taosdata/driver-go/v3/taosSql"
 )
 
 type config struct {

docs/zh/28-releases.md

+---
+sidebar_label: 发布历史
+title: 发布历史
+---
+
+import Release from "/components/Release";
+
+
+<Release versionPrefix="3.0" />

include/common/tcommon.h

@@ -107,6 +107,7 @@ typedef struct SDataBlockInfo {
   int32_t     childId;  // used for stream, do not serialize
   EStreamType type;     // used for stream, do not serialize
   STimeWindow calWin;   // used for stream, do not serialize
+  TSKEY       watermark;// used for stream
 } SDataBlockInfo;
 
 typedef struct SSDataBlock {

include/common/tmsg.h

@@ -1369,6 +1369,7 @@ typedef struct {
   int64_t skey;
   int64_t ekey;
   int64_t version;  // for stream
+  TSKEY   watermark;// for stream
   char    data[];
 } SRetrieveTableRsp;
 
@@ -2656,6 +2657,34 @@ typedef struct {
   SEpSet  epSet;
 } SVgEpSet;
 
+typedef struct {
+  int64_t refId;
+  int64_t suid;
+  int8_t  level;
+} SRSmaFetchMsg;
+
+static FORCE_INLINE int32_t tEncodeSRSmaFetchMsg(SEncoder* pCoder, const SRSmaFetchMsg* pReq) {
+  if (tStartEncode(pCoder) < 0) return -1;
+
+  if (tEncodeI64(pCoder, pReq->refId) < 0) return -1;
+  if (tEncodeI64(pCoder, pReq->suid) < 0) return -1;
+  if (tEncodeI8(pCoder, pReq->level) < 0) return -1;
+
+  tEndEncode(pCoder);
+  return 0;
+}
+
+static FORCE_INLINE int32_t tDecodeSRSmaFetchMsg(SDecoder* pCoder, SRSmaFetchMsg* pReq) {
+  if (tStartDecode(pCoder) < 0) return -1;
+
+  if (tDecodeI64(pCoder, &pReq->refId) < 0) return -1;
+  if (tDecodeI64(pCoder, &pReq->suid) < 0) return -1;
+  if (tDecodeI8(pCoder, &pReq->level) < 0) return -1;
+
+  tEndDecode(pCoder);
+  return 0;
+}
+
 typedef struct {
   int8_t         version;       // for compatibility(default 0)
   int8_t         intervalUnit;  // MACRO: TIME_UNIT_XXX
@@ -3035,6 +3064,7 @@ int32_t tEncodeDeleteRes(SEncoder* pCoder, const SDeleteRes* pRes);
 int32_t tDecodeDeleteRes(SDecoder* pCoder, SDeleteRes* pRes);
 
 typedef struct {
+  int32_t msgIdx;
   int32_t msgType;
   int32_t msgLen;
   void*   msg;
@@ -3048,6 +3078,7 @@ typedef struct {
 
 typedef struct {
   int32_t reqType;
+  int32_t msgIdx;
   int32_t msgLen;
   int32_t rspCode;
   void*   msg;

include/libs/catalog/catalog.h

@@ -58,12 +58,17 @@ typedef struct SDbInfo {
   int64_t dbId;
 } SDbInfo;
 
+typedef struct STablesReq {
+  char    dbFName[TSDB_DB_FNAME_LEN];
+  SArray* pTables;
+} STablesReq;
+
 typedef struct SCatalogReq {
   SArray* pDbVgroup;      // element is db full name
   SArray* pDbCfg;         // element is db full name
   SArray* pDbInfo;        // element is db full name
-  SArray* pTableMeta;     // element is SNAME
-  SArray* pTableHash;     // element is SNAME
+  SArray* pTableMeta;     // element is STablesReq
+  SArray* pTableHash;     // element is STablesReq
   SArray* pUdf;           // element is udf name
   SArray* pIndex;         // element is index name
   SArray* pUser;          // element is SUserAuthInfo

include/libs/executor/executor.h

@@ -107,7 +107,7 @@ int32_t qUpdateQualifiedTableId(qTaskInfo_t tinfo, const SArray* tableIdList, bo
  * @return
  */
 int32_t qCreateExecTask(SReadHandle* readHandle, int32_t vgId, uint64_t taskId, struct SSubplan* pPlan,
-                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, const char* sql, EOPTR_EXEC_MODEL model);
+                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, char* sql, EOPTR_EXEC_MODEL model);
 
 /**
  *

include/libs/function/functionMgt.h

@@ -202,6 +202,7 @@ bool fmIsForbidStreamFunc(int32_t funcId);
 bool fmIsIntervalInterpoFunc(int32_t funcId);
 bool fmIsInterpFunc(int32_t funcId);
 bool fmIsLastRowFunc(int32_t funcId);
+bool fmIsNotNullOutputFunc(int32_t funcId);
 bool fmIsSelectValueFunc(int32_t funcId);
 bool fmIsSystemInfoFunc(int32_t funcId);
 bool fmIsImplicitTsFunc(int32_t funcId);

include/libs/transport/trpc.h

@@ -43,11 +43,12 @@ typedef struct SRpcHandleInfo {
   // rpc info
   void   *handle;         // rpc handle returned to app
   int64_t refId;          // refid, used by server
-  int32_t  noResp;         // has response or not(default 0, 0: resp, 1: no resp);
-  int32_t  persistHandle;  // persist handle or not
-  STraceId traceId;
+  int8_t  noResp;         // has response or not(default 0, 0: resp, 1: no resp)
+  int8_t  persistHandle;  // persist handle or not
   int8_t  hasEpSet;
 
+  STraceId traceId;
+
   // app info
   void *ahandle;  // app handle set by client
   void *wrapper;  // wrapper handle
@@ -69,8 +70,9 @@ typedef struct SRpcMsg {
   SRpcHandleInfo info;
 } SRpcMsg;
 
-typedef void (*RpcCfp)(void *parent, SRpcMsg *, SEpSet *rf);
+typedef void (*RpcCfp)(void *parent, SRpcMsg *, SEpSet *epset);
 typedef bool (*RpcRfp)(int32_t code, tmsg_t msgType);
+typedef bool (*RpcTfp)(int32_t code, tmsg_t msgType);
 
 typedef struct SRpcInit {
   char     localFqdn[TSDB_FQDN_LEN];
@@ -84,12 +86,15 @@ typedef struct SRpcInit {
   // the following is for client app ecurity only
   char *user;  // user name
 
-  // call back to process incoming msg, code shall be ignored by server app
+  // call back to process incoming msg
   RpcCfp cfp;
 
-  // user defined retry func
+  // retry not not for particular msg
   RpcRfp rfp;
 
+  // set up timeout for particular msg
+  RpcTfp tfp;
+
   void *parent;
 } SRpcInit;
 

include/util/taoserror.h

@@ -46,6 +46,7 @@ int32_t* taosGetErrno();
 #define TSDB_CODE_RPC_FQDN_ERROR                TAOS_DEF_ERROR_CODE(0, 0x0015)
 #define TSDB_CODE_RPC_PORT_EADDRINUSE           TAOS_DEF_ERROR_CODE(0, 0x0017)
 #define TSDB_CODE_RPC_BROKEN_LINK               TAOS_DEF_ERROR_CODE(0, 0x0018)
+#define TSDB_CODE_RPC_TIMEOUT                   TAOS_DEF_ERROR_CODE(0, 0x0019)
 
 //common & util
 #define TSDB_CODE_TIME_UNSYNCED                 TAOS_DEF_ERROR_CODE(0, 0x0013) 

include/util/tlockfree.h

@@ -27,7 +27,7 @@ typedef void (*_ref_fn_t)(const void *pObj);
 
 #define T_REF_DECLARE()   \
   struct {                \
-    int32_t val;        \
+    volatile int32_t val; \
   } _ref;
 
 #define T_REF_REGISTER_FUNC(s, e) \

source/client/src/clientEnv.c

@@ -109,6 +109,14 @@ static bool clientRpcRfp(int32_t code, tmsg_t msgType) {
   }
 }
 
+// start timer for particular msgType
+static bool clientRpcTfp(int32_t code, tmsg_t msgType) {
+  if (msgType == TDMT_VND_SUBMIT || msgType == TDMT_VND_CREATE_TABLE) {
+    return true;
+  }
+  return false;
+}
+
 // TODO refactor
 void *openTransporter(const char *user, const char *auth, int32_t numOfThread) {
   SRpcInit rpcInit;
@@ -118,6 +126,7 @@ void *openTransporter(const char *user, const char *auth, int32_t numOfThread) {
   rpcInit.numOfThreads = numOfThread;
   rpcInit.cfp = processMsgFromServer;
   rpcInit.rfp = clientRpcRfp;
+  rpcInit.tfp = clientRpcTfp;
   rpcInit.sessions = 1024;
   rpcInit.connType = TAOS_CONN_CLIENT;
   rpcInit.user = (char *)user;

source/client/src/clientImpl.c

@@ -1308,8 +1308,8 @@ int32_t doProcessMsgFromServer(void* param) {
   char      tbuf[40] = {0};
   TRACE_TO_STR(trace, tbuf);
 
-  tscDebug("processMsgFromServer handle %p, message: %s, code: %s, gtid: %s", pMsg->info.handle,
-           TMSG_INFO(pMsg->msgType), tstrerror(pMsg->code), tbuf);
+  tscDebug("processMsgFromServer handle %p, message: %s, size:%d, code: %s, gtid: %s", pMsg->info.handle,
+           TMSG_INFO(pMsg->msgType), pMsg->contLen, tstrerror(pMsg->code), tbuf);
 
   if (pSendInfo->requestObjRefId != 0) {
     SRequestObj* pRequest = (SRequestObj*)taosAcquireRef(clientReqRefPool, pSendInfo->requestObjRefId);
@@ -1922,7 +1922,7 @@ _OVER:
   return code;
 }
 
-int32_t appendTbToReq(SArray* pList, int32_t pos1, int32_t len1, int32_t pos2, int32_t len2, const char* str,
+int32_t appendTbToReq(SHashObj* pHash, int32_t pos1, int32_t len1, int32_t pos2, int32_t len2, const char* str,
                       int32_t acctId, char* db) {
   SName name;
 
@@ -1957,20 +1957,33 @@ int32_t appendTbToReq(SArray* pList, int32_t pos1, int32_t len1, int32_t pos2, i
     return -1;
   }
 
-  taosArrayPush(pList, &name);
+  char dbFName[TSDB_DB_FNAME_LEN];
+  sprintf(dbFName, "%d.%.*s", acctId, dbLen, dbName);
+
+  STablesReq* pDb = taosHashGet(pHash, dbFName, strlen(dbFName));
+  if (pDb) {
+    taosArrayPush(pDb->pTables, &name);
+  } else {
+    STablesReq db;
+    db.pTables = taosArrayInit(20, sizeof(SName));
+    strcpy(db.dbFName, dbFName);
+    taosArrayPush(db.pTables, &name);
+    taosHashPut(pHash, dbFName, strlen(dbFName), &db, sizeof(db));
+  }
 
   return TSDB_CODE_SUCCESS;
 }
 
 int32_t transferTableNameList(const char* tbList, int32_t acctId, char* dbName, SArray** pReq) {
-  *pReq = taosArrayInit(10, sizeof(SName));
-  if (NULL == *pReq) {
+  SHashObj* pHash = taosHashInit(3, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY), false, HASH_NO_LOCK);
+  if (NULL == pHash) {
     terrno = TSDB_CODE_OUT_OF_MEMORY;
     return terrno;
   }
 
   bool    inEscape = false;
   int32_t code = 0;
+  void *pIter = NULL;
 
   int32_t vIdx = 0;
   int32_t vPos[2];
@@ -1985,7 +1998,7 @@ int32_t transferTableNameList(const char* tbList, int32_t acctId, char* dbName,
         vLen[vIdx] = i - vPos[vIdx];
       }
 
-      code = appendTbToReq(*pReq, vPos[0], vLen[0], vPos[1], vLen[1], tbList, acctId, dbName);
+      code = appendTbToReq(pHash, vPos[0], vLen[0], vPos[1], vLen[1], tbList, acctId, dbName);
       if (code) {
         goto _return;
       }
@@ -2035,7 +2048,7 @@ int32_t transferTableNameList(const char* tbList, int32_t acctId, char* dbName,
         vLen[vIdx] = i - vPos[vIdx];
       }
 
-      code = appendTbToReq(*pReq, vPos[0], vLen[0], vPos[1], vLen[1], tbList, acctId, dbName);
+      code = appendTbToReq(pHash, vPos[0], vLen[0], vPos[1], vLen[1], tbList, acctId, dbName);
       if (code) {
         goto _return;
       }
@@ -2067,14 +2080,31 @@ int32_t transferTableNameList(const char* tbList, int32_t acctId, char* dbName,
     goto _return;
   }
 
+  int32_t dbNum = taosHashGetSize(pHash);
+  *pReq = taosArrayInit(dbNum, sizeof(STablesReq));
+  pIter = taosHashIterate(pHash, NULL);
+  while (pIter) {
+    STablesReq* pDb = (STablesReq*)pIter;
+    taosArrayPush(*pReq, pDb);
+    pIter = taosHashIterate(pHash, pIter);
+  }
+
+  taosHashCleanup(pHash);
+
   return TSDB_CODE_SUCCESS;
 
 _return:
 
   terrno = TSDB_CODE_TSC_INVALID_OPERATION;
 
-  taosArrayDestroy(*pReq);
-  *pReq = NULL;
+  pIter = taosHashIterate(pHash, NULL);
+  while (pIter) {
+    STablesReq* pDb = (STablesReq*)pIter;
+    taosArrayDestroy(pDb->pTables);
+    pIter = taosHashIterate(pHash, pIter);
+  }
+
+  taosHashCleanup(pHash);
 
   return terrno;
 }

source/common/src/systable.c

@@ -308,9 +308,9 @@ static const SSysDbTableSchema offsetSchema[] = {
 };
 
 static const SSysDbTableSchema querySchema[] = {
-    {.name = "query_id", .bytes = TSDB_QUERY_ID_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
-    {.name = "req_id", .bytes = 8, .type = TSDB_DATA_TYPE_UBIGINT},
-    {.name = "connId", .bytes = 4, .type = TSDB_DATA_TYPE_UINT},
+    {.name = "kill_id", .bytes = TSDB_QUERY_ID_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
+    {.name = "query_id", .bytes = 8, .type = TSDB_DATA_TYPE_UBIGINT},
+    {.name = "conn_id", .bytes = 4, .type = TSDB_DATA_TYPE_UINT},
     {.name = "app", .bytes = TSDB_APP_NAME_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
     {.name = "pid", .bytes = 4, .type = TSDB_DATA_TYPE_INT},
     {.name = "user", .bytes = TSDB_USER_LEN + VARSTR_HEADER_SIZE, .type = TSDB_DATA_TYPE_VARCHAR},
@@ -354,11 +354,19 @@ static const SSysTableMeta perfsMeta[] = {
     {TSDB_PERFS_TABLE_APPS, appSchema, tListLen(appSchema)}};
 
 void getInfosDbMeta(const SSysTableMeta** pInfosTableMeta, size_t* size) {
+  if (pInfosTableMeta) {
     *pInfosTableMeta = infosMeta;
+  }
+  if (size) {
     *size = tListLen(infosMeta);
+  }
 }
 
 void getPerfDbMeta(const SSysTableMeta** pPerfsTableMeta, size_t* size) {
+  if (pPerfsTableMeta) {
     *pPerfsTableMeta = perfsMeta;
+  }
+  if (size) {
     *size = tListLen(perfsMeta);
+  }
 }

source/common/src/tdatablock.c

@@ -1273,8 +1273,7 @@ int32_t assignOneDataBlock(SSDataBlock* dst, const SSDataBlock* src) {
     colDataAssign(pDst, pSrc, src->info.rows, &src->info);
   }
 
-  dst->info.rows = src->info.rows;
-  dst->info.capacity = src->info.rows;
+  dst->info = src->info;
   return 0;
 }
 

source/common/src/tglobal.c

@@ -89,7 +89,7 @@ bool tsSmlDataFormat =
 
 // query
 int32_t tsQueryPolicy = 1;
-int32_t tsQuerySmaOptimize = 1;
+int32_t tsQuerySmaOptimize = 0;
 
 /*
  * denote if the server needs to compress response message at the application layer to client, including query rsp,

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

@@ -1691,13 +1691,17 @@ static int32_t mndRetrieveDbs(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBloc
   if (!pShow->sysDbRsp) {
     SDbObj infoschemaDb = {0};
     setInformationSchemaDbCfg(&infoschemaDb);
-    dumpDbInfoData(pBlock, &infoschemaDb, pShow, numOfRows, 14, true, 0, 1);
+    size_t numOfTables = 0;
+    getInfosDbMeta(NULL, &numOfTables);
+    dumpDbInfoData(pBlock, &infoschemaDb, pShow, numOfRows, numOfTables, true, 0, 1);
 
     numOfRows += 1;
 
     SDbObj perfschemaDb = {0};
     setPerfSchemaDbCfg(&perfschemaDb);
-    dumpDbInfoData(pBlock, &perfschemaDb, pShow, numOfRows, 3, true, 0, 1);
+    numOfTables = 0;
+    getPerfDbMeta(NULL, &numOfTables);
+    dumpDbInfoData(pBlock, &perfschemaDb, pShow, numOfRows, numOfTables, true, 0, 1);
 
     numOfRows += 1;
     pShow->sysDbRsp = true;

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

@@ -84,6 +84,9 @@ int32_t mndProcessBatchMetaMsg(SRpcMsg *pMsg) {
   }
   
   for (int32_t i = 0; i < msgNum; ++i) {
+    req.msgIdx = ntohl(*(int32_t*)((char*)pMsg->pCont + offset));
+    offset += sizeof(req.msgIdx);
+
     req.msgType = ntohl(*(int32_t*)((char*)pMsg->pCont + offset));
     offset += sizeof(req.msgType);
 
@@ -111,6 +114,7 @@ int32_t mndProcessBatchMetaMsg(SRpcMsg *pMsg) {
     } else {
       rsp.rspCode = 0;
     }
+    rsp.msgIdx = req.msgIdx;
     rsp.reqType = reqMsg.msgType;
     rsp.msgLen = reqMsg.info.rspLen;
     rsp.msg = reqMsg.info.rsp;
@@ -136,6 +140,8 @@ int32_t mndProcessBatchMetaMsg(SRpcMsg *pMsg) {
     
     *(int32_t*)((char*)pRsp + offset) = htonl(p->reqType);
     offset += sizeof(p->reqType);
+    *(int32_t*)((char*)pRsp + offset) = htonl(p->msgIdx);
+    offset += sizeof(p->msgIdx);
     *(int32_t*)((char*)pRsp + offset) = htonl(p->msgLen);
     offset += sizeof(p->msgLen);
     *(int32_t*)((char*)pRsp + offset) = htonl(p->rspCode);

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

@@ -489,7 +489,7 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   smaObj.uid = mndGenerateUid(pCreate->name, TSDB_TABLE_FNAME_LEN);
   ASSERT(smaObj.uid != 0);
   char resultTbName[TSDB_TABLE_FNAME_LEN + 16] = {0};
-  snprintf(resultTbName, TSDB_TABLE_FNAME_LEN + 16, "td.tsma.rst.tb.%s", pCreate->name);
+  snprintf(resultTbName, TSDB_TABLE_FNAME_LEN + 16, "%s_td_tsma_rst_tb",pCreate->name);
   memcpy(smaObj.dstTbName, resultTbName, TSDB_TABLE_FNAME_LEN);
   smaObj.dstTbUid = mndGenerateUid(smaObj.dstTbName, TSDB_TABLE_FNAME_LEN);
   smaObj.stbUid = pStb->uid;
@@ -603,6 +603,9 @@ static int32_t mndCreateSma(SMnode *pMnode, SRpcMsg *pReq, SMCreateSmaReq *pCrea
   if (mndPersistStream(pMnode, pTrans, &streamObj) != 0) goto _OVER;
   if (mndTransPrepare(pMnode, pTrans) != 0) goto _OVER;
 
+  mDebug("mndSma: create sma index %s %" PRIi64 " on stb:%" PRIi64 ", dstSuid:%" PRIi64 " dstTb:%s dstVg:%d",
+         pCreate->name, smaObj.uid, smaObj.stbUid, smaObj.dstTbUid, smaObj.dstTbName, smaObj.dstVgId);
+
   code = 0;
 
 _OVER:
@@ -795,11 +798,12 @@ static int32_t mndDropSma(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SSmaObj *p
   pStb = mndAcquireStb(pMnode, pSma->stb);
   if (pStb == NULL) goto _OVER;
 
-  pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_DB, pReq);
+  pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_DB, pReq);
   if (pTrans == NULL) goto _OVER;
 
   mDebug("trans:%d, used to drop sma:%s", pTrans->id, pSma->name);
   mndTransSetDbName(pTrans, pDb->name, NULL);
+  mndTransSetSerial(pTrans);
 
   char streamName[TSDB_TABLE_FNAME_LEN] = {0};
   mndGetStreamNameFromSmaName(streamName, pSma->name);
@@ -834,9 +838,6 @@ static int32_t mndDropSma(SMnode *pMnode, SRpcMsg *pReq, SDbObj *pDb, SSmaObj *p
   code = 0;
 
 _OVER:
-  if(code != 0) {
-    ASSERT(0);
-  }
   mndTransDrop(pTrans);
   mndReleaseVgroup(pMnode, pVgroup);
   mndReleaseStb(pMnode, pStb);
@@ -855,6 +856,7 @@ int32_t mndDropSmasByStb(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SStbObj *p
     if (pIter == NULL) break;
 
     if (pSma->stbUid == pStb->uid) {
+      mndTransSetSerial(pTrans);
       pVgroup = mndAcquireVgroup(pMnode, pSma->dstVgId);
       if (pVgroup == NULL) goto _OVER;
 
@@ -935,7 +937,6 @@ static int32_t mndProcessDropSmaReq(SRpcMsg *pReq) {
       goto _OVER;
     } else {
       terrno = TSDB_CODE_MND_SMA_NOT_EXIST;
-      ASSERT(0);
       goto _OVER;
     }
   }

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

@@ -115,24 +115,29 @@ struct SSmaStat {
 #define RSMA_FS_LOCK(r)       (&(r)->lock)
 
 struct SRSmaInfoItem {
-  void   *taskInfo;  // qTaskInfo_t
-  int64_t refId;
-  tmr_h   tmrId;
-  int32_t maxDelay;
   int8_t  level;
   int8_t  triggerStat;
+  int32_t maxDelay;
+  tmr_h   tmrId;
 };
 
 struct SRSmaInfo {
   STSchema *pTSchema;
   int64_t   suid;
+  int64_t   refId;  // refId of SRSmaStat
   int8_t    delFlag;
   T_REF_DECLARE()
   SRSmaInfoItem items[TSDB_RETENTION_L2];
+  void         *taskInfo[TSDB_RETENTION_L2];   // qTaskInfo_t
+  void         *iTaskInfo[TSDB_RETENTION_L2];  // immutable
 };
-#define RSMA_INFO_HEAD_LEN   24
+
+#define RSMA_INFO_HEAD_LEN     32
 #define RSMA_INFO_IS_DEL(r)    ((r)->delFlag == 1)
 #define RSMA_INFO_SET_DEL(r)   ((r)->delFlag = 1)
+#define RSMA_INFO_QTASK(r, i)  ((r)->taskInfo[i])
+#define RSMA_INFO_IQTASK(r, i) ((r)->iTaskInfo[i])
+#define RSMA_INFO_ITEM(r, i)   (&(r)->items[i])
 
 enum {
   TASK_TRIGGER_STAT_INIT = 0,
@@ -168,8 +173,8 @@ int32_t tdUnRefSmaStat(SSma *pSma, SSmaStat *pStat);
 int32_t tdRefRSmaInfo(SSma *pSma, SRSmaInfo *pRSmaInfo);
 int32_t tdUnRefRSmaInfo(SSma *pSma, SRSmaInfo *pRSmaInfo);
 
-void   *tdAcquireSmaRef(int32_t rsetId, int64_t refId, const char *tags, int32_t ln);
-int32_t tdReleaseSmaRef(int32_t rsetId, int64_t refId, const char *tags, int32_t ln);
+void   *tdAcquireSmaRef(int32_t rsetId, int64_t refId);
+int32_t tdReleaseSmaRef(int32_t rsetId, int64_t refId);
 
 int32_t tdCheckAndInitSmaEnv(SSma *pSma, int8_t smaType);
 
@@ -223,12 +228,11 @@ static FORCE_INLINE void tdSmaStatSetDropped(STSmaStat *pTStat) {
 
 void           tdRSmaQTaskInfoGetFileName(int32_t vid, int64_t version, char *outputName);
 void           tdRSmaQTaskInfoGetFullName(int32_t vid, int64_t version, const char *path, char *outputName);
-int32_t        tdCloneRSmaInfo(SSma *pSma, SRSmaInfo *pDest, SRSmaInfo *pSrc);
+int32_t        tdCloneRSmaInfo(SSma *pSma, SRSmaInfo **pDest, SRSmaInfo *pSrc);
 void           tdFreeQTaskInfo(qTaskInfo_t *taskHandle, int32_t vgId, int32_t level);
 static int32_t tdDestroySmaState(SSmaStat *pSmaStat, int8_t smaType);
 void          *tdFreeSmaState(SSmaStat *pSmaStat, int8_t smaType);
 void          *tdFreeRSmaInfo(SSma *pSma, SRSmaInfo *pInfo, bool isDeepFree);
-void           tdRemoveRSmaInfoBySuid(SSma *pSma, int64_t suid);
 int32_t        tdRSmaPersistExecImpl(SRSmaStat *pRSmaStat, SHashObj *pInfoHash);
 
 int32_t tdProcessRSmaCreateImpl(SSma *pSma, SRSmaParam *param, int64_t suid, const char *tbName);

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

@@ -381,6 +381,8 @@ int metaCreateTable(SMeta *pMeta, int64_t version, SVCreateTbReq *pReq) {
     terrno = TSDB_CODE_TDB_TABLE_ALREADY_EXIST;
     metaReaderClear(&mr);
     return -1;
+  } else if (terrno == TSDB_CODE_PAR_TABLE_NOT_EXIST) {
+    terrno = TSDB_CODE_SUCCESS;
   }
   metaReaderClear(&mr);
 

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

@@ -308,12 +308,12 @@ static int32_t tdProcessRSmaSyncPostCommitImpl(SSma *pSma) {
  * @return int32_t
  */
 static int32_t tdProcessRSmaAsyncPreCommitImpl(SSma *pSma) {
-  SSmaEnv *pSmaEnv = SMA_RSMA_ENV(pSma);
-  if (!pSmaEnv) {
+  SSmaEnv *pEnv = SMA_RSMA_ENV(pSma);
+  if (!pEnv) {
     return TSDB_CODE_SUCCESS;
   }
 
-  SSmaStat  *pStat = SMA_ENV_STAT(pSmaEnv);
+  SSmaStat  *pStat = SMA_ENV_STAT(pEnv);
   SRSmaStat *pRSmaStat = SMA_RSMA_STAT(pStat);
 
   // step 1: set rsma stat
@@ -337,18 +337,26 @@ static int32_t tdProcessRSmaAsyncPreCommitImpl(SSma *pSma) {
   }
 
   // step 3:  swap rsmaInfoHash and iRsmaInfoHash
-  ASSERT(!RSMA_IMU_INFO_HASH(pRSmaStat));
+  // lock
+  taosWLockLatch(SMA_ENV_LOCK(pEnv));
+
   ASSERT(RSMA_INFO_HASH(pRSmaStat));
+  ASSERT(!RSMA_IMU_INFO_HASH(pRSmaStat));
 
   RSMA_IMU_INFO_HASH(pRSmaStat) = RSMA_INFO_HASH(pRSmaStat);
   RSMA_INFO_HASH(pRSmaStat) =
       taosHashInit(RSMA_TASK_INFO_HASH_SLOT, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BIGINT), true, HASH_ENTRY_LOCK);
 
   if (!RSMA_INFO_HASH(pRSmaStat)) {
+    // unlock
+    taosWUnLockLatch(SMA_ENV_LOCK(pEnv));
     smaError("vgId:%d, rsma async commit failed since %s", SMA_VID(pSma), terrstr());
     return TSDB_CODE_FAILED;
   }
 
+  // unlock
+  taosWUnLockLatch(SMA_ENV_LOCK(pEnv));
+
   // step 4: others
   pRSmaStat->commitAppliedVer = pSma->pVnode->state.applied;
 
@@ -383,26 +391,52 @@ static int32_t tdProcessRSmaAsyncCommitImpl(SSma *pSma) {
  * @return int32_t
  */
 static int32_t tdProcessRSmaAsyncPostCommitImpl(SSma *pSma) {
-  SSmaEnv *pSmaEnv = SMA_RSMA_ENV(pSma);
-  if (!pSmaEnv) {
+  SSmaEnv *pEnv = SMA_RSMA_ENV(pSma);
+  if (!pEnv) {
     return TSDB_CODE_SUCCESS;
   }
 
-  SSmaStat  *pStat = SMA_ENV_STAT(pSmaEnv);
+  SSmaStat  *pStat = SMA_ENV_STAT(pEnv);
   SRSmaStat *pRSmaStat = SMA_RSMA_STAT(pStat);
 
   // step 1: merge rsmaInfoHash and iRsmaInfoHash
-  taosWLockLatch(SMA_ENV_LOCK(pSmaEnv));
-
+  // lock
+  taosWLockLatch(SMA_ENV_LOCK(pEnv));
+#if 0
   if (taosHashGetSize(RSMA_INFO_HASH(pRSmaStat)) <= 0) {
-    // TODO: optimization - just switch the hash pointer if rsmaInfoHash is empty
+    // just switch the hash pointer if rsmaInfoHash is empty
+    if (taosHashGetSize(RSMA_IMU_INFO_HASH(pRSmaStat)) > 0) {
+      SHashObj *infoHash = RSMA_INFO_HASH(pRSmaStat);
+      RSMA_INFO_HASH(pRSmaStat) = RSMA_IMU_INFO_HASH(pRSmaStat);
+      RSMA_IMU_INFO_HASH(pRSmaStat) = infoHash;
     }
-
+  } else {
+#endif
+#if 1
   void *pIter = taosHashIterate(RSMA_IMU_INFO_HASH(pRSmaStat), NULL);
   while (pIter) {
     tb_uid_t *pSuid = (tb_uid_t *)taosHashGetKey(pIter, NULL);
 
     if (!taosHashGet(RSMA_INFO_HASH(pRSmaStat), pSuid, sizeof(tb_uid_t))) {
+      SRSmaInfo *pRSmaInfo = *(SRSmaInfo **)pIter;
+      if (RSMA_INFO_IS_DEL(pRSmaInfo)) {
+        int32_t refVal = T_REF_VAL_GET(pRSmaInfo);
+        if (refVal == 0) {
+          tdFreeRSmaInfo(pSma, pRSmaInfo, true);
+          smaDebug(
+              "vgId:%d, rsma async post commit, free rsma info since already deleted and ref is 0 for "
+              "table:%" PRIi64,
+              SMA_VID(pSma), *pSuid);
+        } else {
+          smaDebug(
+              "vgId:%d, rsma async post commit, not free rsma info since ref is %d although already deleted for "
+              "table:%" PRIi64,
+              SMA_VID(pSma), refVal, *pSuid);
+        }
+
+        pIter = taosHashIterate(RSMA_IMU_INFO_HASH(pRSmaStat), pIter);
+        continue;
+      }
       taosHashPut(RSMA_INFO_HASH(pRSmaStat), pSuid, sizeof(tb_uid_t), pIter, sizeof(pIter));
       smaDebug("vgId:%d, rsma async post commit, migrated from iRsmaInfoHash for table:%" PRIi64, SMA_VID(pSma),
                *pSuid);
@@ -416,11 +450,14 @@ static int32_t tdProcessRSmaAsyncPostCommitImpl(SSma *pSma) {
 
     pIter = taosHashIterate(RSMA_IMU_INFO_HASH(pRSmaStat), pIter);
   }
+#endif
+  // }
 
   taosHashCleanup(RSMA_IMU_INFO_HASH(pRSmaStat));
   RSMA_IMU_INFO_HASH(pRSmaStat) = NULL;
 
-  taosWUnLockLatch(SMA_ENV_LOCK(pSmaEnv));
+  // unlock
+  taosWUnLockLatch(SMA_ENV_LOCK(pEnv));
 
   // step 2: cleanup outdated qtaskinfo files
   tdCleanupQTaskInfoFiles(pSma, pRSmaStat);

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

@@ -183,9 +183,6 @@ int32_t tdUnRefRSmaInfo(SSma *pSma, SRSmaInfo *pRSmaInfo) {
   int ref = T_REF_DEC(pRSmaInfo);
   smaDebug("vgId:%d, unref rsma info:%p, val:%d", SMA_VID(pSma), pRSmaInfo, ref);
 
-  if (ref == 0) {
-    tdRemoveRSmaInfoBySuid(pSma, pRSmaInfo->suid);
-  }
   return 0;
 }
 

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

@@ -116,8 +116,10 @@ int32_t tdProcessTSmaCreateImpl(SSma *pSma, int64_t version, const char *pMsg) {
     }
 
     // create stable to save tsma result in dstVgId
+    SName stbFullName = {0};
+    tNameFromString(&stbFullName, pCfg->dstTbName, T_NAME_ACCT | T_NAME_DB | T_NAME_TABLE);
     SVCreateStbReq pReq = {0};
-    pReq.name = pCfg->dstTbName;
+    pReq.name = (char*)tNameGetTableName(&stbFullName);
     pReq.suid = pCfg->dstTbUid;
     pReq.schemaRow = pCfg->schemaRow;
     pReq.schemaTag = pCfg->schemaTag;
@@ -125,6 +127,12 @@ int32_t tdProcessTSmaCreateImpl(SSma *pSma, int64_t version, const char *pMsg) {
     if (metaCreateSTable(SMA_META(pSma), version, &pReq) < 0) {
       return -1;
     }
+
+    smaDebug("vgId:%d, success to create sma index %s %" PRIi64 " on stb:%" PRIi64 ", dstSuid:%" PRIi64
+             " dstTb:%s dstVg:%d",
+             SMA_VID(pSma), pCfg->indexName, pCfg->indexUid, pCfg->tableUid, pCfg->dstTbUid, pReq.name, pCfg->dstVgId);
+  } else {
+    ASSERT(0);
   }
 
   return 0;

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

@@ -287,22 +287,22 @@ int32_t tdRemoveTFile(STFile *pTFile) {
 }
 
 // smaXXXUtil ================
-void *tdAcquireSmaRef(int32_t rsetId, int64_t refId, const char *tags, int32_t ln) {
+void *tdAcquireSmaRef(int32_t rsetId, int64_t refId) {
   void *pResult = taosAcquireRef(rsetId, refId);
   if (!pResult) {
-    smaWarn("%s:%d taosAcquireRef for rsetId:%" PRIi64 " refId:%d failed since %s", tags, ln, rsetId, refId, terrstr());
+    smaWarn("rsma acquire ref for rsetId:%" PRIi64 " refId:%d failed since %s", rsetId, refId, terrstr());
   } else {
-    smaDebug("%s:%d taosAcquireRef for rsetId:%" PRIi64 " refId:%d success", tags, ln, rsetId, refId);
+    smaDebug("rsma acquire ref for rsetId:%" PRIi64 " refId:%d success", rsetId, refId);
   }
   return pResult;
 }
 
-int32_t tdReleaseSmaRef(int32_t rsetId, int64_t refId, const char *tags, int32_t ln) {
+int32_t tdReleaseSmaRef(int32_t rsetId, int64_t refId) {
   if (taosReleaseRef(rsetId, refId) < 0) {
-    smaWarn("%s:%d taosReleaseRef for rsetId:%" PRIi64 " refId:%d failed since %s", tags, ln, rsetId, refId, terrstr());
+    smaWarn("rsma release ref for rsetId:%" PRIi64 " refId:%d failed since %s", rsetId, refId, terrstr());
     return TSDB_CODE_FAILED;
   }
-  smaDebug("%s:%d taosReleaseRef for rsetId:%" PRIi64 " refId:%d success", tags, ln, rsetId, refId);
+  smaDebug("rsma release ref for rsetId:%" PRIi64 " refId:%d success", rsetId, refId);
 
   return TSDB_CODE_SUCCESS;
 }
@@ -313,7 +313,7 @@ static int32_t tdCloneQTaskInfo(SSma *pSma, qTaskInfo_t dstTaskInfo, qTaskInfo_t
   char   *pOutput = NULL;
   int32_t len = 0;
 
-  if (qSerializeTaskStatus(srcTaskInfo, &pOutput, &len) < 0) {
+  if ((terrno = qSerializeTaskStatus(srcTaskInfo, &pOutput, &len)) < 0) {
     smaError("vgId:%d, rsma clone, table %" PRIi64 " serialize qTaskInfo failed since %s", TD_VID(pVnode), suid,
              terrstr());
     goto _err;
@@ -337,13 +337,15 @@ static int32_t tdCloneQTaskInfo(SSma *pSma, qTaskInfo_t dstTaskInfo, qTaskInfo_t
     goto _err;
   }
 
-  smaError("vgId:%d, rsma clone, restore rsma task for table:%" PRIi64 " succeed", TD_VID(pVnode), suid);
+  smaDebug("vgId:%d, rsma clone, restore rsma task for table:%" PRIi64 " succeed", TD_VID(pVnode), suid);
 
   taosMemoryFreeClear(pOutput);
   return TSDB_CODE_SUCCESS;
 _err:
   taosMemoryFreeClear(pOutput);
   tdFreeQTaskInfo(dstTaskInfo, TD_VID(pVnode), idx + 1);
+  smaError("vgId:%d, rsma clone, restore rsma task for table:%" PRIi64 " failed since %s", TD_VID(pVnode), suid,
+           terrstr());
   return TSDB_CODE_FAILED;
 }
 
@@ -355,23 +357,14 @@ _err:
  * @param pSrc
  * @return int32_t
  */
-int32_t tdCloneRSmaInfo(SSma *pSma, SRSmaInfo *pDest, SRSmaInfo *pSrc) {
+int32_t tdCloneRSmaInfo(SSma *pSma, SRSmaInfo **pDest, SRSmaInfo *pSrc) {
   SVnode     *pVnode = pSma->pVnode;
   SRSmaParam *param = NULL;
   if (!pSrc) {
+    *pDest = NULL;
     return TSDB_CODE_SUCCESS;
   }
 
-  if (!pDest) {
-    pDest = taosMemoryCalloc(1, sizeof(SRSmaInfo));
-    if (!pDest) {
-      terrno = TSDB_CODE_OUT_OF_MEMORY;
-      return TSDB_CODE_FAILED;
-    }
-  }
-
-  memcpy(pDest, pSrc, sizeof(SRSmaInfo));
-
   SMetaReader mr = {0};
   metaReaderInit(&mr, SMA_META(pSma), 0);
   smaDebug("vgId:%d, rsma clone, suid is %" PRIi64, TD_VID(pVnode), pSrc->suid);
@@ -384,21 +377,22 @@ int32_t tdCloneRSmaInfo(SSma *pSma, SRSmaInfo *pDest, SRSmaInfo *pSrc) {
   ASSERT(mr.me.uid == pSrc->suid);
   if (TABLE_IS_ROLLUP(mr.me.flags)) {
     param = &mr.me.stbEntry.rsmaParam;
-    for (int i = 0; i < TSDB_RETENTION_L2; ++i) {
-      SRSmaInfoItem *pItem = &pSrc->items[i];
-      if (pItem->taskInfo) {
-        tdCloneQTaskInfo(pSma, pDest->items[i].taskInfo, pItem->taskInfo, param, pSrc->suid, i);
+    for (int32_t i = 0; i < TSDB_RETENTION_L2; ++i) {
+      if (tdCloneQTaskInfo(pSma, pSrc->iTaskInfo[i], pSrc->taskInfo[i], param, pSrc->suid, i) < 0) {
+        goto _err;
       }
     }
     smaDebug("vgId:%d, rsma clone env success for %" PRIi64, TD_VID(pVnode), pSrc->suid);
   }
 
   metaReaderClear(&mr);
+
+  *pDest = pSrc;  // pointer copy
+
   return TSDB_CODE_SUCCESS;
 _err:
+  *pDest = NULL;
   metaReaderClear(&mr);
-  tdFreeRSmaInfo(pSma, pDest, false);
+  smaError("vgId:%d, rsma clone env failed for %" PRIi64 " since %s", TD_VID(pVnode), pSrc->suid, terrstr());
   return TSDB_CODE_FAILED;
 }
\ No newline at end of file
-
-// ...
\ No newline at end of file

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

@@ -223,6 +223,8 @@ int vnodeCommit(SVnode *pVnode) {
   vnodeBufPoolUnRef(pVnode->inUse);
   pVnode->inUse = NULL;
 
+  pVnode->state.commitTerm = pVnode->state.applyTerm;
+
   // save info
   info.config = pVnode->config;
   info.state.committed = pVnode->state.applied;

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

@@ -273,6 +273,9 @@ int32_t vnodeGetBatchMeta(SVnode *pVnode, SRpcMsg *pMsg) {
   }
 
   for (int32_t i = 0; i < msgNum; ++i) {
+    req.msgIdx = ntohl(*(int32_t *)((char *)pMsg->pCont + offset));
+    offset += sizeof(req.msgIdx);
+
     req.msgType = ntohl(*(int32_t *)((char *)pMsg->pCont + offset));
     offset += sizeof(req.msgType);
 
@@ -301,6 +304,7 @@ int32_t vnodeGetBatchMeta(SVnode *pVnode, SRpcMsg *pMsg) {
         break;
     }
 
+    rsp.msgIdx = req.msgIdx;
     rsp.reqType = reqMsg.msgType;
     rsp.msgLen = reqMsg.contLen;
     rsp.rspCode = reqMsg.code;
@@ -327,6 +331,8 @@ int32_t vnodeGetBatchMeta(SVnode *pVnode, SRpcMsg *pMsg) {
 
     *(int32_t *)((char *)pRsp + offset) = htonl(p->reqType);
     offset += sizeof(p->reqType);
+    *(int32_t *)((char *)pRsp + offset) = htonl(p->msgIdx);
+    offset += sizeof(p->msgIdx);
     *(int32_t *)((char *)pRsp + offset) = htonl(p->msgLen);
     offset += sizeof(p->msgLen);
     *(int32_t *)((char *)pRsp + offset) = htonl(p->rspCode);

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

@@ -351,7 +351,6 @@ int32_t vnodeProcessFetchMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo) {
 // TODO: remove the function
 void smaHandleRes(void *pVnode, int64_t smaId, const SArray *data) {
   // TODO
-
   blockDebugShowDataBlocks(data, __func__);
   tdProcessTSmaInsert(((SVnode *)pVnode)->pSma, smaId, (const char *)data);
 }
@@ -852,6 +851,7 @@ static int32_t vnodeProcessSubmitReq(SVnode *pVnode, int64_t version, void *pReq
       if (metaCreateTable(pVnode->pMeta, version, &createTbReq) < 0) {
         if (terrno != TSDB_CODE_TDB_TABLE_ALREADY_EXIST) {
           submitBlkRsp.code = terrno;
+          pRsp->code = terrno;
           tDecoderClear(&decoder);
           taosArrayDestroy(createTbReq.ctb.tagName);
           goto _exit;

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

@@ -708,8 +708,8 @@ int32_t vnodeSyncOpen(SVnode *pVnode, char *path) {
   }
 
   setPingTimerMS(pVnode->sync, 5000);
-  setElectTimerMS(pVnode->sync, 2800);
-  setHeartbeatTimerMS(pVnode->sync, 900);
+  setElectTimerMS(pVnode->sync, 4000);
+  setHeartbeatTimerMS(pVnode->sync, 700);
   return 0;
 }
 

source/libs/catalog/inc/catalogInt.h

@@ -32,6 +32,7 @@ extern "C" {
 #define CTG_DEFAULT_RENT_SLOT_SIZE 10
 #define CTG_DEFAULT_MAX_RETRY_TIMES 3
 #define CTG_DEFAULT_BATCH_NUM 64
+#define CTG_DEFAULT_FETCH_NUM 8
 
 #define CTG_RENT_SLOT_SECOND 1.5
 
@@ -80,6 +81,8 @@ typedef enum {
   CTG_TASK_GET_UDF,
   CTG_TASK_GET_USER,
   CTG_TASK_GET_SVR_VER,
+  CTG_TASK_GET_TB_META_BATCH,
+  CTG_TASK_GET_TB_HASH_BATCH,
 } CTG_TASK_TYPE;
 
 typedef enum {
@@ -110,6 +113,23 @@ typedef struct SCtgTbMetaCtx {
   int32_t flag;
 } SCtgTbMetaCtx;
 
+typedef struct SCtgFetch {
+  int32_t dbIdx;
+  int32_t tbIdx;
+  int32_t fetchIdx;
+  int32_t resIdx;
+  int32_t flag;
+  SCtgTbCacheInfo tbInfo;
+  int32_t vgId;
+} SCtgFetch;
+
+typedef struct SCtgTbMetasCtx {
+  int32_t fetchNum;
+  SArray* pNames;
+  SArray* pResList;
+  SArray* pFetchs;
+} SCtgTbMetasCtx;
+
 typedef struct SCtgTbIndexCtx {
   SName* pName;
 } SCtgTbIndexCtx;
@@ -137,6 +157,14 @@ typedef struct SCtgTbHashCtx {
   SName* pName;
 } SCtgTbHashCtx;
 
+typedef struct SCtgTbHashsCtx {
+  int32_t fetchNum;
+  SArray* pNames;
+  SArray* pResList;
+  SArray* pFetchs;
+} SCtgTbHashsCtx;
+
+
 typedef struct SCtgIndexCtx {
   char indexFName[TSDB_INDEX_FNAME_LEN];
 } SCtgIndexCtx;
@@ -211,6 +239,7 @@ typedef struct SCtgBatch {
   SRequestConnInfo conn;
   char    dbFName[TSDB_DB_FNAME_LEN];
   SArray* pTaskIds;
+  SArray* pMsgIdxs;
 } SCtgBatch;
 
 typedef struct SCtgJob {
@@ -218,6 +247,7 @@ typedef struct SCtgJob {
   int32_t           batchId;
   SHashObj*         pBatchs;
   SArray*           pTasks;
+  int32_t           subTaskNum;
   int32_t           taskDone;
   SMetaData         jobRes;
   int32_t           jobResCode;
@@ -258,6 +288,7 @@ typedef struct SCtgTaskCallbackParam {
   SArray*                 taskId;
   int32_t                 reqType;
   int32_t                 batchId;
+  SArray*                 msgIdx;
 } SCtgTaskCallbackParam;
 
 
@@ -276,6 +307,7 @@ typedef struct SCtgTask {
   int32_t         taskId;
   SCtgJob*        pJob;
   void*           taskCtx;
+  SArray*         msgCtxs;
   SCtgMsgCtx      msgCtx;
   int32_t         code;
   void*           res;
@@ -286,9 +318,14 @@ typedef struct SCtgTask {
   SHashObj*       pBatchs;
 } SCtgTask;
 
+typedef struct SCtgTaskReq {
+  SCtgTask* pTask;
+  int32_t   msgIdx;
+} SCtgTaskReq;
+
 typedef int32_t (*ctgInitTaskFp)(SCtgJob*, int32_t, void*);
 typedef int32_t (*ctgLanchTaskFp)(SCtgTask*);
-typedef int32_t (*ctgHandleTaskMsgRspFp)(SCtgTask*, int32_t, const SDataBuf *, int32_t);
+typedef int32_t (*ctgHandleTaskMsgRspFp)(SCtgTaskReq*, int32_t, const SDataBuf *, int32_t);
 typedef int32_t (*ctgDumpTaskResFp)(SCtgTask*);
 typedef int32_t (*ctgCloneTaskResFp)(SCtgTask*, void**);
 typedef int32_t (*ctgCompTaskFp)(SCtgTask*, void*, bool*);
@@ -487,6 +524,8 @@ typedef struct SCtgOperation {
 #define CTG_FLAG_MAKE_STB(_isStb) (((_isStb) == 1) ? CTG_FLAG_STB : ((_isStb) == 0 ? CTG_FLAG_NOT_STB : CTG_FLAG_UNKNOWN_STB))
 #define CTG_FLAG_MATCH_STB(_flag, tbType) (CTG_FLAG_IS_UNKNOWN_STB(_flag) || (CTG_FLAG_IS_STB(_flag) && (tbType) == TSDB_SUPER_TABLE) || (CTG_FLAG_IS_NOT_STB(_flag) && (tbType) != TSDB_SUPER_TABLE))
 
+#define CTG_GET_TASK_MSGCTX(_task, _id) (((CTG_TASK_GET_TB_META_BATCH == (_task)->type) || (CTG_TASK_GET_TB_HASH_BATCH == (_task)->type)) ? taosArrayGet((_task)->msgCtxs, (_id)) : &(_task)->msgCtx)
+
 #define CTG_META_SIZE(pMeta) (sizeof(STableMeta) + ((pMeta)->tableInfo.numOfTags + (pMeta)->tableInfo.numOfColumns) * sizeof(SSchema))
 
 #define CTG_TABLE_NOT_EXIST(code) (code == CTG_ERR_CODE_TABLE_NOT_EXIST) 
@@ -586,6 +625,7 @@ int32_t ctgdShowCacheInfo(void);
 
 int32_t ctgRemoveTbMetaFromCache(SCatalog* pCtg, SName* pTableName, bool syncReq);
 int32_t ctgGetTbMetaFromCache(SCatalog* pCtg, SRequestConnInfo *pConn, SCtgTbMetaCtx* ctx, STableMeta** pTableMeta);
+int32_t ctgGetTbMetasFromCache(SCatalog* pCtg, SRequestConnInfo *pConn, SCtgTbMetasCtx* ctx, int32_t dbIdx, int32_t *fetchIdx, int32_t baseResIdx, SArray* pList);
 
 int32_t ctgOpUpdateVgroup(SCtgCacheOperation *action);
 int32_t ctgOpUpdateTbMeta(SCtgCacheOperation *action);
@@ -631,7 +671,7 @@ int32_t ctgGetTbHashVgroupFromCache(SCatalog *pCtg, const SName *pTableName, SVg
 
 
 int32_t ctgProcessRspMsg(void* out, int32_t reqType, char* msg, int32_t msgSize, int32_t rspCode, char* target);
-int32_t ctgGetDBVgInfoFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, SBuildUseDBInput *input, SUseDbOutput *out, SCtgTask* pTask);
+int32_t ctgGetDBVgInfoFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, SBuildUseDBInput *input, SUseDbOutput *out, SCtgTaskReq* tReq);
 int32_t ctgGetQnodeListFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, SArray *out, SCtgTask* pTask);
 int32_t ctgGetDnodeListFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, SArray **out, SCtgTask* pTask);
 int32_t ctgGetDBCfgFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const char *dbFName, SDbCfgInfo *out, SCtgTask* pTask);
@@ -639,9 +679,9 @@ int32_t ctgGetIndexInfoFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const
 int32_t ctgGetTbIndexFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, SName *name, STableIndex* out, SCtgTask* pTask);
 int32_t ctgGetUdfInfoFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const char *funcName, SFuncInfo *out, SCtgTask* pTask);
 int32_t ctgGetUserDbAuthFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const char *user, SGetUserAuthRsp *out, SCtgTask* pTask);
-int32_t ctgGetTbMetaFromMnodeImpl(SCatalog* pCtg, SRequestConnInfo *pConn, char *dbFName, char* tbName, STableMetaOutput* out, SCtgTask* pTask);
-int32_t ctgGetTbMetaFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, STableMetaOutput* out, SCtgTask* pTask);
-int32_t ctgGetTbMetaFromVnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, SVgroupInfo *vgroupInfo, STableMetaOutput* out, SCtgTask* pTask);
+int32_t ctgGetTbMetaFromMnodeImpl(SCatalog* pCtg, SRequestConnInfo *pConn, char *dbFName, char* tbName, STableMetaOutput* out, SCtgTaskReq* tReq);
+int32_t ctgGetTbMetaFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, STableMetaOutput* out, SCtgTaskReq* tReq);
+int32_t ctgGetTbMetaFromVnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, SVgroupInfo *vgroupInfo, STableMetaOutput* out, SCtgTaskReq* tReq);
 int32_t ctgGetTableCfgFromVnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, SVgroupInfo *vgroupInfo, STableCfg **out, SCtgTask* pTask);
 int32_t ctgGetTableCfgFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, const SName* pTableName, STableCfg **out, SCtgTask* pTask);
 int32_t ctgGetSvrVerFromMnode(SCatalog* pCtg, SRequestConnInfo *pConn, char **out, SCtgTask* pTask);
@@ -664,6 +704,7 @@ void    ctgFreeJob(void* job);
 void    ctgFreeHandleImpl(SCatalog* pCtg);
 void    ctgFreeVgInfo(SDBVgInfo *vgInfo);
 int32_t ctgGetVgInfoFromHashValue(SCatalog *pCtg, SDBVgInfo *dbInfo, const SName *pTableName, SVgroupInfo *pVgroup);
+int32_t ctgGetVgInfosFromHashValue(SCatalog *pCtg, SCtgTaskReq* tReq, SDBVgInfo *dbInfo, SCtgTbHashsCtx *pCtx, char* dbFName, SArray* pNames, bool update);
 void    ctgResetTbMetaTask(SCtgTask* pTask);
 void    ctgFreeDbCache(SCtgDBCache *dbCache);
 int32_t ctgStbVersionSortCompare(const void* key1, const void* key2);
@@ -672,8 +713,11 @@ int32_t ctgStbVersionSearchCompare(const void* key1, const void* key2);
 int32_t ctgDbVgVersionSearchCompare(const void* key1, const void* key2);
 void    ctgFreeSTableMetaOutput(STableMetaOutput* pOutput);
 int32_t ctgUpdateMsgCtx(SCtgMsgCtx* pCtx, int32_t reqType, void* out, char* target);
+int32_t ctgAddMsgCtx(SArray* pCtxs, int32_t reqType, void* out, char* target);
 char *  ctgTaskTypeStr(CTG_TASK_TYPE type);
 int32_t ctgUpdateSendTargetInfo(SMsgSendInfo *pMsgSendInfo, int32_t msgType, char* dbFName, int32_t vgId);
+int32_t ctgGetTablesReqNum(SArray *pList);
+int32_t ctgAddFetch(SArray** pFetchs, int32_t dbIdx, int32_t tbIdx, int32_t *fetchIdx, int32_t resIdx, int32_t flag);
 int32_t ctgCloneTableIndex(SArray* pIndex, SArray** pRes);
 void    ctgFreeSTableIndex(void *info);
 void    ctgClearSubTaskRes(SCtgSubRes *pRes);
@@ -682,6 +726,7 @@ void    ctgClearHandle(SCatalog* pCtg);
 void    ctgFreeTbCacheImpl(SCtgTbCache *pCache);
 int32_t ctgRemoveTbMeta(SCatalog* pCtg, SName* pTableName);
 int32_t ctgGetTbHashVgroup(SCatalog *pCtg, SRequestConnInfo *pConn, const SName *pTableName, SVgroupInfo *pVgroup);
+SName*  ctgGetFetchName(SArray* pNames, SCtgFetch* pFetch);
 
 
 extern SCatalogMgmt gCtgMgmt;

source/libs/catalog/src/ctgCache.c

@@ -242,7 +242,6 @@ int32_t ctgAcquireTbMetaFromCache(SCatalog* pCtg, char *dbFName, char* tbName, S
     goto _return;
   }
   
-  int32_t sz = 0;
   pCache = taosHashAcquire(dbCache->tbCache, tbName, strlen(tbName));
   if (NULL == pCache) {
     ctgDebug("tb %s not in cache, dbFName:%s", tbName, dbFName);
@@ -282,7 +281,6 @@ int32_t ctgAcquireStbMetaFromCache(SCatalog* pCtg, char *dbFName, uint64_t suid,
     goto _return;
   }
   
-  int32_t sz = 0;
   char* stName = taosHashAcquire(dbCache->stbCache, &suid, sizeof(suid));
   if (NULL == stName) {
     ctgDebug("stb 0x%" PRIx64 " not in cache, dbFName:%s", suid, dbFName);
@@ -2152,6 +2150,254 @@ int32_t ctgGetTbMetaFromCache(SCatalog* pCtg, SRequestConnInfo *pConn, SCtgTbMet
   return TSDB_CODE_SUCCESS;
 }
 
+#if 0
+int32_t ctgGetTbMetaBFromCache(SCatalog* pCtg, SRequestConnInfo *pConn, SCtgTbMetasCtx* ctx, SArray** pResList) {
+  int32_t tbNum = taosArrayGetSize(ctx->pNames);
+  SName* fName = taosArrayGet(ctx->pNames, 0);
+  int32_t fIdx = 0;
+  
+  for (int32_t i = 0; i < tbNum; ++i) {
+    SName* pName = taosArrayGet(ctx->pNames, i);
+    SCtgTbMetaCtx nctx = {0};
+    nctx.flag = CTG_FLAG_UNKNOWN_STB;
+    nctx.pName = pName;
+    
+    if (IS_SYS_DBNAME(pName->dbname)) {
+      CTG_FLAG_SET_SYS_DB(nctx.flag);
+    }
+
+    STableMeta *pTableMeta = NULL;
+    CTG_ERR_RET(ctgReadTbMetaFromCache(pCtg, &nctx, &pTableMeta));
+    SMetaRes res = {0};
+    
+    if (pTableMeta) {
+      if (CTG_FLAG_MATCH_STB(nctx.flag, pTableMeta->tableType) &&
+          ((!CTG_FLAG_IS_FORCE_UPDATE(nctx.flag)) || (CTG_FLAG_IS_SYS_DB(nctx.flag)))) {
+        res.pRes = pTableMeta;
+      } else {
+        taosMemoryFreeClear(pTableMeta);
+      }
+    }
+
+    if (NULL == res.pRes) {
+      if (NULL == ctx->pFetchs) {
+        ctx->pFetchs = taosArrayInit(tbNum, sizeof(SCtgFetch));
+      }
+      
+      if (CTG_FLAG_IS_UNKNOWN_STB(nctx.flag)) {
+        CTG_FLAG_SET_STB(nctx.flag, nctx.tbInfo.tbType);
+      }
+
+      SCtgFetch fetch = {0};
+      fetch.tbIdx = i;
+      fetch.fetchIdx = fIdx++;
+      fetch.flag = nctx.flag;
+
+      taosArrayPush(ctx->pFetchs, &fetch);
+    }
+    
+    taosArrayPush(ctx->pResList, &res);
+  }
+
+  if (NULL == ctx->pFetchs) {
+    TSWAP(*pResList, ctx->pResList);
+  }
+
+  return TSDB_CODE_SUCCESS;
+}
+#endif
+
+int32_t ctgGetTbMetasFromCache(SCatalog* pCtg, SRequestConnInfo *pConn, SCtgTbMetasCtx* ctx, int32_t dbIdx, int32_t *fetchIdx, int32_t baseResIdx, SArray* pList) {
+  int32_t tbNum = taosArrayGetSize(pList);
+  SName* pName = taosArrayGet(pList, 0);
+  char dbFName[TSDB_DB_FNAME_LEN] = {0};
+  int32_t flag = CTG_FLAG_UNKNOWN_STB;
+  uint64_t lastSuid = 0;
+  STableMeta* lastTableMeta = NULL;
+  
+  if (IS_SYS_DBNAME(pName->dbname)) {
+    CTG_FLAG_SET_SYS_DB(flag);
+    strcpy(dbFName, pName->dbname);
+  } else {
+    tNameGetFullDbName(pName, dbFName);
+  }
+
+  SCtgDBCache *dbCache = NULL;
+  SCtgTbCache* pCache = NULL;
+  ctgAcquireDBCache(pCtg, dbFName, &dbCache);
+  
+  if (NULL == dbCache) {
+    ctgDebug("db %s not in cache", dbFName);
+    for (int32_t i = 0; i < tbNum; ++i) {
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+    }
+
+    return TSDB_CODE_SUCCESS;
+  }
+
+  for (int32_t i = 0; i < tbNum; ++i) {
+    SName* pName = taosArrayGet(pList, i);
+
+    pCache = taosHashAcquire(dbCache->tbCache, pName->tname, strlen(pName->tname));
+    if (NULL == pCache) {
+      ctgDebug("tb %s not in cache, dbFName:%s", pName->tname, dbFName);
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+      
+      continue;
+    }
+
+    CTG_LOCK(CTG_READ, &pCache->metaLock);
+    if (NULL == pCache->pMeta) {
+      ctgDebug("tb %s meta not in cache, dbFName:%s", pName->tname, dbFName);
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+      
+      continue;
+    }
+
+    STableMeta* tbMeta = pCache->pMeta;
+
+    SCtgTbMetaCtx nctx = {0};
+    nctx.flag = flag;
+    nctx.tbInfo.inCache = true;
+    nctx.tbInfo.dbId = dbCache->dbId;
+    nctx.tbInfo.suid = tbMeta->suid;
+    nctx.tbInfo.tbType = tbMeta->tableType;
+
+    SMetaRes res = {0};
+    STableMeta* pTableMeta = NULL;
+    if (tbMeta->tableType != TSDB_CHILD_TABLE) {
+      int32_t metaSize = CTG_META_SIZE(tbMeta);
+      pTableMeta = taosMemoryCalloc(1, metaSize);
+      if (NULL == pTableMeta) {
+        ctgReleaseTbMetaToCache(pCtg, dbCache, pCache);
+        CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+      }
+    
+      memcpy(pTableMeta, tbMeta, metaSize);
+      
+      CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+      taosHashRelease(dbCache->tbCache, pCache); 
+      
+      ctgDebug("Got tb %s meta from cache, type:%d, dbFName:%s", pName->tname, tbMeta->tableType, dbFName);
+      
+      res.pRes = pTableMeta;
+      taosArrayPush(ctx->pResList, &res);
+
+      continue;
+    }
+    
+    // PROCESS FOR CHILD TABLE
+
+    if (lastSuid && tbMeta->suid == lastSuid && lastTableMeta) {
+      cloneTableMeta(lastTableMeta, &pTableMeta);
+      memcpy(pTableMeta, tbMeta, sizeof(SCTableMeta));
+
+      CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+      taosHashRelease(dbCache->tbCache, pCache); 
+      
+      ctgDebug("Got tb %s meta from cache, type:%d, dbFName:%s", pName->tname, tbMeta->tableType, dbFName);
+      
+      res.pRes = pTableMeta;
+      taosArrayPush(ctx->pResList, &res);
+      
+      continue;
+    }
+    
+    int32_t metaSize = sizeof(SCTableMeta);
+    pTableMeta = taosMemoryCalloc(1, metaSize);
+    if (NULL == pTableMeta) {
+      ctgReleaseTbMetaToCache(pCtg, dbCache, pCache);    
+      CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+    }
+    
+    memcpy(pTableMeta, tbMeta, metaSize);
+    
+    CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+    taosHashRelease(dbCache->tbCache, pCache); 
+    
+    ctgDebug("Got ctb %s meta from cache, will continue to get its stb meta, type:%d, dbFName:%s", 
+             pName->tname, nctx.tbInfo.tbType, dbFName);
+    
+    char* stName = taosHashAcquire(dbCache->stbCache, &pTableMeta->suid, sizeof(pTableMeta->suid));
+    if (NULL == stName) {
+      ctgDebug("stb 0x%" PRIx64 " not in cache, dbFName:%s", pTableMeta->suid, dbFName);
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+
+      taosMemoryFreeClear(pTableMeta);
+      continue;
+    }
+
+    pCache = taosHashAcquire(dbCache->tbCache, stName, strlen(stName));
+    if (NULL == pCache) {
+      ctgDebug("stb 0x%" PRIx64 " name %s not in cache, dbFName:%s", pTableMeta->suid, stName, dbFName);
+      taosHashRelease(dbCache->stbCache, stName);
+      
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+
+      taosMemoryFreeClear(pTableMeta);      
+      continue;
+    }
+
+    taosHashRelease(dbCache->stbCache, stName);
+
+    CTG_LOCK(CTG_READ, &pCache->metaLock);
+    if (NULL == pCache->pMeta) {
+      ctgDebug("stb 0x%" PRIx64 " meta not in cache, dbFName:%s", pTableMeta->suid, dbFName);
+      CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+      taosHashRelease(dbCache->tbCache, pCache); 
+      
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+
+      taosMemoryFreeClear(pTableMeta);
+
+      continue;
+    }
+    
+    STableMeta* stbMeta = pCache->pMeta;
+    if (stbMeta->suid != nctx.tbInfo.suid) {    
+      CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+      taosHashRelease(dbCache->tbCache, pCache); 
+      
+      ctgError("stb suid 0x%" PRIx64 " in stbCache mis-match, expected suid 0x%"PRIx64 , stbMeta->suid, nctx.tbInfo.suid);
+      
+      ctgAddFetch(&ctx->pFetchs, dbIdx, i, fetchIdx, baseResIdx + i, flag);
+      taosArraySetSize(ctx->pResList, taosArrayGetSize(ctx->pResList) + 1);
+
+      taosMemoryFreeClear(pTableMeta);
+
+      continue;
+    }
+    
+    metaSize = CTG_META_SIZE(stbMeta);
+    pTableMeta = taosMemoryRealloc(pTableMeta, metaSize);
+    if (NULL == pTableMeta) {    
+      ctgReleaseTbMetaToCache(pCtg, dbCache, pCache);
+      CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+    }
+    
+    memcpy(&pTableMeta->sversion, &stbMeta->sversion, metaSize - sizeof(SCTableMeta));
+    
+    CTG_UNLOCK(CTG_READ, &pCache->metaLock);
+    taosHashRelease(dbCache->tbCache, pCache); 
+    
+    res.pRes = pTableMeta;
+    taosArrayPush(ctx->pResList, &res);
+
+    lastSuid = pTableMeta->suid;
+    lastTableMeta = pTableMeta;
+  }
+
+  ctgReleaseDBCache(pCtg, dbCache);
+  
+  return TSDB_CODE_SUCCESS;
+}
+
 
 int32_t ctgRemoveTbMetaFromCache(SCatalog* pCtg, SName* pTableName, bool syncReq) {
   int32_t       code = 0;

source/libs/catalog/src/ctgUtil.c

@@ -26,6 +26,7 @@ void ctgFreeMsgSendParam(void* param) {
 
   SCtgTaskCallbackParam* pParam = (SCtgTaskCallbackParam*)param;
   taosArrayDestroy(pParam->taskId);
+  taosArrayDestroy(pParam->msgIdx);
 
   taosMemoryFree(param);
 }
@@ -88,6 +89,10 @@ char *ctgTaskTypeStr(CTG_TASK_TYPE type) {
       return "[get user]";
     case CTG_TASK_GET_SVR_VER:
       return "[get svr ver]";
+    case CTG_TASK_GET_TB_META_BATCH:
+      return "[bget table meta]";
+    case CTG_TASK_GET_TB_HASH_BATCH:
+      return "[bget table hash]";
     default:
       return "unknown";
   }
@@ -460,6 +465,25 @@ void ctgResetTbMetaTask(SCtgTask* pTask) {
   taosMemoryFreeClear(pTask->res);
 }
 
+void ctgFreeBatchMeta(void* meta) {
+  if (NULL == meta) {
+    return;
+  }
+  
+  SMetaRes* pRes = (SMetaRes*)meta;
+  taosMemoryFreeClear(pRes->pRes);
+}
+
+void ctgFreeBatchHash(void* hash) {
+  if (NULL == hash) {
+    return;
+  }
+  
+  SMetaRes* pRes = (SMetaRes*)hash;
+  taosMemoryFreeClear(pRes->pRes);
+}
+
+
 void ctgFreeTaskRes(CTG_TASK_TYPE type, void **pRes) {
   switch (type) {
     case CTG_TASK_GET_QNODE:
@@ -500,6 +524,24 @@ void ctgFreeTaskRes(CTG_TASK_TYPE type, void **pRes) {
       taosMemoryFreeClear(*pRes);
       break;
     }
+    case CTG_TASK_GET_TB_META_BATCH: {
+      SArray* pArray = (SArray*)*pRes;
+      int32_t num = taosArrayGetSize(pArray);
+      for (int32_t i = 0; i < num; ++i) {
+        ctgFreeBatchMeta(taosArrayGet(pArray, i));
+      }
+      *pRes = NULL; // no need to free it
+      break;
+    }
+    case CTG_TASK_GET_TB_HASH_BATCH: {
+      SArray* pArray = (SArray*)*pRes;
+      int32_t num = taosArrayGetSize(pArray);
+      for (int32_t i = 0; i < num; ++i) {
+        ctgFreeBatchHash(taosArrayGet(pArray, i));
+      }
+      *pRes = NULL; // no need to free it
+      break;
+    }    
     default:
       qError("invalid task type %d", type);
       break;
@@ -554,6 +596,16 @@ void ctgFreeSubTaskRes(CTG_TASK_TYPE type, void **pRes) {
       taosMemoryFreeClear(*pRes);
       break;
     }
+    case CTG_TASK_GET_TB_META_BATCH: {
+      taosArrayDestroyEx(*pRes, ctgFreeBatchMeta);
+      *pRes = NULL;
+      break;
+    }
+    case CTG_TASK_GET_TB_HASH_BATCH: {
+      taosArrayDestroyEx(*pRes, ctgFreeBatchHash);
+      *pRes = NULL;
+      break;
+    }
     default:
       qError("invalid task type %d", type);
       break;
@@ -583,12 +635,38 @@ void ctgFreeTaskCtx(SCtgTask* pTask) {
       taosMemoryFreeClear(pTask->taskCtx);
       break;
     }
+    case CTG_TASK_GET_TB_META_BATCH: {
+      SCtgTbMetasCtx* taskCtx = (SCtgTbMetasCtx*)pTask->taskCtx;
+      taosArrayDestroyEx(taskCtx->pResList, ctgFreeBatchMeta);
+      taosArrayDestroy(taskCtx->pFetchs);
+      // NO NEED TO FREE pNames
+
+      taosArrayDestroyEx(pTask->msgCtxs, (FDelete)ctgFreeMsgCtx);
+      
+      if (pTask->msgCtx.lastOut) {
+        ctgFreeSTableMetaOutput((STableMetaOutput*)pTask->msgCtx.lastOut);
+        pTask->msgCtx.lastOut = NULL;
+      }
+      taosMemoryFreeClear(pTask->taskCtx);
+      break;
+    }
     case CTG_TASK_GET_TB_HASH: {
       SCtgTbHashCtx* taskCtx = (SCtgTbHashCtx*)pTask->taskCtx;
       taosMemoryFreeClear(taskCtx->pName);
       taosMemoryFreeClear(pTask->taskCtx);      
       break;
     }
+    case CTG_TASK_GET_TB_HASH_BATCH: {
+      SCtgTbHashsCtx* taskCtx = (SCtgTbHashsCtx*)pTask->taskCtx;
+      taosArrayDestroyEx(taskCtx->pResList, ctgFreeBatchHash);
+      taosArrayDestroy(taskCtx->pFetchs);
+      // NO NEED TO FREE pNames
+
+      taosArrayDestroyEx(pTask->msgCtxs, (FDelete)ctgFreeMsgCtx);
+      
+      taosMemoryFreeClear(pTask->taskCtx);
+      break;
+    }    
     case CTG_TASK_GET_TB_INDEX: {
       SCtgTbIndexCtx* taskCtx = (SCtgTbIndexCtx*)pTask->taskCtx;
       taosMemoryFreeClear(taskCtx->pName);
@@ -679,6 +757,23 @@ int32_t ctgUpdateMsgCtx(SCtgMsgCtx* pCtx, int32_t reqType, void* out, char* targ
   return TSDB_CODE_SUCCESS;
 }
 
+int32_t ctgAddMsgCtx(SArray* pCtxs, int32_t reqType, void* out, char* target) {
+  SCtgMsgCtx ctx = {0};
+  
+  ctx.reqType = reqType;
+  ctx.out = out;
+  if (target) {
+    ctx.target = strdup(target);
+    if (NULL == ctx.target) {
+      CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+    }
+  }
+
+  taosArrayPush(pCtxs, &ctx);
+
+  return TSDB_CODE_SUCCESS;
+}
+
 
 int32_t ctgGetHashFunction(int8_t hashMethod, tableNameHashFp *fp) {
   switch (hashMethod) {
@@ -780,6 +875,104 @@ int32_t ctgGetVgInfoFromHashValue(SCatalog *pCtg, SDBVgInfo *dbInfo, const SName
   CTG_RET(code);
 }
 
+int32_t ctgGetVgInfosFromHashValue(SCatalog *pCtg, SCtgTaskReq* tReq, SDBVgInfo *dbInfo, SCtgTbHashsCtx *pCtx, char* dbFName, SArray* pNames, bool update) {
+  int32_t code = 0;
+  SCtgTask* pTask = tReq->pTask;
+  SMetaRes res = {0};
+  int32_t vgNum = taosHashGetSize(dbInfo->vgHash);
+  if (vgNum <= 0) {
+    ctgError("db vgroup cache invalid, db:%s, vgroup number:%d", dbFName, vgNum);
+    CTG_ERR_RET(TSDB_CODE_TSC_DB_NOT_SELECTED);
+  }
+
+  tableNameHashFp fp = NULL;
+  SVgroupInfo *vgInfo = NULL;
+
+  CTG_ERR_RET(ctgGetHashFunction(dbInfo->hashMethod, &fp));
+
+  int32_t tbNum = taosArrayGetSize(pNames);
+
+  if (1 == vgNum) {
+    void *pIter = taosHashIterate(dbInfo->vgHash, NULL);
+    for (int32_t i = 0; i < tbNum; ++i) {
+      vgInfo = taosMemoryMalloc(sizeof(SVgroupInfo));
+      if (NULL == vgInfo) {
+        CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+      }
+
+      *vgInfo = *(SVgroupInfo*)pIter;
+
+      ctgDebug("Got tb hash vgroup, vgId:%d, epNum %d, current %s port %d", vgInfo->vgId, vgInfo->epSet.numOfEps,
+        vgInfo->epSet.eps[vgInfo->epSet.inUse].fqdn, vgInfo->epSet.eps[vgInfo->epSet.inUse].port);
+
+      if (update) {
+        SCtgFetch* pFetch = taosArrayGet(pCtx->pFetchs, tReq->msgIdx);
+        SMetaRes *pRes = taosArrayGet(pCtx->pResList, pFetch->resIdx + i);
+        pRes->pRes = vgInfo;
+      } else {
+        res.pRes = vgInfo;
+        taosArrayPush(pCtx->pResList, &res);
+      }
+    }
+
+    return TSDB_CODE_SUCCESS;
+  }
+
+  char tbFullName[TSDB_TABLE_FNAME_LEN];
+  sprintf(tbFullName, "%s.", dbFName);
+  int32_t offset = strlen(tbFullName);
+  SName* pName = NULL;
+  int32_t tbNameLen = 0;
+  
+  for (int32_t i = 0; i < tbNum; ++i) {
+    pName = taosArrayGet(pNames, i);
+
+    tbNameLen = offset + strlen(pName->tname);
+    strcpy(tbFullName + offset, pName->tname);
+
+    uint32_t hashValue = (*fp)(tbFullName, (uint32_t)tbNameLen);
+
+    void *pIter = taosHashIterate(dbInfo->vgHash, NULL);
+    while (pIter) {
+      vgInfo = pIter;
+      if (hashValue >= vgInfo->hashBegin && hashValue <= vgInfo->hashEnd) {
+        taosHashCancelIterate(dbInfo->vgHash, pIter);
+        break;
+      }
+      
+      pIter = taosHashIterate(dbInfo->vgHash, pIter);
+      vgInfo = NULL;
+    }
+
+    if (NULL == vgInfo) {
+      ctgError("no hash range found for hash value [%u], db:%s, numOfVgId:%d", hashValue, dbFName, taosHashGetSize(dbInfo->vgHash));
+      CTG_ERR_RET(TSDB_CODE_CTG_INTERNAL_ERROR);
+    }
+
+    SVgroupInfo* pNewVg = taosMemoryMalloc(sizeof(SVgroupInfo));
+    if (NULL == pNewVg) {
+      CTG_ERR_RET(TSDB_CODE_OUT_OF_MEMORY);
+    }
+
+    *pNewVg = *vgInfo;
+
+    ctgDebug("Got tb %s hash vgroup, vgId:%d, epNum %d, current %s port %d", tbFullName, vgInfo->vgId, vgInfo->epSet.numOfEps,
+             vgInfo->epSet.eps[vgInfo->epSet.inUse].fqdn, vgInfo->epSet.eps[vgInfo->epSet.inUse].port);
+
+    if (update) {
+      SCtgFetch* pFetch = taosArrayGet(pCtx->pFetchs, tReq->msgIdx);
+      SMetaRes *pRes = taosArrayGet(pCtx->pResList, pFetch->resIdx + i);
+      pRes->pRes = pNewVg;
+    } else {
+      res.pRes = pNewVg;
+      taosArrayPush(pCtx->pResList, &res);
+    }    
+  }
+
+  CTG_RET(code);
+}
+
+
 int32_t ctgStbVersionSearchCompare(const void* key1, const void* key2) {
   if (*(uint64_t *)key1 < ((SSTableVersion*)key2)->suid) {
     return -1;
@@ -921,4 +1114,41 @@ int32_t ctgUpdateSendTargetInfo(SMsgSendInfo *pMsgSendInfo, int32_t msgType, cha
   return TSDB_CODE_SUCCESS;
 }
 
+int32_t ctgGetTablesReqNum(SArray *pList) {
+  if (NULL == pList) {
+    return 0;
+  }
+
+  int32_t total = 0;
+  int32_t n = taosArrayGetSize(pList);
+  for (int32_t i = 0; i < n; ++i) {
+    STablesReq *pReq = taosArrayGet(pList, i);
+    total += taosArrayGetSize(pReq->pTables);
+  }
+
+  return total;
+}
+
+int32_t ctgAddFetch(SArray** pFetchs, int32_t dbIdx, int32_t tbIdx, int32_t *fetchIdx, int32_t resIdx, int32_t flag) {
+  if (NULL == (*pFetchs)) {
+    *pFetchs = taosArrayInit(CTG_DEFAULT_FETCH_NUM, sizeof(SCtgFetch));
+  }
+  
+  SCtgFetch fetch = {0};
+  fetch.dbIdx = dbIdx;
+  fetch.tbIdx = tbIdx;
+  fetch.fetchIdx = (*fetchIdx)++;
+  fetch.resIdx = resIdx;
+  fetch.flag = flag;
+  
+  taosArrayPush(*pFetchs, &fetch);
+
+  return TSDB_CODE_SUCCESS;
+}
+
+SName* ctgGetFetchName(SArray* pNames, SCtgFetch* pFetch) {
+  STablesReq* pReq = (STablesReq*)taosArrayGet(pNames, pFetch->dbIdx);
+  return (SName*)taosArrayGet(pReq->pTables, pFetch->tbIdx);
+}
+
 

source/libs/executor/inc/executorimpl.h

@@ -755,6 +755,9 @@ typedef struct STimeSliceOperatorInfo {
   SInterval      interval;
   int64_t        current;
   SArray*        pPrevRow;      // SArray<SGroupValue>
+  SArray*        pNextRow;      // SArray<SGroupValue>
+  bool           isPrevRowSet;
+  bool           isNextRowSet;
   int32_t        fillType;      // fill type
   SColumn        tsCol;         // primary timestamp column
   SExprSupp      scalarSup;     // scalar calculation
@@ -1007,7 +1010,7 @@ int32_t decodeOperator(SOperatorInfo* ops, const char* data, int32_t length);
 
 void    setTaskStatus(SExecTaskInfo* pTaskInfo, int8_t status);
 int32_t createExecTaskInfoImpl(SSubplan* pPlan, SExecTaskInfo** pTaskInfo, SReadHandle* pHandle, uint64_t taskId,
-                               const char* sql, EOPTR_EXEC_MODEL model);
+                               char* sql, EOPTR_EXEC_MODEL model);
 int32_t createDataSinkParam(SDataSinkNode *pNode, void **pParam, qTaskInfo_t* pTaskInfo, SReadHandle* readHandle);
 int32_t getOperatorExplainExecInfo(SOperatorInfo* operatorInfo, SArray* pExecInfoList);
 

source/libs/executor/src/executil.c

@@ -191,6 +191,7 @@ SSDataBlock* createResDataBlock(SDataBlockDescNode* pNode) {
   pBlock->info.blockId = pNode->dataBlockId;
   pBlock->info.type = STREAM_INVALID;
   pBlock->info.calWin = (STimeWindow){.skey = INT64_MIN, .ekey = INT64_MAX};
+  pBlock->info.watermark = INT64_MIN;
 
   for (int32_t i = 0; i < numOfCols; ++i) {
     SSlotDescNode*  pDescNode = (SSlotDescNode*)nodesListGetNode(pNode->pSlots, i);

source/libs/executor/src/executor.c

@@ -357,7 +357,7 @@ int32_t qGetQueryTableSchemaVersion(qTaskInfo_t tinfo, char* dbName, char* table
 }
 
 int32_t qCreateExecTask(SReadHandle* readHandle, int32_t vgId, uint64_t taskId, SSubplan* pSubplan,
-                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, const char* sql, EOPTR_EXEC_MODEL model) {
+                        qTaskInfo_t* pTaskInfo, DataSinkHandle* handle, char* sql, EOPTR_EXEC_MODEL model) {
   assert(pSubplan != NULL);
   SExecTaskInfo** pTask = (SExecTaskInfo**)pTaskInfo;
 

source/libs/executor/src/executorimpl.c

@@ -1437,7 +1437,8 @@ static void setExecutionContext(SOperatorInfo* pOperator, int32_t numOfOutput, u
   pAggInfo->groupId = groupId;
 }
 
-static void doUpdateNumOfRows(SResultRow* pRow, int32_t numOfExprs, const int32_t* rowCellOffset) {
+static void doUpdateNumOfRows(SqlFunctionCtx* pCtx, SResultRow* pRow, int32_t numOfExprs, const int32_t* rowCellOffset) {
+  bool returnNotNull = false;
   for (int32_t j = 0; j < numOfExprs; ++j) {
     struct SResultRowEntryInfo* pResInfo = getResultEntryInfo(pRow, j, rowCellOffset);
     if (!isRowEntryInitialized(pResInfo)) {
@@ -1447,6 +1448,15 @@ static void doUpdateNumOfRows(SResultRow* pRow, int32_t numOfExprs, const int32_
     if (pRow->numOfRows < pResInfo->numOfRes) {
       pRow->numOfRows = pResInfo->numOfRes;
     }
+
+    if (fmIsNotNullOutputFunc(pCtx[j].functionId)) {
+      returnNotNull = true;
+    }
+  }
+  // if all expr skips all blocks, e.g. all null inputs for max function, output one row in final result.
+  //  except for first/last, which require not null output, output no rows
+  if (pRow->numOfRows == 0 && !returnNotNull) {
+    pRow->numOfRows = 1;
   }
 }
 
@@ -1458,7 +1468,7 @@ int32_t finalizeResultRowIntoResultDataBlock(SDiskbasedBuf* pBuf, SResultRowPosi
   SFilePage*  page = getBufPage(pBuf, resultRowPosition->pageId);
   SResultRow* pRow = (SResultRow*)((char*)page + resultRowPosition->offset);
 
-  doUpdateNumOfRows(pRow, numOfExprs, rowCellOffset);
+  doUpdateNumOfRows(pCtx, pRow, numOfExprs, rowCellOffset);
   if (pRow->numOfRows == 0) {
     releaseBufPage(pBuf, page);
     return 0;
@@ -1514,7 +1524,7 @@ int32_t doCopyToSDataBlock(SExecTaskInfo* pTaskInfo, SSDataBlock* pBlock, SExprI
 
     SResultRow* pRow = (SResultRow*)((char*)page + pPos->pos.offset);
 
-    doUpdateNumOfRows(pRow, numOfExprs, rowCellOffset);
+    doUpdateNumOfRows(pCtx, pRow, numOfExprs, rowCellOffset);
     if (pRow->numOfRows == 0) {
       pGroupResInfo->index += 1;
       releaseBufPage(pBuf, page);
@@ -4181,7 +4191,7 @@ SOperatorInfo* createOperatorTree(SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo
     int32_t children = 0;
     pOptr = createStreamFinalSessionAggOperatorInfo(ops[0], pPhyNode, pTaskInfo, children);
   } else if (QUERY_NODE_PHYSICAL_PLAN_STREAM_FINAL_SESSION == type) {
-    int32_t children = 1;
+    int32_t children = pHandle->numOfVgroups;
     pOptr = createStreamFinalSessionAggOperatorInfo(ops[0], pPhyNode, pTaskInfo, children);
   } else if (QUERY_NODE_PHYSICAL_PLAN_PARTITION == type) {
     pOptr = createPartitionOperatorInfo(ops[0], (SPartitionPhysiNode*)pPhyNode, pTaskInfo);
@@ -4493,7 +4503,7 @@ int32_t createDataSinkParam(SDataSinkNode* pNode, void** pParam, qTaskInfo_t* pT
 }
 
 int32_t createExecTaskInfoImpl(SSubplan* pPlan, SExecTaskInfo** pTaskInfo, SReadHandle* pHandle, uint64_t taskId,
-                               const char* sql, EOPTR_EXEC_MODEL model) {
+                               char* sql, EOPTR_EXEC_MODEL model) {
   uint64_t queryId = pPlan->id.queryId;
 
   int32_t code = TSDB_CODE_SUCCESS;
@@ -4504,6 +4514,7 @@ int32_t createExecTaskInfoImpl(SSubplan* pPlan, SExecTaskInfo** pTaskInfo, SRead
   }
 
   (*pTaskInfo)->sql = sql;
+  sql = NULL;
   (*pTaskInfo)->pSubplan = pPlan;
   (*pTaskInfo)->pRoot = createOperatorTree(pPlan->pNode, *pTaskInfo, pHandle, &(*pTaskInfo)->tableqinfoList,
                                            pPlan->pTagCond, pPlan->pTagIndexCond, pPlan->user);
@@ -4516,6 +4527,7 @@ int32_t createExecTaskInfoImpl(SSubplan* pPlan, SExecTaskInfo** pTaskInfo, SRead
   return code;
 
 _complete:
+  taosMemoryFree(sql);
   doDestroyTask(*pTaskInfo);
   terrno = code;
   return code;

source/libs/executor/src/timewindowoperator.c

@@ -1752,30 +1752,11 @@ void increaseTs(SqlFunctionCtx* pCtx) {
   }
 }
 
-SSDataBlock* createDeleteBlock() {
-  SSDataBlock* pBlock = taosMemoryCalloc(1, sizeof(SSDataBlock));
-  pBlock->info.hasVarCol = false;
-  pBlock->info.groupId = 0;
-  pBlock->info.rows = 0;
-  pBlock->info.type = STREAM_DELETE_RESULT;
-  pBlock->info.rowSize = sizeof(TSKEY) + sizeof(uint64_t);
-
-  pBlock->pDataBlock = taosArrayInit(2, sizeof(SColumnInfoData));
-  SColumnInfoData infoData = {0};
-  infoData.info.type = TSDB_DATA_TYPE_TIMESTAMP;
-  infoData.info.bytes = sizeof(TSKEY);
-  // window start ts
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-
-  infoData.info.type = TSDB_DATA_TYPE_UBIGINT;
-  infoData.info.bytes = sizeof(uint64_t);
-  taosArrayPush(pBlock->pDataBlock, &infoData);
-
-  return pBlock;
-}
-
 void initIntervalDownStream(SOperatorInfo* downstream, uint8_t type, SAggSupporter* pSup) {
-  ASSERT(downstream->operatorType == QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN);
+  if (downstream->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) {
+    // Todo(liuyao) support partition by column
+    return;
+  }
   SStreamScanInfo* pScanInfo = downstream->info;
   pScanInfo->sessionSup.parentType = type;
   pScanInfo->sessionSup.pIntervalAggSup = pSup;
@@ -2084,10 +2065,30 @@ static void doKeepPrevRows(STimeSliceOperatorInfo* pSliceInfo, const SSDataBlock
       memcpy(pkey->pData, val, pkey->bytes);
     }
   }
+
+  pSliceInfo->isPrevRowSet = true;
+}
+
+static void doKeepNextRows(STimeSliceOperatorInfo* pSliceInfo, const SSDataBlock* pBlock, int32_t rowIndex) {
+  int32_t numOfCols = taosArrayGetSize(pBlock->pDataBlock);
+  for (int32_t i = 0; i < numOfCols; ++i) {
+    SColumnInfoData* pColInfoData = taosArrayGet(pBlock->pDataBlock, i);
+
+    // null data should not be kept since it can not be used to perform interpolation
+    if (!colDataIsNull_s(pColInfoData, i)) {
+      SGroupKeys* pkey = taosArrayGet(pSliceInfo->pNextRow, i);
+
+      pkey->isNull = false;
+      char* val = colDataGetData(pColInfoData, rowIndex);
+      memcpy(pkey->pData, val, pkey->bytes);
+    }
+  }
+
+  pSliceInfo->isNextRowSet = true;
 }
 
 static void genInterpolationResult(STimeSliceOperatorInfo* pSliceInfo, SExprSupp* pExprSup, SSDataBlock* pBlock,
-                                   int32_t rowIndex, SSDataBlock* pResBlock) {
+                                   SSDataBlock* pResBlock) {
   int32_t rows = pResBlock->info.rows;
 
   // todo set the correct primary timestamp column
@@ -2163,6 +2164,10 @@ static void genInterpolationResult(STimeSliceOperatorInfo* pSliceInfo, SExprSupp
         break;
       }
       case TSDB_FILL_PREV: {
+        if (!pSliceInfo->isPrevRowSet) {
+          break;
+        }
+
         SGroupKeys* pkey = taosArrayGet(pSliceInfo->pPrevRow, srcSlot);
         colDataAppend(pDst, rows, pkey->pData, false);
         pResBlock->info.rows += 1;
@@ -2170,8 +2175,12 @@ static void genInterpolationResult(STimeSliceOperatorInfo* pSliceInfo, SExprSupp
       }
 
       case TSDB_FILL_NEXT: {
-        char* p = colDataGetData(pSrc, rowIndex);
-        colDataAppend(pDst, rows, p, colDataIsNull_s(pSrc, rowIndex));
+        if (!pSliceInfo->isNextRowSet) {
+          break;
+        }
+
+        SGroupKeys* pkey = taosArrayGet(pSliceInfo->pNextRow, srcSlot);
+        colDataAppend(pDst, rows, pkey->pData, false);
         pResBlock->info.rows += 1;
         break;
       }
@@ -2205,6 +2214,35 @@ static int32_t initPrevRowsKeeper(STimeSliceOperatorInfo* pInfo, SSDataBlock* pB
     taosArrayPush(pInfo->pPrevRow, &key);
   }
 
+  pInfo->isPrevRowSet = false;
+
+  return TSDB_CODE_SUCCESS;
+}
+
+static int32_t initNextRowsKeeper(STimeSliceOperatorInfo* pInfo, SSDataBlock* pBlock) {
+  if (pInfo->pNextRow != NULL) {
+    return TSDB_CODE_SUCCESS;
+  }
+
+  pInfo->pNextRow = taosArrayInit(4, sizeof(SGroupKeys));
+  if (pInfo->pNextRow == NULL) {
+    return TSDB_CODE_OUT_OF_MEMORY;
+  }
+
+  int32_t numOfCols = taosArrayGetSize(pBlock->pDataBlock);
+  for (int32_t i = 0; i < numOfCols; ++i) {
+    SColumnInfoData* pColInfo = taosArrayGet(pBlock->pDataBlock, i);
+
+    SGroupKeys key = {0};
+    key.bytes = pColInfo->info.bytes;
+    key.type = pColInfo->info.type;
+    key.isNull = false;
+    key.pData = taosMemoryCalloc(1, pColInfo->info.bytes);
+    taosArrayPush(pInfo->pNextRow, &key);
+  }
+
+  pInfo->isNextRowSet = false;
+
   return TSDB_CODE_SUCCESS;
 }
 
@@ -2234,14 +2272,19 @@ static SSDataBlock* doTimeslice(SOperatorInfo* pOperator) {
 
   blockDataCleanup(pResBlock);
 
-  int32_t numOfRows = 0;
   while (1) {
     SSDataBlock* pBlock = downstream->fpSet.getNextFn(downstream);
     if (pBlock == NULL) {
       break;
     }
 
-    int32_t code = initPrevRowsKeeper(pSliceInfo, pBlock);
+    int32_t code;
+    code = initPrevRowsKeeper(pSliceInfo, pBlock);
+    if (code != TSDB_CODE_SUCCESS) {
+      longjmp(pTaskInfo->env, code);
+    }
+
+    code = initNextRowsKeeper(pSliceInfo, pBlock);
     if (code != TSDB_CODE_SUCCESS) {
       longjmp(pTaskInfo->env, code);
     }
@@ -2263,7 +2306,7 @@ static SSDataBlock* doTimeslice(SOperatorInfo* pOperator) {
           SColumnInfoData* pDst = taosArrayGet(pResBlock->pDataBlock, dstSlot);
 
           char* v = colDataGetData(pSrc, i);
-          colDataAppend(pDst, numOfRows, v, false);
+          colDataAppend(pDst, pResBlock->info.rows, v, false);
         }
 
         pResBlock->info.rows += 1;
@@ -2280,11 +2323,16 @@ static SSDataBlock* doTimeslice(SOperatorInfo* pOperator) {
           break;
         }
       } else if (ts < pSliceInfo->current) {
+        // in case interpolation window starts and ends between two datapoints, fill(prev) need to interpolate
+        doKeepPrevRows(pSliceInfo, pBlock, i);
+
         if (i < pBlock->info.rows - 1) {
+          // in case interpolation window starts and ends between two datapoints, fill(next) need to interpolate
+          doKeepNextRows(pSliceInfo, pBlock, i + 1);
           int64_t nextTs = *(int64_t*)colDataGetData(pTsCol, i + 1);
           if (nextTs > pSliceInfo->current) {
             while (pSliceInfo->current < nextTs && pSliceInfo->current <= pSliceInfo->win.ekey) {
-              genInterpolationResult(pSliceInfo, &pOperator->exprSupp, pBlock, i, pResBlock);
+              genInterpolationResult(pSliceInfo, &pOperator->exprSupp, pBlock, pResBlock);
               pSliceInfo->current =
                   taosTimeAdd(pSliceInfo->current, pInterval->interval, pInterval->intervalUnit, pInterval->precision);
               if (pResBlock->info.rows >= pResBlock->info.capacity) {
@@ -2303,10 +2351,38 @@ static SSDataBlock* doTimeslice(SOperatorInfo* pOperator) {
           doKeepPrevRows(pSliceInfo, pBlock, i);
         }
       } else {  // ts > pSliceInfo->current
+        // in case interpolation window starts and ends between two datapoints, fill(next) need to interpolate
+        doKeepNextRows(pSliceInfo, pBlock, i);
+
         while (pSliceInfo->current < ts && pSliceInfo->current <= pSliceInfo->win.ekey) {
-          genInterpolationResult(pSliceInfo, &pOperator->exprSupp, pBlock, i, pResBlock);
+          genInterpolationResult(pSliceInfo, &pOperator->exprSupp, pBlock, pResBlock);
+          pSliceInfo->current =
+              taosTimeAdd(pSliceInfo->current, pInterval->interval, pInterval->intervalUnit, pInterval->precision);
+          if (pResBlock->info.rows >= pResBlock->info.capacity) {
+            break;
+          }
+        }
+
+        // add current row if timestamp match
+        if (ts == pSliceInfo->current && pSliceInfo->current <= pSliceInfo->win.ekey) {
+          for (int32_t j = 0; j < pOperator->exprSupp.numOfExprs; ++j) {
+            SExprInfo* pExprInfo = &pOperator->exprSupp.pExprInfo[j];
+            int32_t    dstSlot = pExprInfo->base.resSchema.slotId;
+            int32_t    srcSlot = pExprInfo->base.pParam[0].pCol->slotId;
+
+            SColumnInfoData* pSrc = taosArrayGet(pBlock->pDataBlock, srcSlot);
+            SColumnInfoData* pDst = taosArrayGet(pResBlock->pDataBlock, dstSlot);
+
+            char* v = colDataGetData(pSrc, i);
+            colDataAppend(pDst, pResBlock->info.rows, v, false);
+          }
+
+          pResBlock->info.rows += 1;
+          doKeepPrevRows(pSliceInfo, pBlock, i);
+
           pSliceInfo->current =
               taosTimeAdd(pSliceInfo->current, pInterval->interval, pInterval->intervalUnit, pInterval->precision);
+
           if (pResBlock->info.rows >= pResBlock->info.capacity) {
             break;
           }
@@ -2318,6 +2394,17 @@ static SSDataBlock* doTimeslice(SOperatorInfo* pOperator) {
         }
       }
     }
+
+    // check if need to interpolate after ts range
+    // except for fill(next)
+    while (pSliceInfo->current <= pSliceInfo->win.ekey && pSliceInfo->fillType != TSDB_FILL_NEXT) {
+      genInterpolationResult(pSliceInfo, &pOperator->exprSupp, pBlock, pResBlock);
+      pSliceInfo->current =
+          taosTimeAdd(pSliceInfo->current, pInterval->interval, pInterval->intervalUnit, pInterval->precision);
+      if (pResBlock->info.rows >= pResBlock->info.capacity) {
+        break;
+      }
+    }
   }
 
   // restore the value
@@ -2836,13 +2923,6 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
       // process the rest of the data
       return pInfo->pUpdateRes;
     }
-    // doBuildPullDataBlock(pInfo->pPullWins, &pInfo->pullIndex, pInfo->pPullDataRes);
-    // if (pInfo->pPullDataRes->info.rows != 0) {
-    //   // process the rest of the data
-    //   ASSERT(IS_FINAL_OP(pInfo));
-    //   printDataBlock(pInfo->pPullDataRes, IS_FINAL_OP(pInfo) ? "interval final" : "interval semi");
-    //   return pInfo->pPullDataRes;
-    // }
     doBuildDeleteResult(pInfo->pDelWins, &pInfo->delIndex, pInfo->pDelRes);
     if (pInfo->pDelRes->info.rows != 0) {
       // process the rest of the data
@@ -2862,6 +2942,7 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
     }
     printDataBlock(pBlock, IS_FINAL_OP(pInfo) ? "interval final recv" : "interval semi recv");
     maxTs = TMAX(maxTs, pBlock->info.window.ekey);
+    maxTs = TMAX(maxTs, pBlock->info.watermark);
 
     if (pBlock->info.type == STREAM_NORMAL || pBlock->info.type == STREAM_PULL_DATA ||
         pBlock->info.type == STREAM_INVALID) {
@@ -2950,6 +3031,8 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
     closeIntervalWindow(pInfo->aggSup.pResultRowHashTable, &pInfo->twAggSup, &pInfo->interval, pInfo->pPullDataMap,
                         pUpdated, pInfo->pRecycledPages, pInfo->aggSup.pResultBuf);
     closeChildIntervalWindow(pInfo->pChildren, pInfo->twAggSup.maxTs);
+  } else {
+    pInfo->binfo.pRes->info.watermark = pInfo->twAggSup.maxTs;
   }
 
   finalizeUpdatedResult(pOperator->exprSupp.numOfExprs, pInfo->aggSup.pResultBuf, pUpdated, pSup->rowEntryInfoOffset);
@@ -2984,7 +3067,6 @@ static SSDataBlock* doStreamFinalIntervalAgg(SOperatorInfo* pOperator) {
     printDataBlock(pInfo->pDelRes, IS_FINAL_OP(pInfo) ? "interval final" : "interval semi");
     return pInfo->pDelRes;
   }
-  // ASSERT(false);
   return NULL;
 }
 
@@ -2996,6 +3078,7 @@ SSDataBlock* createSpecialDataBlock(EStreamType type) {
   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};
@@ -3185,7 +3268,6 @@ void destroyStreamSessionAggOperatorInfo(void* param, int32_t numOfOutput) {
       SStreamSessionAggOperatorInfo* pChInfo = pChild->info;
       destroyStreamSessionAggOperatorInfo(pChInfo, numOfOutput);
       taosMemoryFreeClear(pChild);
-      taosMemoryFreeClear(pChInfo);
     }
   }
   colDataDestroy(&pInfo->twAggSup.timeWindowData);
@@ -3950,6 +4032,7 @@ static SSDataBlock* doStreamSessionAgg(SOperatorInfo* pOperator) {
       doStreamSessionAggImpl(pChildOp, pBlock, NULL, NULL, true);
     }
     maxTs = TMAX(maxTs, pBlock->info.window.ekey);
+    maxTs = TMAX(maxTs, pBlock->info.watermark);
   }
 
   pInfo->twAggSup.maxTs = TMAX(pInfo->twAggSup.maxTs, maxTs);
@@ -4073,6 +4156,7 @@ static SSDataBlock* doStreamSessionSemiAgg(SOperatorInfo* pOperator) {
   }
 
   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

source/libs/function/src/functionMgt.c

@@ -221,6 +221,18 @@ bool fmIsLastRowFunc(int32_t funcId) {
   return FUNCTION_TYPE_LAST_ROW == funcMgtBuiltins[funcId].type;
 }
 
+bool fmIsNotNullOutputFunc(int32_t funcId) {
+  if (funcId < 0 || funcId >= funcMgtBuiltinsNum) {
+    return false;
+  }
+  return FUNCTION_TYPE_LAST == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_LAST_PARTIAL == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_LAST_MERGE == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST_PARTIAL == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST_MERGE == funcMgtBuiltins[funcId].type;
+}
+
 bool fmIsSelectValueFunc(int32_t funcId) {
   if (funcId < 0 || funcId >= funcMgtBuiltinsNum) {
     return false;

source/libs/parser/inc/parUtil.h

@@ -39,6 +39,11 @@ typedef struct SMsgBuf {
   char*   buf;
 } SMsgBuf;
 
+typedef struct SParseTablesMetaReq {
+  char      dbFName[TSDB_DB_FNAME_LEN];
+  SHashObj* pTables;
+} SParseTablesMetaReq;
+
 typedef struct SParseMetaCache {
   SHashObj* pTableMeta;    // key is tbFName, element is STableMeta*
   SHashObj* pDbVgroup;     // key is dbFName, element is SArray<SVgroupInfo>*
@@ -95,7 +100,7 @@ int32_t getUdfInfoFromCache(SParseMetaCache* pMetaCache, const char* pFunc, SFun
 int32_t getTableIndexFromCache(SParseMetaCache* pMetaCache, const SName* pName, SArray** pIndexes);
 int32_t getTableCfgFromCache(SParseMetaCache* pMetaCache, const SName* pName, STableCfg** pOutput);
 int32_t getDnodeListFromCache(SParseMetaCache* pMetaCache, SArray** pDnodes);
-void    destoryParseMetaCache(SParseMetaCache* pMetaCache);
+void    destoryParseMetaCache(SParseMetaCache* pMetaCache, bool request);
 
 #ifdef __cplusplus
 }

source/libs/parser/src/parser.c

@@ -187,7 +187,7 @@ int32_t qParseSqlSyntax(SParseContext* pCxt, SQuery** pQuery, struct SCatalogReq
   if (TSDB_CODE_SUCCESS == code) {
     code = buildCatalogReq(&metaCache, pCatalogReq);
   }
-  destoryParseMetaCache(&metaCache);
+  destoryParseMetaCache(&metaCache, true);
   terrno = code;
   return code;
 }
@@ -203,7 +203,7 @@ int32_t qAnalyseSqlSemantic(SParseContext* pCxt, const struct SCatalogReq* pCata
       code = analyseSemantic(pCxt, pQuery, &metaCache);
     }
   }
-  destoryParseMetaCache(&metaCache);
+  destoryParseMetaCache(&metaCache, false);
   terrno = code;
   return code;
 }

source/libs/parser/test/mockCatalogService.h

@@ -50,6 +50,7 @@ struct MockTableMeta {
 class MockCatalogServiceImpl;
 class MockCatalogService {
  public:
+  static void destoryTablesReq(void* p);
   static void destoryCatalogReq(SCatalogReq* pReq);
   static void destoryMetaRes(void* p);
   static void destoryMetaArrayRes(void* p);

source/libs/planner/src/planner.c

@@ -85,7 +85,7 @@ static int32_t setSubplanExecutionNode(SPhysiNode* pNode, int32_t groupId, SDown
 }
 
 int32_t qSetSubplanExecutionNode(SSubplan* subplan, int32_t groupId, SDownstreamSourceNode* pSource) {
-  planDebug("QID:0x%" PRIx64 " set subplan execution node, groupId:%d", subplan->id.groupId, groupId);
+  planDebug("QID:0x%" PRIx64 " set subplan execution node, groupId:%d", subplan->id.queryId, groupId);
   return setSubplanExecutionNode(subplan->pNode, groupId, pSource);
 }
 
@@ -104,7 +104,10 @@ static void clearSubplanExecutionNode(SPhysiNode* pNode) {
   FOREACH(pChild, pNode->pChildren) { clearSubplanExecutionNode((SPhysiNode*)pChild); }
 }
 
-void qClearSubplanExecutionNode(SSubplan* pSubplan) { clearSubplanExecutionNode(pSubplan->pNode); }
+void qClearSubplanExecutionNode(SSubplan* pSubplan) {
+  planDebug("QID:0x%" PRIx64 " clear subplan execution node, groupId:%d", pSubplan->id.queryId, pSubplan->id.groupId);
+  clearSubplanExecutionNode(pSubplan->pNode);
+}
 
 int32_t qSubPlanToString(const SSubplan* pSubplan, char** pStr, int32_t* pLen) {
   if (SUBPLAN_TYPE_MODIFY == pSubplan->subplanType && NULL == pSubplan->pNode) {

source/libs/qcom/src/queryUtil.c

@@ -462,3 +462,5 @@ int32_t cloneDbVgInfo(SDBVgInfo* pSrc, SDBVgInfo** pDst) {
 
   return TSDB_CODE_SUCCESS;
 }
+
+