Merge pull request #22456 from taosdata/3.0

merge 3.0 to feat/TD-22970

cmake/apr-util_CMakeLists.txt.in

+# apr-util
+ExternalProject_Add(aprutil-1
+        URL https://dlcdn.apache.org//apr/apr-util-1.6.3.tar.gz
+        URL_HASH SHA256=2b74d8932703826862ca305b094eef2983c27b39d5c9414442e9976a9acf1983
+        DOWNLOAD_NO_PROGRESS 1
+        DOWNLOAD_DIR "${TD_CONTRIB_DIR}/deps-download"
+        #GIT_REPOSITORY https://github.com/apache/apr-util.git
+	#GIT_TAG 1.5.4
+        SOURCE_DIR "${TD_CONTRIB_DIR}/apr-util"
+        #BINARY_DIR ""
+        BUILD_IN_SOURCE TRUE
+        BUILD_ALWAYS 1
+        #UPDATE_COMMAND ""
+        CONFIGURE_COMMAND ./configure --prefix=$ENV{HOME}/.cos-local.1/ --with-apr=$ENV{HOME}/.cos-local.1
+        #CONFIGURE_COMMAND ./configure --with-apr=/usr/local/apr
+        BUILD_COMMAND make
+        INSTALL_COMMAND make install
+        TEST_COMMAND ""
+)

cmake/apr_CMakeLists.txt.in

+# apr
+ExternalProject_Add(apr-1
+        URL https://dlcdn.apache.org//apr/apr-1.7.4.tar.gz
+        URL_HASH SHA256=a4137dd82a185076fa50ba54232d920a17c6469c30b0876569e1c2a05ff311d9
+        DOWNLOAD_NO_PROGRESS 1
+        DOWNLOAD_DIR "${TD_CONTRIB_DIR}/deps-download"
+        #GIT_REPOSITORY https://github.com/apache/apr.git
+	#GIT_TAG 1.5.2
+        SOURCE_DIR "${TD_CONTRIB_DIR}/apr"
+        BUILD_IN_SOURCE TRUE
+        UPDATE_DISCONNECTED TRUE
+        BUILD_ALWAYS 1
+        #UPDATE_COMMAND ""
+        CONFIGURE_COMMAND ./configure --prefix=$ENV{HOME}/.cos-local.1/ --enable-shared=no
+        #CONFIGURE_COMMAND ./configure
+        BUILD_COMMAND make
+        INSTALL_COMMAND make install
+        TEST_COMMAND ""
+)

cmake/cmake.define

 cmake_minimum_required(VERSION 3.0)
-
 set(CMAKE_VERBOSE_MAKEFILE ON)
 set(TD_BUILD_TAOSA_INTERNAL FALSE)
 
@@ -78,6 +77,12 @@ ELSE ()
     SET(TD_TAOS_TOOLS TRUE)
 ENDIF ()
 
+IF (${TD_WINDOWS})
+    SET(TAOS_LIB taos_static)
+ELSE ()
+    SET(TAOS_LIB taos)
+ENDIF ()
+
 IF (TD_WINDOWS)
     MESSAGE("${Yellow} set compiler flag for Windows! ${ColourReset}")
     SET(COMMON_FLAGS "/w /D_WIN32 /DWIN32 /Zi /MTd")

cmake/cmake.options

@@ -125,6 +125,16 @@ option(
     ON
 )
 
+IF(${TD_LINUX})
+
+option(
+    BUILD_WITH_COS
+    "If build with cos"
+    ON
+)
+
+ENDIF ()
+
 option(
     BUILD_WITH_SQLITE
     "If build with sqlite" 

cmake/cos_CMakeLists.txt.in

+# cos
+ExternalProject_Add(cos
+        GIT_REPOSITORY https://github.com/tencentyun/cos-c-sdk-v5.git
+	GIT_TAG v5.0.16
+        SOURCE_DIR "${TD_CONTRIB_DIR}/cos-c-sdk-v5"
+        BINARY_DIR ""
+        #BUILD_IN_SOURCE TRUE
+        CONFIGURE_COMMAND ""
+        BUILD_COMMAND ""
+        INSTALL_COMMAND ""
+        TEST_COMMAND ""
+)

cmake/curl_CMakeLists.txt.in

+# curl
+ExternalProject_Add(curl
+        URL  https://curl.se/download/curl-8.2.1.tar.gz
+        DOWNLOAD_NO_PROGRESS 1
+        DOWNLOAD_DIR "${TD_CONTRIB_DIR}/deps-download"
+        #GIT_REPOSITORY https://github.com/curl/curl.git
+        #GIT_TAG curl-7_88_1
+        SOURCE_DIR "${TD_CONTRIB_DIR}/curl"
+        BUILD_IN_SOURCE TRUE
+        BUILD_ALWAYS 1
+        #UPDATE_COMMAND ""
+        CONFIGURE_COMMAND ./configure --prefix=$ENV{HOME}/.cos-local.1 --without-ssl --enable-shared=no --disable-ldap --disable-ldaps --without-brotli
+        #CONFIGURE_COMMAND ./configure --without-ssl
+        BUILD_COMMAND make
+        INSTALL_COMMAND make install
+        TEST_COMMAND ""
+)

cmake/mxml_CMakeLists.txt.in

+# cos
+ExternalProject_Add(mxml
+        GIT_REPOSITORY https://github.com/michaelrsweet/mxml.git
+	GIT_TAG release-2.10
+        SOURCE_DIR "${TD_CONTRIB_DIR}/mxml"
+        #BINARY_DIR ""
+        BUILD_IN_SOURCE TRUE
+        #UPDATE_COMMAND ""
+        CONFIGURE_COMMAND ./configure --prefix=$ENV{HOME}/.cos-local.1 --enable-shared=no
+        #CONFIGURE_COMMAND ./configure
+        BUILD_COMMAND make
+        INSTALL_COMMAND make install
+        TEST_COMMAND ""
+)

contrib/CMakeLists.txt

@@ -6,6 +6,39 @@ function(cat IN_FILE OUT_FILE)
     file(APPEND ${OUT_FILE} "${CONTENTS}")
 endfunction(cat IN_FILE OUT_FILE)
 
+if(${TD_LINUX})
+
+set(CONTRIB_TMP_FILE3 "${CMAKE_BINARY_DIR}/deps_tmp_CMakeLists.txt.in3")
+configure_file("${TD_SUPPORT_DIR}/deps_CMakeLists.txt.in" ${CONTRIB_TMP_FILE3})
+
+if(${BUILD_WITH_COS})
+    file(MAKE_DIRECTORY $ENV{HOME}/.cos-local.1/)
+    cat("${TD_SUPPORT_DIR}/mxml_CMakeLists.txt.in" ${CONTRIB_TMP_FILE3})
+    cat("${TD_SUPPORT_DIR}/apr_CMakeLists.txt.in" ${CONTRIB_TMP_FILE3})
+    cat("${TD_SUPPORT_DIR}/curl_CMakeLists.txt.in" ${CONTRIB_TMP_FILE3})
+endif(${BUILD_WITH_COS})
+
+configure_file(${CONTRIB_TMP_FILE3} "${TD_CONTRIB_DIR}/deps-download/CMakeLists.txt")
+execute_process(COMMAND "${CMAKE_COMMAND}" -G "${CMAKE_GENERATOR}" .
+        WORKING_DIRECTORY "${TD_CONTRIB_DIR}/deps-download")
+execute_process(COMMAND "${CMAKE_COMMAND}" --build .
+        WORKING_DIRECTORY "${TD_CONTRIB_DIR}/deps-download")
+
+set(CONTRIB_TMP_FILE2 "${CMAKE_BINARY_DIR}/deps_tmp_CMakeLists.txt.in2")
+configure_file("${TD_SUPPORT_DIR}/deps_CMakeLists.txt.in" ${CONTRIB_TMP_FILE2})
+
+if(${BUILD_WITH_COS})
+    cat("${TD_SUPPORT_DIR}/apr-util_CMakeLists.txt.in" ${CONTRIB_TMP_FILE2})
+endif(${BUILD_WITH_COS})
+
+configure_file(${CONTRIB_TMP_FILE2} "${TD_CONTRIB_DIR}/deps-download/CMakeLists.txt")
+execute_process(COMMAND "${CMAKE_COMMAND}" -G "${CMAKE_GENERATOR}" .
+        WORKING_DIRECTORY "${TD_CONTRIB_DIR}/deps-download")
+execute_process(COMMAND "${CMAKE_COMMAND}" --build .
+        WORKING_DIRECTORY "${TD_CONTRIB_DIR}/deps-download")
+
+endif(${TD_LINUX})
+
 set(CONTRIB_TMP_FILE "${CMAKE_BINARY_DIR}/deps_tmp_CMakeLists.txt.in")
 configure_file("${TD_SUPPORT_DIR}/deps_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
 
@@ -122,6 +155,16 @@ if(${BUILD_WITH_SQLITE})
     cat("${TD_SUPPORT_DIR}/sqlite_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
 endif(${BUILD_WITH_SQLITE})
 
+# cos
+if(${BUILD_WITH_COS})
+    #cat("${TD_SUPPORT_DIR}/mxml_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+    #cat("${TD_SUPPORT_DIR}/apr_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+    #cat("${TD_SUPPORT_DIR}/apr-util_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+    #cat("${TD_SUPPORT_DIR}/curl_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+    cat("${TD_SUPPORT_DIR}/cos_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+    add_definitions(-DUSE_COS)
+endif(${BUILD_WITH_COS})
+
 # lucene
 if(${BUILD_WITH_LUCENE})
     cat("${TD_SUPPORT_DIR}/lucene_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
@@ -347,6 +390,31 @@ if (${BUILD_WITH_ROCKSDB})
     endif()
 endif()
 
+# cos
+if(${BUILD_WITH_COS})
+  if(${TD_LINUX})
+    set(CMAKE_PREFIX_PATH $ENV{HOME}/.cos-local.1)
+    #ADD_DEFINITIONS(-DMINIXML_LIBRARY=${CMAKE_BINARY_DIR}/build/lib/libxml.a)
+    option(ENABLE_TEST "Enable the tests" OFF)
+    INCLUDE_DIRECTORIES($ENV{HOME}/.cos-local.1/include)
+    MESSAGE("$ENV{HOME}/.cos-local.1/include")
+
+    set(CMAKE_BUILD_TYPE debug)
+    set(ORIG_CMAKE_PROJECT_NAME ${CMAKE_PROJECT_NAME})
+    set(CMAKE_PROJECT_NAME cos_c_sdk)
+
+    add_subdirectory(cos-c-sdk-v5 EXCLUDE_FROM_ALL)
+    target_include_directories(
+      cos_c_sdk
+      PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/cos-c-sdk-v5/cos_c_sdk>
+      )
+
+    set(CMAKE_PROJECT_NAME ${ORIG_CMAKE_PROJECT_NAME})
+  else()
+
+  endif(${TD_LINUX})
+endif(${BUILD_WITH_COS})
+
 # lucene
 # To support build on ubuntu: sudo apt-get install libboost-all-dev
 if(${BUILD_WITH_LUCENE})

contrib/test/CMakeLists.txt

@@ -3,6 +3,11 @@ if(${BUILD_WITH_ROCKSDB})
     add_subdirectory(rocksdb)
 endif(${BUILD_WITH_ROCKSDB})
 
+# cos
+if(${BUILD_WITH_COS})
+    add_subdirectory(cos)
+endif(${BUILD_WITH_COS})
+
 if(${BUILD_WITH_LUCENE})
     add_subdirectory(lucene)
 endif(${BUILD_WITH_LUCENE})

contrib/test/cos/CMakeLists.txt

+add_executable(cosTest "")
+target_sources(cosTest
+  PRIVATE
+  "${CMAKE_CURRENT_SOURCE_DIR}/main.c"
+  )
+
+#find_path(APR_INCLUDE_DIR apr-1/apr_time.h)
+#find_path(APR_UTIL_INCLUDE_DIR apr/include/apr-1/apr_md5.h)
+#find_path(MINIXML_INCLUDE_DIR mxml.h)
+#find_path(CURL_INCLUDE_DIR curl/curl.h)
+
+#include_directories (${MINIXML_INCLUDE_DIR})
+#include_directories (${CURL_INCLUDE_DIR})
+FIND_PROGRAM(APR_CONFIG_BIN NAMES apr-config apr-1-config PATHS /usr/bin /usr/local/bin /usr/local/apr/bin/)
+#FIND_PROGRAM(APU_CONFIG_BIN NAMES apu-config apu-1-config PATHS /usr/bin /usr/local/bin /usr/local/apr/bin/)
+
+IF (APR_CONFIG_BIN)
+  EXECUTE_PROCESS(
+    COMMAND ${APR_CONFIG_BIN} --includedir
+    OUTPUT_VARIABLE APR_INCLUDE_DIR
+    OUTPUT_STRIP_TRAILING_WHITESPACE
+    )
+ENDIF()
+#IF (APU_CONFIG_BIN)
+#  EXECUTE_PROCESS(
+#    COMMAND ${APU_CONFIG_BIN} --includedir
+#    OUTPUT_VARIABLE APR_UTIL_INCLUDE_DIR
+#    OUTPUT_STRIP_TRAILING_WHITESPACE
+#    )
+#ENDIF()
+
+include_directories (${APR_INCLUDE_DIR})
+#include_directories (${APR_UTIL_INCLUDE_DIR})
+
+target_include_directories(
+  cosTest
+  PUBLIC "${TD_SOURCE_DIR}/contrib/cos-c-sdk-v5/cos_c_sdk"
+  )
+
+#find_library(APR_LIBRARY apr-1 PATHS /usr/local/apr/lib/)
+#find_library(APR_UTIL_LIBRARY aprutil-1 PATHS /usr/local/apr/lib/)
+#find_library(MINIXML_LIBRARY mxml)
+#find_library(CURL_LIBRARY curl)
+
+target_link_libraries(cosTest cos_c_sdk)
+target_link_libraries(cosTest apr-1})
+target_link_libraries(cosTest aprutil-1})
+target_link_libraries(cosTest mxml)
+target_link_libraries(cosTest curl)

docs/en/07-develop/06-stream.md

@@ -52,7 +52,7 @@ CREATE TABLE d1004 USING meters TAGS ("California.LosAngeles", 3);
 ### Create a Stream
 
 ```sql
-create stream current_stream into current_stream_output_stb as select _wstart as start, _wend as end, max(current) as max_current from meters where voltage <= 220 interval (5s);
+create stream current_stream trigger at_once into current_stream_output_stb as select _wstart as wstart, _wend as wend, max(current) as max_current from meters where voltage <= 220 interval (5s);
 ```
 
 ### Write Data

docs/en/07-develop/07-tmq.mdx

@@ -23,7 +23,20 @@ By subscribing to a topic, a consumer can obtain the latest data in that topic i
 
 To implement these features, TDengine indexes its write-ahead log (WAL) file for fast random access and provides configurable methods for replacing and retaining this file. You can define a retention period and size for this file. For information, see the CREATE DATABASE statement. In this way, the WAL file is transformed into a persistent storage engine that remembers the order in which events occur. However, note that configuring an overly long retention period for your WAL files makes database compression inefficient. TDengine then uses the WAL file instead of the time-series database as its storage engine for queries in the form of topics. TDengine reads the data from the WAL file; uses a unified query engine instance to perform filtering, transformations, and other operations; and finally pushes the data to consumers.
 
-Tips: Data subscription is to consume data from the wal. If some wal files are deleted according to WAL retention policy, the deleted data can't be consumed any more. So you need to set a reasonable value for parameter `WAL_RETENTION_PERIOD` or `WAL_RETENTION_SIZE` when creating the database and make sure your application consume the data in a timely way to make sure there is no data loss. This behavior is similar to Kafka and other widely used message queue products.
+Tips:(c interface for example)
+1. A consumption group consumes all data under the same topic, and different consumption groups are independent of each other;
+2. A consumption group consumes all vgroups of the same topic, which can be composed of multiple consumers, but a vgroup is only consumed by one consumer. If the number of consumers exceeds the number of vgroups, the excess consumers do not consume data;
+3. On the server side, only one offset is saved for each vgroup, and the offsets for each vgroup are monotonically increasing, but not necessarily continuous. There is no correlation between the offsets of various vgroups;
+4. Each poll server will return a result block, which belongs to a vgroup and may contain data from multiple versions of wal. This block can be accessed through tmq_get_vgroup_offset. The offset interface obtains the offset of the first record in the block;
+5. If a consumer group has never committed an offset, when its member consumers restart and pull data again, they start consuming from the set value of the parameter auto.offset.reset; In a consumer lifecycle, the client locally records the offset of the most recent pull data and will not pull duplicate data;
+6. If a consumer terminates abnormally (without calling tmq_close), they need to wait for about 12 seconds to trigger their consumer group rebalance. The consumer's status on the server will change to LOST, and after about 1 day, the consumer will be automatically deleted; Exit normally, and after exiting, the consumer will be deleted; Add a new consumer, wait for about 2 seconds to trigger Rebalance, and the consumer's status on the server will change to ready;
+7. The consumer group Rebalance will reassign Vgroups to all consumer members in the ready state of the group, and consumers can only assign/see/commit/poll operations to the Vgroups they are responsible for;
+8. Consumers can tmq_position to obtain the offset of the current consumption, seek to the specified offset, and consume again;
+9. Seek points the position to the specified offset without executing the commit operation. Once the seek is successful, it can poll the specified offset and subsequent data;
+10. Before the seek operation, tmq must be call tmq_get_topic_assignment, The assignment interface obtains the vgroup ID and offset range of the consumer. The seek operation will detect whether the vgroup ID and offset are legal, and if they are illegal, an error will be reported;
+11. Due to the existence of a WAL expiration deletion mechanism, even if the seek operation is successful, it is possible that the offset has expired when polling data. If the offset of poll is less than the WAL minimum version number, it will be consumed from the WAL minimum version number;
+12. The tmq_get_vgroup_offset interface obtains the offset of the first data in the result block where the record is located. When seeking to this offset, it will consume all the data in this block. Refer to point four;
+13. Data subscription is to consume data from the wal. If some wal files are deleted according to WAL retention policy, the deleted data can't be consumed any more. So you need to set a reasonable value for parameter `WAL_RETENTION_PERIOD` or `WAL_RETENTION_SIZE` when creating the database and make sure your application consume the data in a timely way to make sure there is no data loss. This behavior is similar to Kafka and other widely used message queue products.
 
 ## Data Schema and API
 
@@ -33,39 +46,58 @@ The related schemas and APIs in various languages are described as follows:
 <TabItem value="c" label="C">
 
 ```c
-typedef struct tmq_t      tmq_t;
-typedef struct tmq_conf_t tmq_conf_t;
-typedef struct tmq_list_t tmq_list_t;
-
-typedef void(tmq_commit_cb(tmq_t *, int32_t code, void *param));
-
-DLL_EXPORT tmq_list_t *tmq_list_new();
-DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
-DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
-DLL_EXPORT tmq_t *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
-DLL_EXPORT const char *tmq_err2str(int32_t code);
-
-DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
-DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
-DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
-DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
-DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg);
-DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
-
-enum tmq_conf_res_t {
-  TMQ_CONF_UNKNOWN = -2,
-  TMQ_CONF_INVALID = -1,
-  TMQ_CONF_OK = 0,
-};
-typedef enum tmq_conf_res_t tmq_conf_res_t;
-
-DLL_EXPORT tmq_conf_t    *tmq_conf_new();
-DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
-DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
-DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
-```
-
-For more information, see [C/C++ Connector](/reference/connector/cpp).
+    typedef struct tmq_t      tmq_t;
+    typedef struct tmq_conf_t tmq_conf_t;
+    typedef struct tmq_list_t tmq_list_t;
+
+    typedef void(tmq_commit_cb(tmq_t *tmq, int32_t code, void *param));
+
+    typedef enum tmq_conf_res_t {
+        TMQ_CONF_UNKNOWN = -2,
+        TMQ_CONF_INVALID = -1,
+        TMQ_CONF_OK = 0,
+    } tmq_conf_res_t;
+
+    typedef struct tmq_topic_assignment {
+        int32_t vgId;
+        int64_t currentOffset;
+        int64_t begin;
+        int64_t end;  // The last version of wal + 1
+    } tmq_topic_assignment;
+
+    DLL_EXPORT tmq_conf_t    *tmq_conf_new();
+    DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
+    DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
+    DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
+
+    DLL_EXPORT tmq_list_t *tmq_list_new();
+    DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
+    DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
+    DLL_EXPORT int32_t     tmq_list_get_size(const tmq_list_t *);
+    DLL_EXPORT char      **tmq_list_to_c_array(const tmq_list_t *);
+
+    DLL_EXPORT tmq_t    *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
+    DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
+    DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
+    DLL_EXPORT int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics);
+    DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
+    DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
+    DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg); //Commit the msg’s offset + 1
+    DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
+    DLL_EXPORT int32_t   tmq_commit_offset_sync(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
+    DLL_EXPORT void      tmq_commit_offset_async(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset, tmq_commit_cb *cb, void *param);
+    DLL_EXPORT int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment,int32_t *numOfAssignment);
+    DLL_EXPORT void      tmq_free_assignment(tmq_topic_assignment* pAssignment);
+    DLL_EXPORT int32_t   tmq_offset_seek(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
+    DLL_EXPORT int64_t   tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId);  // The current offset is the offset of the last consumed message + 1
+    DLL_EXPORT int64_t   tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId);
+
+    DLL_EXPORT const char *tmq_get_topic_name(TAOS_RES *res);
+    DLL_EXPORT const char *tmq_get_db_name(TAOS_RES *res);
+    DLL_EXPORT int32_t     tmq_get_vgroup_id(TAOS_RES *res);
+    DLL_EXPORT int64_t     tmq_get_vgroup_offset(TAOS_RES* res);  // Get current offset of the result
+    DLL_EXPORT const char *tmq_err2str(int32_t code);
+```
 
 The following example is based on the smart meter table described in Data Models. For complete sample code, see the C language section below.
 

docs/en/10-deployment/01-deploy.md

@@ -62,12 +62,13 @@ serverPort            6030
 
 For all the dnodes in a TDengine cluster, the below parameters must be configured exactly the same, any node whose configuration is different from dnodes already in the cluster can't join the cluster.
 
-| **#** | **Parameter**      | **Definition**                                                                    |
-| ----- | ------------------ | ------------------------------------------- |
-| 1     | statusInterval     | The interval by which dnode reports its status to mnode                           |
-| 2     | timezone           | Timezone                                                                          |
-| 3     | locale             | System region and encoding                       |
-| 4     | charset            | Character set |
+| **#** | **Parameter**    | **Definition**                                                                |
+| ----- | ---------------- | ----------------------------------------------------------------------------- |
+| 1     | statusInterval   | The interval by which dnode reports its status to mnode                       |
+| 2     | timezone         | Timezone                                                                      |
+| 3     | locale           | System region and encoding                                                    |
+| 4     | charset          | Character set                                                                 |
+| 5     | ttlChangeOnWrite | Whether the ttl expiration time changes with the table modification operation |
 
 ## Start Cluster
 
@@ -97,7 +98,7 @@ Then, on the first dnode i.e. h1.tdengine.com in our example, use TDengine CLI `
 CREATE DNODE "h2.taos.com:6030";
 ````
 
-This adds the end point of the new dnode (from Step 4) into the end point list of the cluster. In the command "fqdn:port" should be quoted using double quotes. Change `"h2.taos.com:6030"` to the end point of your new dnode. 
+This adds the end point of the new dnode (from Step 4) into the end point list of the cluster. In the command "fqdn:port" should be quoted using double quotes. Change `"h2.taos.com:6030"` to the end point of your new dnode.
 
 Then on the first dnode h1.tdengine.com, execute `show dnodes` in `taos`
 

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

@@ -698,7 +698,7 @@ ELAPSED(ts_primary_key [, time_unit])
 LEASTSQUARES(expr, start_val, step_val)
 ```
 
-**Description**: The linear regression function of the specified column and the timestamp column (primary key), `start_val` is the initial value and `step_val` is the step value.
+**Description**: The linear regression function of a specified column, `start_val` is the initial value and `step_val` is the step value.
 
 **Return value type**: A string in the format of "(slope, intercept)"
 

docs/en/12-taos-sql/25-grant.md

@@ -4,7 +4,7 @@ sidebar_label: Access Control
 description: This document describes how to manage users and permissions in TDengine.
 ---
 
-This document describes how to manage permissions in TDengine.
+User and Access control is a distingguished feature of TDengine enterprise edition. In this section, only the most fundamental functionalities of user and access control are demonstrated. To get the full knowledge of user and access control, please contact the TDengine team.
 
 ## Create a User
 

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

@@ -378,7 +378,7 @@ In addition to writing data using the SQL method or the parameter binding API, w
 - `TAOS_RES* taos_schemaless_insert(TAOS* taos, const char* lines[], int numLines, int protocol, int precision)`
 
   **Function description**
-  This interface writes the text data of the line protocol to TDengine.
+  - This interface writes the text data of the line protocol to TDengine.
 
   **Parameter description**
   - taos: database connection, established by the `taos_connect()` function.
@@ -387,12 +387,13 @@ In addition to writing data using the SQL method or the parameter binding API, w
   - protocol: the protocol type of the lines, used to identify the text data format.
   - precision: precision string for the timestamp in the text data.
 
-  **return value**
-  TAOS_RES structure, application can get error message by using `taos_errstr()` and also error code by using `taos_errno()`.
+  **Return value**
+  - TAOS_RES structure, application can get error message by using `taos_errstr()` and also error code by using `taos_errno()`.
   In some cases, the returned TAOS_RES is `NULL`, and it is still possible to call `taos_errno()` to safely get the error code information.
   The returned TAOS_RES needs to be freed by the caller in order to avoid memory leaks.
 
   **Description**
+  
   The protocol type is enumerated and contains the following three formats.
 
   - TSDB_SML_LINE_PROTOCOL: InfluxDB line protocol (Line Protocol)
@@ -427,3 +428,89 @@ In addition to writing data using the SQL method or the parameter binding API, w
   - Within _raw interfaces represent data through the passed parameters lines and len. In order to solve the problem that the original interface data contains '\0' and is truncated. The totalRows pointer returns the number of parsed data rows.
   - Within _ttl interfaces can pass the ttl parameter to control the ttl expiration time of the table.
   - Within _reqid interfaces can track the entire call chain by passing the reqid parameter.
+
+### Subscription API
+
+- `int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment, int32_t *numOfAssignment)`
+- `void      tmq_free_assignment(tmq_topic_assignment* pAssignment)`
+  
+  tmq_topic_assignment defined as follows:
+  ```c
+  typedef struct tmq_topic_assignment {
+    int32_t vgId;
+    int64_t currentOffset;
+    int64_t begin;
+    int64_t end;
+  } tmq_topic_assignment;
+  ```
+  **Function description**
+  - tmq_get_topic_assignment get the current vgroup information of this consumer
+
+  **Parameter description**
+  - numOfAssignment:the num of vgroups assigned to this consumer
+  - assignment:the information of vgroups, needed to be freed by tmq_free_assignment
+
+  **Return value**
+  - zero success,none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
+
+- `int64_t   tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId)`
+  **Function description**
+  - get the committed offset
+
+  **Return value**
+  - the value of committed offset, -2147467247 means no committed value, Other values less than 0 indicate failure
+
+- `int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg)`
+- `void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param)`
+- `int32_t   tmq_commit_offset_sync(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset)`
+- `void      tmq_commit_offset_async(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset, tmq_commit_cb *cb, void *param)`
+  
+  **Function description**
+
+  The commit interface is divided into two types, each with synchronous and asynchronous interfaces:
+  - The first type: based on message submission, submit the progress in the message. If the message passes NULL, submit the current progress of all vgroups consumed by the current consumer: tmq_commit_sync/tmq_commit_async
+  - The second type: submit based on the offset of a Vgroup in a topic: tmq_commit_offset_sync/tmq_commit_offset_async
+  
+  **Parameter description**
+  - msg：Message consumed, If the message passes NULL, submit the current progress of all vgroups consumed by the current consumer
+
+  **Return value**
+  - zero success, none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
+
+- `int64_t     tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId)`
+
+  **Function description**
+  - Obtain the current consumption location, which is the next location of the data consumed
+  
+  **Return value**
+  - the current consumption location, none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
+
+- `int32_t   tmq_offset_seek(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset)`
+
+  **Function description**
+  - Set the offset position of the consumer in a Vgroup of a certain topic to start consumption
+
+  **Return value**
+  - zero success, none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
+  
+- `int32_t   int64_t     tmq_get_vgroup_offset(TAOS_RES* res)`
+
+  **Function description**
+  - Obtain the starting offset of the consumed data
+
+  **Parameter description**
+  - msg：Message consumed
+
+  **Return value**
+  - the starting offset of the consumed data, none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
+  
+- `int32_t   int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics)`
+
+  **Function description**
+  - Obtain a list of topics subscribed by consumers
+
+  **Parameter description**
+  - topics: a list of topics subscribed by consumers，need to be freed by tmq_list_destroy
+
+  **Return value**
+  - zero success，none zero failed, wrong message can be obtained through `char *tmq_err2str(int32_t code)`
\ No newline at end of file

docs/en/14-reference/03-connector/04-java.mdx

@@ -1091,8 +1091,6 @@ public abstract class ConsumerLoop {
         config.setProperty("client.id", "1");
         config.setProperty("value.deserializer", "com.taosdata.jdbc.tmq.ConsumerTest.ConsumerLoop$ResultDeserializer");
         config.setProperty("value.deserializer.encoding", "UTF-8");
-        config.setProperty("experimental.snapshot.enable", "true");
-
 
         this.consumer = new TaosConsumer<>(config);
         this.topics = Collections.singletonList("topic_speed");
@@ -1176,7 +1174,6 @@ public abstract class ConsumerLoop {
         config.setProperty("client.id", "1");
         config.setProperty("value.deserializer", "com.taosdata.jdbc.tmq.ConsumerTest.ConsumerLoop$ResultDeserializer");
         config.setProperty("value.deserializer.encoding", "UTF-8");
-        config.setProperty("experimental.snapshot.enable", "true");
 
         this.consumer = new TaosConsumer<>(config);
         this.topics = Collections.singletonList("topic_speed");

docs/en/14-reference/08-taos-shell.md

@@ -81,6 +81,14 @@ For example:
 taos -h h1.taos.com -s "use db; show tables;"
 ```
 
+## Export query results to a file
+
+- You can use ">>" to export the query results to a file, the syntax is like `select * from table >> file`. If there is only file name without path, the file will be generated under the current working directory of TDegnine CLI.
+
+## Import data from CSV file
+
+- You can use `insert into table_name file 'fileName'` to import the data from the specified file into the specified table. For example, `insert into d0 file '/root/d0.csv';` means importing the data in file "/root/d0.csv" into table "d0". If there is only file name without path, that means the file is located under current working directory of TDengine CLI. 
+
 ## TDengine CLI tips
 
 - You can use the up and down keys to iterate the history of commands entered
@@ -89,3 +97,5 @@ taos -h h1.taos.com -s "use db; show tables;"
 - Execute `RESET QUERY CACHE` to clear the local cache of the table schema
 - Execute SQL statements in batches. You can store a series of shell commands (ending with ;, one line for each SQL command) in a script file and execute the command `source <file-name>` in the TDengine CLI to execute all SQL commands in that file automatically
 - Enter `q` to exit TDengine CLI
+
+  

docs/examples/c/tmq_example.c

@@ -227,12 +227,6 @@ tmq_t* build_consumer() {
     return NULL;
   }
 
-  code = tmq_conf_set(conf, "experimental.snapshot.enable", "false");
-  if (TMQ_CONF_OK != code) {
-    tmq_conf_destroy(conf);
-    return NULL;
-  }
-
   tmq_conf_set_auto_commit_cb(conf, tmq_commit_cb_print, NULL);
 
   tmq_t* tmq = tmq_consumer_new(conf, NULL, 0);

docs/examples/go/sub/main.go

@@ -35,7 +35,6 @@ func main() {
 		"td.connect.port":              "6030",
 		"client.id":                    "test_tmq_client",
 		"enable.auto.commit":           "false",
-		"experimental.snapshot.enable": "true",
 		"msg.with.table.name":          "true",
 	})
 	if err != nil {

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

@@ -25,7 +25,20 @@ import CDemo from "./_sub_c.mdx";
 
 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。
 
-注意：数据订阅是从 WAL 消费数据，如果一些 WAL 文件被基于 WAL 保留策略删除，则已经删除的 WAL 文件中的数据就无法再消费到。需要根据业务需要在创建数据库时合理设置 `WAL_RETENTION_PERIOD` 或 `WAL_RETENTION_SIZE` ，并确保应用及时消费数据，这样才不会产生数据丢失的现象。数据订阅的行为与 Kafka 等广泛使用的消息队列类产品的行为相似。
+说明（以c接口为例）：
+1. 一个消费组消费同一个topic下的所有数据，不同消费组之间相互独立；
+2. 一个消费组消费同一个topic所有的vgroup，消费组可由多个消费者组成，但一个vgroup仅被一个消费者消费，如果消费者数量超过了vgroup数量，多余的消费者不消费数据；
+3. 在服务端每个vgroup仅保存一个offset，每个vgroup的offset是单调递增的，但不一定连续。各个vgroup的offset之间没有关联；
+4. 每次poll服务端会返回一个结果block，该block属于一个vgroup，可能包含多个wal版本的数据，可以通过 tmq_get_vgroup_offset 接口获得是该block第一条记录的offset；
+5. 一个消费组如果从未commit过offset，当其成员消费者重启重新拉取数据时，均从参数auto.offset.reset设定值开始消费；在一个消费者生命周期中，客户端本地记录了最近一次拉取数据的offset，不会拉取重复数据；
+6. 消费者如果异常终止（没有调用tmq_close），需等约12秒后触发其所属消费组rebalance，该消费者在服务端状态变为LOST，约1天后该消费者自动被删除；正常退出，退出后就会删除消费者；新增消费者，需等约2秒触发rebalance，该消费者在服务端状态变为ready；
+7. 消费组rebalance会对该组所有ready状态的消费者成员重新进行vgroup分配，消费者仅能对自己负责的vgroup进行assignment/seek/commit/poll操作；
+8. 消费者可利用 tmq_position 获得当前消费的offset，并seek到指定offset，重新消费；
+9. seek将position指向指定offset，不执行commit操作，一旦seek成功，可poll拉取指定offset及以后的数据；
+10. seek 操作之前须调用 tmq_get_topic_assignment 接口获取该consumer的vgroup ID和offset范围。seek 操作会检测vgroup ID 和 offset是否合法，如非法将报错；
+11. tmq_get_vgroup_offset接口获取的是记录所在结果block块里的第一条数据的offset，当seek至该offset时，将消费到这个block里的全部数据。参见第四点；
+12. 由于存在 WAL 过期删除机制，即使seek 操作成功，poll数据时有可能offset已失效。如果poll 的offset 小于 WAL 最小版本号，将会从WAL最小版本号消费；
+13. 数据订阅是从 WAL 消费数据，如果一些 WAL 文件被基于 WAL 保留策略删除，则已经删除的 WAL 文件中的数据就无法再消费到。需要根据业务需要在创建数据库时合理设置 `WAL_RETENTION_PERIOD` 或 `WAL_RETENTION_SIZE` ，并确保应用及时消费数据，这样才不会产生数据丢失的现象。数据订阅的行为与 Kafka 等广泛使用的消息队列类产品的行为相似；
 
 ## 主要数据结构和 API
 
@@ -35,39 +48,60 @@ import CDemo from "./_sub_c.mdx";
 <TabItem value="c" label="C">
 
 ```c
-typedef struct tmq_t      tmq_t;
-typedef struct tmq_conf_t tmq_conf_t;
-typedef struct tmq_list_t tmq_list_t;
-
-typedef void(tmq_commit_cb(tmq_t *, int32_t code, void *param));
-
-DLL_EXPORT tmq_list_t *tmq_list_new();
-DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
-DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
-DLL_EXPORT tmq_t *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
-DLL_EXPORT const char *tmq_err2str(int32_t code);
-
-DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
-DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
-DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
-DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
-DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg);
-DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
-
-enum tmq_conf_res_t {
-  TMQ_CONF_UNKNOWN = -2,
-  TMQ_CONF_INVALID = -1,
-  TMQ_CONF_OK = 0,
-};
-typedef enum tmq_conf_res_t tmq_conf_res_t;
-
-DLL_EXPORT tmq_conf_t    *tmq_conf_new();
-DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
-DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
-DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
-```
-
-这些 API 的文档请见 [C/C++ Connector](../../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），完整的示例代码请见下面 C 语言的示例代码。
+    typedef struct tmq_t      tmq_t;
+    typedef struct tmq_conf_t tmq_conf_t;
+    typedef struct tmq_list_t tmq_list_t;
+
+    typedef void(tmq_commit_cb(tmq_t *tmq, int32_t code, void *param));
+
+    typedef enum tmq_conf_res_t {
+    TMQ_CONF_UNKNOWN = -2,
+    TMQ_CONF_INVALID = -1,
+    TMQ_CONF_OK = 0,
+} tmq_conf_res_t;
+
+    typedef struct tmq_topic_assignment {
+    int32_t vgId;
+    int64_t currentOffset;
+    int64_t begin;
+    int64_t end;
+} tmq_topic_assignment;
+
+    DLL_EXPORT tmq_conf_t    *tmq_conf_new();
+    DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
+    DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
+    DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
+
+    DLL_EXPORT tmq_list_t *tmq_list_new();
+    DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
+    DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
+    DLL_EXPORT int32_t     tmq_list_get_size(const tmq_list_t *);
+    DLL_EXPORT char      **tmq_list_to_c_array(const tmq_list_t *);
+
+    DLL_EXPORT tmq_t    *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
+    DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
+    DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
+    DLL_EXPORT int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics);
+    DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
+    DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
+    DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg);
+    DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
+    DLL_EXPORT int32_t   tmq_commit_offset_sync(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
+    DLL_EXPORT void      tmq_commit_offset_async(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset, tmq_commit_cb *cb, void *param);
+    DLL_EXPORT int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment,int32_t *numOfAssignment);
+    DLL_EXPORT void      tmq_free_assignment(tmq_topic_assignment* pAssignment);
+    DLL_EXPORT int32_t   tmq_offset_seek(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
+    DLL_EXPORT int64_t   tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId);
+    DLL_EXPORT int64_t   tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId);
+
+    DLL_EXPORT const char *tmq_get_topic_name(TAOS_RES *res);
+    DLL_EXPORT const char *tmq_get_db_name(TAOS_RES *res);
+    DLL_EXPORT int32_t     tmq_get_vgroup_id(TAOS_RES *res);
+    DLL_EXPORT int64_t     tmq_get_vgroup_offset(TAOS_RES* res);
+    DLL_EXPORT const char *tmq_err2str(int32_t code);DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
+```
+
+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），完整的示例代码请见下面 C 语言的示例代码。
 
 </TabItem>
 <TabItem value="java" label="Java">

docs/zh/08-connector/10-cpp.mdx

@@ -467,21 +467,22 @@ TDengine 的异步 API 均采用非阻塞调用模式。应用程序可以用多
 - `TAOS_RES* taos_schemaless_insert(TAOS* taos, const char* lines[], int numLines, int protocol, int precision)`
 
   **功能说明**
-  该接口将行协议的文本数据写入到 TDengine 中。
+  - 该接口将行协议的文本数据写入到 TDengine 中。
 
   **参数说明**
-  taos: 数据库连接，通过 `taos_connect()` 函数建立的数据库连接。
-  lines：文本数据。满足解析格式要求的无模式文本字符串。
-  numLines:文本数据的行数，不能为 0 。
-  protocol: 行协议类型，用于标识文本数据格式。
-  precision：文本数据中的时间戳精度字符串。
+  - taos: 数据库连接，通过 `taos_connect()` 函数建立的数据库连接。
+  - lines：文本数据。满足解析格式要求的无模式文本字符串。
+  - numLines:文本数据的行数，不能为 0 。
+  - protocol: 行协议类型，用于标识文本数据格式。
+  - precision：文本数据中的时间戳精度字符串。
 
   **返回值**
-  TAOS_RES 结构体，应用可以通过使用 `taos_errstr()` 获得错误信息，也可以使用 `taos_errno()` 获得错误码。
+  - TAOS_RES 结构体，应用可以通过使用 `taos_errstr()` 获得错误信息，也可以使用 `taos_errno()` 获得错误码。
   在某些情况下，返回的 TAOS_RES 为 `NULL`，此时仍然可以调用 `taos_errno()` 来安全地获得错误码信息。
   返回的 TAOS_RES 需要调用方来负责释放，否则会出现内存泄漏。
 
   **说明**
+  
   协议类型是枚举类型，包含以下三种格式：
 
   - TSDB_SML_LINE_PROTOCOL：InfluxDB 行协议（Line Protocol)
@@ -515,3 +516,90 @@ TDengine 的异步 API 均采用非阻塞调用模式。应用程序可以用多
   - 带_raw的接口通过传递的参数lines指针和长度len来表示数据，为了解决原始接口数据包含'\0'而被截断的问题。totalRows指针返回解析出来的数据行数。
   - 带_ttl的接口可以传递ttl参数来控制建表的ttl到期时间。
   - 带_reqid的接口可以通过传递reqid参数来追踪整个的调用链。
+
+### 数据订阅 API
+
+- `int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment, int32_t *numOfAssignment)`
+- `void      tmq_free_assignment(tmq_topic_assignment* pAssignment)`
+  
+  tmq_topic_assignment结构体定义如下：
+  ```c
+  typedef struct tmq_topic_assignment {
+    int32_t vgId;
+    int64_t currentOffset;
+    int64_t begin;
+    int64_t end;
+  } tmq_topic_assignment;
+  ```
+  **功能说明**
+  - tmq_get_topic_assignment 接口返回当前consumer分配的vgroup的信息，每个vgroup的信息包括vgId，wal的最大最小offset，以及当前消费到的offset。
+
+  **参数说明**
+  - numOfAssignment ：分配给该consumer有效的vgroup个数。
+  - assignment ：分配的信息，数据大小为numOfAssignment，需要通过 tmq_free_assignment 接口释放。
+
+  **返回值**
+  - 错误码，0成功，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息。
+
+- `int64_t   tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId)`
+  **功能说明**
+  - 获取当前 consumer 在某个 topic 和 vgroup上的 commit 位置。
+
+  **返回值**
+  - 当前commit的位置，-2147467247表示没有消费进度，其他小于0的值表示失败，错误码就是返回值
+
+- `int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg)`
+- `void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param)`
+- `int32_t   tmq_commit_offset_sync(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset)`
+- `void      tmq_commit_offset_async(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset, tmq_commit_cb *cb, void *param)`
+  
+  **功能说明**
+  
+  commit接口分为两种类型，每种类型有同步和异步接口：
+  - 第一种类型：根据消息提交，提交消息里的进度，如果消息传NULL，提交当前consumer所有消费的vgroup的当前进度 : tmq_commit_sync/tmq_commit_async
+  - 第二种类型：根据某个topic的某个vgroup的offset提交 : tmq_commit_offset_sync/tmq_commit_offset_async
+  
+  **参数说明**
+  - msg：消费到的消息结构，如果msg传NULL，提交当前consumer所有消费的vgroup的当前进度
+
+  **返回值**
+  - 错误码，0成功，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息
+
+- `int64_t     tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId)`
+
+  **功能说明**
+  - 获取当前消费位置，为消费到的数据位置的下一个位置
+  
+  **返回值**
+  - 消费位置，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息
+
+- `int32_t   tmq_offset_seek(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset)`
+
+  **功能说明**
+  - 设置 consumer 在某个topic的某个vgroup的 offset位置，开始消费
+
+  **返回值**
+  - 错误码，0成功，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息
+  
+- `int32_t   int64_t     tmq_get_vgroup_offset(TAOS_RES* res)`
+
+  **功能说明**
+
+  获取 poll 消费到的数据的起始offset
+  
+  **参数说明**
+  - msg：消费到的消息结构
+
+  **返回值**
+  - 消费到的offset，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息
+  
+- `int32_t   int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics)`
+
+  **功能说明**
+
+  获取消费者订阅的 topic 列表
+  **参数说明**
+  - topics: 获取的 topic 列表存储在这个结构中，接口内分配内存，需调用tmq_list_destroy释放
+
+  **返回值**
+  - 错误码，0成功，非0失败，可通过 `char *tmq_err2str(int32_t code)` 函数获取错误信息
\ No newline at end of file

docs/zh/10-deployment/01-deploy.md

@@ -62,12 +62,13 @@ serverPort            6030
 
 加入到集群中的数据节点 dnode，下表中涉及集群相关的参数必须完全相同，否则不能成功加入到集群中。
 
-| **#** | **配置参数名称**   | **含义**                                    |
-| ----- | ------------------ | ------------------------------------------- |
-| 1     | statusInterval     | dnode 向 mnode 报告状态时长                 |
-| 2     | timezone           | 时区                                        |
-| 3     | locale             | 系统区位信息及编码格式                       |
-| 4     | charset            | 字符集编码                                 |
+| **#** | **配置参数名称** | **含义**                             |
+| ----- | ---------------- | ------------------------------------ |
+| 1     | statusInterval   | dnode 向 mnode 报告状态时长          |
+| 2     | timezone         | 时区                                 |
+| 3     | locale           | 系统区位信息及编码格式               |
+| 4     | charset          | 字符集编码                           |
+| 5     | ttlChangeOnWrite | ttl 到期时间是否伴随表的修改操作改变 |
 
 ## 启动集群
 
@@ -196,10 +197,10 @@ dnodeID 是集群自动分配的，不得人工指定。它在生成时是递增
 1、建立集群时使用 CREATE DNODE 增加新节点后，新节点始终显示 offline 状态？
 ```sql
   1）首先要检查增加的新节点上的 taosd 服务是否已经正常启动
-  
+
   2）如果已经启动，再检查到新节点的网络是否通畅，可以使用 ping fqdn 验证下
-  
+
   3）如果前面两步都没有问题，这一步要检查新节点做为独立集群在运行了，可以使用 taos -h fqdn 连接上后，show dnodes; 命令查看.
-    如果显示的列表与你主节点上显示的不一致，说明此节点自己单独成立了一个集群，解决的方法是停止新节点上的服务，然后清空新节点上 
+    如果显示的列表与你主节点上显示的不一致，说明此节点自己单独成立了一个集群，解决的方法是停止新节点上的服务，然后清空新节点上
     taos.cfg 中配置的 dataDir 目录下的所有文件，重新启动新节点服务即可解决。
-```  
+```

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

@@ -700,7 +700,7 @@ ELAPSED(ts_primary_key [, time_unit])
 LEASTSQUARES(expr, start_val, step_val)
 ```
 
-**功能说明**：统计表中某列的值是主键（时间戳）的拟合直线方程。start_val 是自变量初始值，step_val 是自变量的步长值。
+**功能说明**：统计表中某列的值的拟合直线方程。start_val 是自变量初始值，step_val 是自变量的步长值。
 
 **返回数据类型**：字符串表达式（斜率, 截距）。
 

docs/zh/12-taos-sql/25-grant.md

@@ -4,7 +4,7 @@ title: 权限管理
 description: 企业版中才具有的权限管理功能
 ---
 
-本节讲述如何在 TDengine 中进行权限管理的相关操作。
+本节讲述如何在 TDengine 中进行权限管理的相关操作。权限管理是 TDengine 企业版的特有功能，本节只列举了一些基本的权限管理功能作为示例，更丰富的权限管理请联系 TDengine 销售或市场团队。
 
 ## 创建用户
 

docs/zh/14-reference/08-taos-shell.md

@@ -89,3 +89,11 @@ taos -h h1.taos.com -s "use db; show tables;"
 - 执行 `RESET QUERY CACHE` 可清除本地表 Schema 的缓存
 - 批量执行 SQL 语句。可以将一系列的 TDengine CLI 命令（以英文 ; 结尾，每个 SQL 语句为一行）按行存放在文件里，在 TDengine CLI 里执行命令 `source <file-name>` 自动执行该文件里所有的 SQL 语句
 - 输入 `q` 或 `quit` 或 `exit` 回车，可以退出 TDengine CLI
+
+## TDengine CLI 导出查询结果到文件中
+
+- 可以使用符号 “>>” 导出查询结果到某个文件中，语法为： sql 查询语句 >> ‘输出文件名’; 输出文件如果不写路径的话，将输出至当前目录下。如 select * from d0 >> ‘/root/d0.csv’;  将把查询结果输出到 /root/d0.csv 中。
+
+## TDengine CLI 导入文件中的数据到表中
+
+- 可以使用 insert into table_name file '输入文件名'，把上一步中导出的数据文件再导入到指定表中。如 insert into d0 file '/root/d0.csv'; 表示把上面导出的数据全部再导致至 d0 表中。

examples/c/tmq.c

@@ -228,11 +228,6 @@ tmq_t* build_consumer() {
     tmq_conf_destroy(conf);
     return NULL;
   }
-  code = tmq_conf_set(conf, "experimental.snapshot.enable", "false");
-  if (TMQ_CONF_OK != code) {
-    tmq_conf_destroy(conf);
-    return NULL;
-  }
 
   tmq_conf_set_auto_commit_cb(conf, tmq_commit_cb_print, NULL);
   tmq = tmq_consumer_new(conf, NULL, 0);

include/client/taos.h

@@ -260,19 +260,14 @@ typedef struct tmq_t      tmq_t;
 typedef struct tmq_conf_t tmq_conf_t;
 typedef struct tmq_list_t tmq_list_t;
 
-typedef void(tmq_commit_cb(tmq_t *, int32_t code, void *param));
+typedef void(tmq_commit_cb(tmq_t *tmq, int32_t code, void *param));
 
-DLL_EXPORT tmq_list_t *tmq_list_new();
-DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
-DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
-DLL_EXPORT int32_t     tmq_list_get_size(const tmq_list_t *);
-DLL_EXPORT char      **tmq_list_to_c_array(const tmq_list_t *);
-
-DLL_EXPORT tmq_t *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
-
-DLL_EXPORT const char *tmq_err2str(int32_t code);
+typedef enum tmq_conf_res_t {
+  TMQ_CONF_UNKNOWN = -2,
+  TMQ_CONF_INVALID = -1,
+  TMQ_CONF_OK = 0,
+} tmq_conf_res_t;
 
-/* ------------------------TMQ CONSUMER INTERFACE------------------------ */
 typedef struct tmq_topic_assignment {
   int32_t vgId;
   int64_t currentOffset;
@@ -280,43 +275,38 @@ typedef struct tmq_topic_assignment {
   int64_t end;
 } tmq_topic_assignment;
 
+DLL_EXPORT tmq_conf_t    *tmq_conf_new();
+DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
+DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
+DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
+
+DLL_EXPORT tmq_list_t *tmq_list_new();
+DLL_EXPORT int32_t     tmq_list_append(tmq_list_t *, const char *);
+DLL_EXPORT void        tmq_list_destroy(tmq_list_t *);
+DLL_EXPORT int32_t     tmq_list_get_size(const tmq_list_t *);
+DLL_EXPORT char      **tmq_list_to_c_array(const tmq_list_t *);
+
+DLL_EXPORT tmq_t    *tmq_consumer_new(tmq_conf_t *conf, char *errstr, int32_t errstrLen);
 DLL_EXPORT int32_t   tmq_subscribe(tmq_t *tmq, const tmq_list_t *topic_list);
 DLL_EXPORT int32_t   tmq_unsubscribe(tmq_t *tmq);
 DLL_EXPORT int32_t   tmq_subscription(tmq_t *tmq, tmq_list_t **topics);
 DLL_EXPORT TAOS_RES *tmq_consumer_poll(tmq_t *tmq, int64_t timeout);
 DLL_EXPORT int32_t   tmq_consumer_close(tmq_t *tmq);
-DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg);
+DLL_EXPORT int32_t   tmq_commit_sync(tmq_t *tmq, const TAOS_RES *msg); //Commit the msg’s offset + 1
 DLL_EXPORT void      tmq_commit_async(tmq_t *tmq, const TAOS_RES *msg, tmq_commit_cb *cb, void *param);
 DLL_EXPORT int32_t   tmq_commit_offset_sync(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
 DLL_EXPORT void      tmq_commit_offset_async(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset, tmq_commit_cb *cb, void *param);
-DLL_EXPORT int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment,
-                                              int32_t *numOfAssignment);
+DLL_EXPORT int32_t   tmq_get_topic_assignment(tmq_t *tmq, const char *pTopicName, tmq_topic_assignment **assignment,int32_t *numOfAssignment);
 DLL_EXPORT void      tmq_free_assignment(tmq_topic_assignment* pAssignment);
 DLL_EXPORT int32_t   tmq_offset_seek(tmq_t *tmq, const char *pTopicName, int32_t vgId, int64_t offset);
+DLL_EXPORT int64_t   tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId);  // The current offset is the offset of the last consumed message + 1
+DLL_EXPORT int64_t   tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId);
 
 DLL_EXPORT const char *tmq_get_topic_name(TAOS_RES *res);
 DLL_EXPORT const char *tmq_get_db_name(TAOS_RES *res);
 DLL_EXPORT int32_t     tmq_get_vgroup_id(TAOS_RES *res);
 DLL_EXPORT int64_t     tmq_get_vgroup_offset(TAOS_RES* res);
-DLL_EXPORT int64_t     tmq_position(tmq_t *tmq, const char *pTopicName, int32_t vgId);
-DLL_EXPORT int64_t     tmq_committed(tmq_t *tmq, const char *pTopicName, int32_t vgId);
-
-/* ----------------------TMQ CONFIGURATION INTERFACE---------------------- */
-
-enum tmq_conf_res_t {
-  TMQ_CONF_UNKNOWN = -2,
-  TMQ_CONF_INVALID = -1,
-  TMQ_CONF_OK = 0,
-};
-
-typedef enum tmq_conf_res_t tmq_conf_res_t;
-
-DLL_EXPORT tmq_conf_t    *tmq_conf_new();
-DLL_EXPORT tmq_conf_res_t tmq_conf_set(tmq_conf_t *conf, const char *key, const char *value);
-DLL_EXPORT void           tmq_conf_destroy(tmq_conf_t *conf);
-DLL_EXPORT void           tmq_conf_set_auto_commit_cb(tmq_conf_t *conf, tmq_commit_cb *cb, void *param);
-
-/* -------------------------TMQ MSG HANDLE INTERFACE---------------------- */
+DLL_EXPORT const char *tmq_err2str(int32_t code);
 
 /* ------------------------------ TAOSX -----------------------------------*/
 // note: following apis are unstable

include/common/tcommon.h

@@ -152,6 +152,8 @@ enum {
   STREAM_INPUT__DATA_RETRIEVE,
   STREAM_INPUT__GET_RES,
   STREAM_INPUT__CHECKPOINT,
+  STREAM_INPUT__CHECKPOINT_TRIGGER,
+  STREAM_INPUT__TRANS_STATE,
   STREAM_INPUT__REF_DATA_BLOCK,
   STREAM_INPUT__DESTROY,
 };
@@ -168,7 +170,9 @@ typedef enum EStreamType {
   STREAM_PULL_DATA,
   STREAM_PULL_OVER,
   STREAM_FILL_OVER,
+  STREAM_CHECKPOINT,
   STREAM_CREATE_CHILD_TABLE,
+  STREAM_TRANS_STATE,
 } EStreamType;
 
 #pragma pack(push, 1)

include/common/tglobal.h

@@ -199,6 +199,7 @@ extern bool    tsFilterScalarMode;
 extern int32_t tsKeepTimeOffset;
 extern int32_t tsMaxStreamBackendCache;
 extern int32_t tsPQSortMemThreshold;
+extern int32_t tsResolveFQDNRetryTime;
 
 // #define NEEDTO_COMPRESSS_MSG(size) (tsCompressMsgSize != -1 && (size) > tsCompressMsgSize)
 

include/common/tmsgdef.h

@@ -255,7 +255,6 @@ enum {
   TD_DEF_MSG_TYPE(TDMT_STREAM_RETRIEVE, "stream-retrieve", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_STREAM_SCAN_HISTORY, "stream-scan-history", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_STREAM_SCAN_HISTORY_FINISH, "stream-scan-history-finish", NULL, NULL)
-  TD_DEF_MSG_TYPE(TDMT_STREAM_TRANSFER_STATE, "stream-transfer-state", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_STREAM_TASK_CHECK, "stream-task-check", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_STREAM_TASK_CHECKPOINT, "stream-checkpoint", NULL, NULL)
   TD_DEF_MSG_TYPE(TDMT_STREAM_TASK_REPORT_CHECKPOINT, "stream-report-checkpoint", NULL, NULL)

include/libs/executor/storageapi.h

@@ -98,6 +98,16 @@ typedef struct SMTbCursor {
   int8_t      paused;
 } SMTbCursor;
 
+typedef struct SMCtbCursor {
+  SMeta   *pMeta;
+  void    *pCur;
+  tb_uid_t suid;
+  void    *pKey;
+  void    *pVal;
+  int      kLen;
+  int      vLen;
+} SMCtbCursor;
+
 typedef struct SRowBuffPos {
   void* pRowBuff;
   void* pKey;
@@ -278,13 +288,15 @@ typedef struct SStoreMeta {
   void (*getBasicInfo)(void* pVnode, const char** dbname, int32_t* vgId, int64_t* numOfTables,
                        int64_t* numOfNormalTables);  // vnodeGetInfo(void *pVnode, const char **dbname, int32_t *vgId) &
                                                      // metaGetTbNum(SMeta *pMeta) & metaGetNtbNum(SMeta *pMeta);
-
   int64_t (*getNumOfRowsInMem)(void* pVnode);
   /**
 int32_t vnodeGetCtbIdList(void *pVnode, int64_t suid, SArray *list);
 int32_t vnodeGetCtbIdListByFilter(void *pVnode, int64_t suid, SArray *list, bool (*filter)(void *arg), void *arg);
 int32_t vnodeGetStbIdList(void *pVnode, int64_t suid, SArray *list);
  */
+  SMCtbCursor* (*openCtbCursor)(void *pVnode, tb_uid_t uid, int lock);
+  void         (*closeCtbCursor)(SMCtbCursor *pCtbCur, int lock);
+  tb_uid_t     (*ctbCursorNext)(SMCtbCursor* pCur);
 } SStoreMeta;
 
 typedef struct SStoreMetaReader {

include/libs/function/functionMgt.h

@@ -157,6 +157,8 @@ typedef enum EFunctionType {
   FUNCTION_TYPE_AVG_MERGE,
   FUNCTION_TYPE_STDDEV_PARTIAL,
   FUNCTION_TYPE_STDDEV_MERGE,
+  FUNCTION_TYPE_IRATE_PARTIAL,
+  FUNCTION_TYPE_IRATE_MERGE,
 
   // geometry functions
   FUNCTION_TYPE_GEOM_FROM_TEXT = 4250,

include/libs/nodes/nodes.h

@@ -169,7 +169,7 @@ typedef enum ENodeType {
   QUERY_NODE_REVOKE_STMT,
   QUERY_NODE_SHOW_DNODES_STMT,
   QUERY_NODE_SHOW_MNODES_STMT,
-  QUERY_NODE_SHOW_MODULES_STMT,
+//  QUERY_NODE_SHOW_MODULES_STMT,
   QUERY_NODE_SHOW_QNODES_STMT,
   QUERY_NODE_SHOW_SNODES_STMT,
   QUERY_NODE_SHOW_BNODES_STMT,

include/libs/nodes/plannodes.h

@@ -107,6 +107,7 @@ typedef struct SScanLogicNode {
   bool          sortPrimaryKey;
   bool          igLastNull;
   bool          groupOrderScan;
+  bool          onlyMetaCtbIdx; // for tag scan with no tbname
 } SScanLogicNode;
 
 typedef struct SJoinLogicNode {
@@ -334,7 +335,11 @@ typedef struct SScanPhysiNode {
   bool       groupOrderScan;
 } SScanPhysiNode;
 
-typedef SScanPhysiNode STagScanPhysiNode;
+typedef struct STagScanPhysiNode {
+  SScanPhysiNode scan;
+  bool       onlyMetaCtbIdx; //no tbname, tag index not used.
+} STagScanPhysiNode;
+
 typedef SScanPhysiNode SBlockDistScanPhysiNode;
 
 typedef struct SLastRowScanPhysiNode {
@@ -603,6 +608,8 @@ typedef struct SSubplan {
   SNode*         pTagCond;
   SNode*         pTagIndexCond;
   bool           showRewrite;
+  int32_t        rowsThreshold;
+  bool           dynamicRowThreshold;
 } SSubplan;
 
 typedef enum EExplainMode { EXPLAIN_MODE_DISABLE = 1, EXPLAIN_MODE_STATIC, EXPLAIN_MODE_ANALYZE } EExplainMode;

include/libs/stream/tstream.h

@@ -122,6 +122,7 @@ typedef struct {
   int8_t type;
 
   int32_t srcVgId;
+  int32_t srcTaskId;
   int32_t childId;
   int64_t sourceVer;
   int64_t reqId;
@@ -251,6 +252,7 @@ typedef struct SStreamChildEpInfo {
   int32_t nodeId;
   int32_t childId;
   int32_t taskId;
+  int8_t  dataAllowed;
   SEpSet  epSet;
 } SStreamChildEpInfo;
 
@@ -272,6 +274,7 @@ typedef struct SStreamStatus {
   int8_t schedStatus;
   int8_t keepTaskStatus;
   bool   transferState;
+  bool   appendTranstateBlock;  // has append the transfer state data block already, todo: remove it
   int8_t timerActive;   // timer is active
   int8_t pauseAllowed;  // allowed task status to be set to be paused
 } SStreamStatus;
@@ -399,8 +402,9 @@ typedef struct {
 
 typedef struct {
   int64_t streamId;
+  int32_t type;
   int32_t taskId;
-  int32_t dataSrcVgId;
+  int32_t srcVgId;
   int32_t upstreamTaskId;
   int32_t upstreamChildId;
   int32_t upstreamNodeId;
@@ -570,8 +574,6 @@ int32_t tDecodeStreamDispatchReq(SDecoder* pDecoder, SStreamDispatchReq* pReq);
 int32_t tDecodeStreamRetrieveReq(SDecoder* pDecoder, SStreamRetrieveReq* pReq);
 void    tDeleteStreamRetrieveReq(SStreamRetrieveReq* pReq);
 
-int32_t tInitStreamDispatchReq(SStreamDispatchReq* pReq, const SStreamTask* pTask, int32_t vgId, int32_t numOfBlocks,
-                               int64_t dstTaskId);
 void    tDeleteStreamDispatchReq(SStreamDispatchReq* pReq);
 
 int32_t streamSetupScheduleTrigger(SStreamTask* pTask);
@@ -579,6 +581,8 @@ int32_t streamSetupScheduleTrigger(SStreamTask* pTask);
 int32_t streamProcessRunReq(SStreamTask* pTask);
 int32_t streamProcessDispatchMsg(SStreamTask* pTask, SStreamDispatchReq* pReq, SRpcMsg* pMsg, bool exec);
 int32_t streamProcessDispatchRsp(SStreamTask* pTask, SStreamDispatchRsp* pRsp, int32_t code);
+void streamTaskCloseUpstreamInput(SStreamTask* pTask, int32_t taskId);
+void streamTaskOpenAllUpstreamInput(SStreamTask* pTask);
 
 int32_t streamProcessRetrieveReq(SStreamTask* pTask, SStreamRetrieveReq* pReq, SRpcMsg* pMsg);
 
@@ -589,7 +593,6 @@ int32_t streamTaskOutputResultBlock(SStreamTask* pTask, SStreamDataBlock* pBlock
 bool    streamTaskShouldStop(const SStreamStatus* pStatus);
 bool    streamTaskShouldPause(const SStreamStatus* pStatus);
 bool    streamTaskIsIdle(const SStreamTask* pTask);
-int32_t streamTaskEndScanWAL(SStreamTask* pTask);
 
 SStreamChildEpInfo* streamTaskGetUpstreamTaskEpInfo(SStreamTask* pTask, int32_t taskId);
 int32_t             streamScanExec(SStreamTask* pTask, int32_t batchSize);
@@ -626,7 +629,7 @@ int32_t streamSetParamForStreamScannerStep2(SStreamTask* pTask, SVersionRange* p
 int32_t streamSourceScanHistoryData(SStreamTask* pTask);
 int32_t streamDispatchScanHistoryFinishMsg(SStreamTask* pTask);
 
-int32_t streamDispatchTransferStateMsg(SStreamTask* pTask);
+int32_t appendTranstateIntoInputQ(SStreamTask* pTask);
 
 // agg level
 int32_t streamTaskScanHistoryPrepare(SStreamTask* pTask);

include/libs/wal/wal.h

@@ -153,7 +153,6 @@ struct SWalReader {
   int64_t        capacity;
   TdThreadMutex  mutex;
   SWalFilterCond cond;
-  // TODO remove it
   SWalCkHead *pHead;
 };
 
@@ -207,10 +206,9 @@ void        walReaderValidVersionRange(SWalReader *pReader, int64_t *sver, int64
 void        walReaderVerifyOffset(SWalReader *pWalReader, STqOffsetVal* pOffset);
 
 // only for tq usage
-void    walSetReaderCapacity(SWalReader *pRead, int32_t capacity);
-int32_t walFetchHead(SWalReader *pRead, int64_t ver, SWalCkHead *pHead);
-int32_t walFetchBody(SWalReader *pRead, SWalCkHead **ppHead);
-int32_t walSkipFetchBody(SWalReader *pRead, const SWalCkHead *pHead);
+int32_t walFetchHead(SWalReader *pRead, int64_t ver);
+int32_t walFetchBody(SWalReader *pRead);
+int32_t walSkipFetchBody(SWalReader *pRead);
 
 void walRefFirstVer(SWal *, SWalRef *);
 void walRefLastVer(SWal *, SWalRef *);

include/os/osFile.h

@@ -76,7 +76,7 @@ int32_t taosUnLockFile(TdFilePtr pFile);
 
 int32_t taosUmaskFile(int32_t maskVal);
 
-int32_t taosStatFile(const char *path, int64_t *size, int32_t *mtime);
+int32_t taosStatFile(const char *path, int64_t *size, int32_t *mtime, int32_t *atime);
 int32_t taosDevInoFile(TdFilePtr pFile, int64_t *stDev, int64_t *stIno);
 int32_t taosFStatFile(TdFilePtr pFile, int64_t *size, int32_t *mtime);
 bool    taosCheckExistFile(const char *pathname);

include/os/osRand.h

@@ -32,6 +32,8 @@ void     taosSeedRand(uint32_t seed);
 uint32_t taosRand(void);
 uint32_t taosRandR(uint32_t* pSeed);
 void     taosRandStr(char* str, int32_t size);
+void     taosRandStr2(char* str, int32_t size);
+
 uint32_t taosSafeRand(void);
 
 #ifdef __cplusplus

include/util/taoserror.h

@@ -191,6 +191,7 @@ int32_t* taosGetErrno();
 // #define TSDB_CODE_MND_FAILED_TO_CREATE_DIR   TAOS_DEF_ERROR_CODE(0, 0x0313) // 2.x
 // #define TSDB_CODE_MND_FAILED_TO_INIT_STEP    TAOS_DEF_ERROR_CODE(0, 0x0314) // 2.x
 #define TSDB_CODE_MND_USER_DISABLED             TAOS_DEF_ERROR_CODE(0, 0x0315)
+#define TSDB_CODE_MND_INVALID_PLATFORM          TAOS_DEF_ERROR_CODE(0, 0x0316)
 
 // mnode-sdb
 #define TSDB_CODE_SDB_OBJ_ALREADY_THERE         TAOS_DEF_ERROR_CODE(0, 0x0320) // internal

source/client/CMakeLists.txt

@@ -16,7 +16,7 @@ target_include_directories(
 target_link_libraries(
     taos
     INTERFACE api
-    PRIVATE os util common transport nodes parser command planner catalog scheduler function qcom
+    PRIVATE os util common transport nodes parser command planner catalog scheduler function qcom geometry
 )
 
 if(TD_DARWIN_ARM64)
@@ -57,7 +57,7 @@ target_include_directories(
 target_link_libraries(
     taos_static
     INTERFACE api
-    PRIVATE os util common transport nodes parser command planner catalog scheduler function qcom
+    PRIVATE os util common transport nodes parser command planner catalog scheduler function qcom geometry
 )
 
 if(${BUILD_TEST})

source/client/inc/clientSml.h

@@ -33,6 +33,7 @@ extern "C" {
 #include "ttime.h"
 #include "ttypes.h"
 #include "cJSON.h"
+#include "geosWrapper.h"
 
 #if (defined(__GNUC__) && (__GNUC__ >= 3)) || (defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 800)) || defined(__clang__)
 #  define expect(expr,value)    (__builtin_expect ((expr),(value)) )
@@ -192,7 +193,7 @@ typedef struct {
   //
   SArray      *preLineTagKV;
   SArray      *maxTagKVs;
-  SArray      *masColKVs;
+  SArray      *maxColKVs;
 
   SSmlLineInfo preLine;
   STableMeta  *currSTableMeta;

source/client/src/clientSml.c

@@ -1073,6 +1073,7 @@ static int32_t smlModifyDBSchemas(SSmlHandle *info) {
   return 0;
 
 end:
+  taosHashCancelIterate(info->superTables, tmp);
   taosHashCleanup(hashTmp);
   taosMemoryFreeClear(pTableMeta);
   catalogRefreshTableMeta(info->pCatalog, &conn, &pName, 1);
@@ -1191,6 +1192,7 @@ void freeSSmlKv(void *data) {
   SSmlKv *kv = (SSmlKv *)data;
   if (kv->keyEscaped) taosMemoryFree((void *)(kv->key));
   if (kv->valueEscaped) taosMemoryFree((void *)(kv->value));
+  if (kv->type == TSDB_DATA_TYPE_GEOMETRY) geosFreeBuffer((void *)(kv->value));
 }
 
 void smlDestroyInfo(SSmlHandle *info) {
@@ -1433,6 +1435,7 @@ static int32_t smlInsertData(SSmlHandle *info) {
     code = smlCheckAuth(info, &conn, pName.tname, AUTH_TYPE_WRITE);
     if(code != TSDB_CODE_SUCCESS){
       taosMemoryFree(measure);
+      taosHashCancelIterate(info->childTables, oneTable);
       return code;
     }
 
@@ -1441,6 +1444,7 @@ static int32_t smlInsertData(SSmlHandle *info) {
     if (code != TSDB_CODE_SUCCESS) {
       uError("SML:0x%" PRIx64 " catalogGetTableHashVgroup failed. table name: %s", info->id, tableData->childTableName);
       taosMemoryFree(measure);
+      taosHashCancelIterate(info->childTables, oneTable);
       return code;
     }
     taosHashPut(info->pVgHash, (const char *)&vg.vgId, sizeof(vg.vgId), (char *)&vg, sizeof(vg));
@@ -1450,6 +1454,7 @@ static int32_t smlInsertData(SSmlHandle *info) {
     if (unlikely(NULL == pMeta || NULL == (*pMeta)->tableMeta)) {
       uError("SML:0x%" PRIx64 " NULL == pMeta. table name: %s", info->id, tableData->childTableName);
       taosMemoryFree(measure);
+      taosHashCancelIterate(info->childTables, oneTable);
       return TSDB_CODE_SML_INTERNAL_ERROR;
     }
 
@@ -1465,6 +1470,7 @@ static int32_t smlInsertData(SSmlHandle *info) {
     taosMemoryFree(measure);
     if (code != TSDB_CODE_SUCCESS) {
       uError("SML:0x%" PRIx64 " smlBindData failed", info->id);
+      taosHashCancelIterate(info->childTables, oneTable);
       return code;
     }
     oneTable = (SSmlTableInfo **)taosHashIterate(info->childTables, oneTable);

source/client/src/clientSmlLine.c

@@ -102,6 +102,30 @@ int32_t smlParseValue(SSmlKv *pVal, SSmlMsgBuf *msg) {
     return TSDB_CODE_TSC_INVALID_VALUE;
   }
 
+  if (pVal->value[0] == 'g' || pVal->value[0] == 'G') {  // geometry
+    if (pVal->value[1] == '"' && pVal->value[pVal->length - 1] == '"' && pVal->length >= sizeof("POINT")+3) {
+      int32_t code = initCtxGeomFromText();
+      if (code != TSDB_CODE_SUCCESS) {
+        return code;
+      }
+      char* tmp = taosMemoryCalloc(pVal->length, 1);
+      memcpy(tmp, pVal->value + 2, pVal->length - 3);
+      code = doGeomFromText(tmp, (unsigned char **)&pVal->value, &pVal->length);
+      taosMemoryFree(tmp);
+      if (code != TSDB_CODE_SUCCESS) {
+        return code;
+      }
+
+      pVal->type = TSDB_DATA_TYPE_GEOMETRY;
+      if (pVal->length > TSDB_MAX_BINARY_LEN - VARSTR_HEADER_SIZE) {
+        geosFreeBuffer((void*)(pVal->value));
+        return TSDB_CODE_PAR_INVALID_VAR_COLUMN_LEN;
+      }
+      return TSDB_CODE_SUCCESS;
+    }
+    return TSDB_CODE_TSC_INVALID_VALUE;
+  }
+
   if (pVal->value[0] == 't' || pVal->value[0] == 'T') {
     if (pVal->length == 1 ||
         (pVal->length == 4 && (pVal->value[1] == 'r' || pVal->value[1] == 'R') &&
@@ -390,14 +414,14 @@ static int32_t smlParseColKv(SSmlHandle *info, char **sql, char *sqlEnd, SSmlLin
           SSmlKv   kv = {.key = tag->name, .keyLen = strlen(tag->name), .type = tag->type};
           if (tag->type == TSDB_DATA_TYPE_NCHAR) {
             kv.length = (tag->bytes - VARSTR_HEADER_SIZE) / TSDB_NCHAR_SIZE;
-          } else if (tag->type == TSDB_DATA_TYPE_BINARY) {
+          } else if (tag->type == TSDB_DATA_TYPE_BINARY || tag->type == TSDB_DATA_TYPE_GEOMETRY) {
             kv.length = tag->bytes - VARSTR_HEADER_SIZE;
           }
           taosArrayPush((*tmp)->cols, &kv);
         }
       }
       info->currSTableMeta = (*tmp)->tableMeta;
-      info->masColKVs = (*tmp)->cols;
+      info->maxColKVs = (*tmp)->cols;
     }
   }
 
@@ -512,13 +536,13 @@ static int32_t smlParseColKv(SSmlHandle *info, char **sql, char *sqlEnd, SSmlLin
         freeSSmlKv(&kv);
         return TSDB_CODE_SUCCESS;
       }
-      if (cnt >= taosArrayGetSize(info->masColKVs)) {
+      if (cnt >= taosArrayGetSize(info->maxColKVs)) {
         info->dataFormat = false;
         info->reRun = true;
         freeSSmlKv(&kv);
         return TSDB_CODE_SUCCESS;
       }
-      SSmlKv *maxKV = (SSmlKv *)taosArrayGet(info->masColKVs, cnt);
+      SSmlKv *maxKV = (SSmlKv *)taosArrayGet(info->maxColKVs, cnt);
       if (kv.type != maxKV->type) {
         info->dataFormat = false;
         info->reRun = true;
@@ -663,14 +687,15 @@ int32_t smlParseInfluxString(SSmlHandle *info, char *sql, char *sqlEnd, SSmlLine
   if (info->dataFormat) {
     uDebug("SML:0x%" PRIx64 " smlParseInfluxString format true, ts:%" PRId64, info->id, ts);
     ret = smlBuildCol(info->currTableDataCtx, info->currSTableMeta->schema, &kv, 0);
-    if (ret != TSDB_CODE_SUCCESS) {
-      return ret;
+    if (ret == TSDB_CODE_SUCCESS) {
+      ret = smlBuildRow(info->currTableDataCtx);
     }
-    ret = smlBuildRow(info->currTableDataCtx);
-    if (ret != TSDB_CODE_SUCCESS) {
+
+    clearColValArray(info->currTableDataCtx->pValues);
+    if (unlikely(ret != TSDB_CODE_SUCCESS)) {
+      smlBuildInvalidDataMsg(&info->msgBuf, "smlBuildCol error", NULL);
       return ret;
     }
-    clearColValArray(info->currTableDataCtx->pValues);
   } else {
     uDebug("SML:0x%" PRIx64 " smlParseInfluxString format false, ts:%" PRId64, info->id, ts);
     taosArraySet(elements->colArray, 0, &kv);

source/client/src/clientTmq.c

@@ -1863,10 +1863,10 @@ static int32_t tmqHandleNoPollRsp(tmq_t* tmq, SMqRspWrapper* rspWrapper, bool* p
   return 0;
 }
 
-static void updateVgInfo(SMqClientVg* pVg, STqOffsetVal* reqOffset, STqOffsetVal* rspOffset, int64_t sver, int64_t ever, int64_t consumerId){
+static void updateVgInfo(SMqClientVg* pVg, STqOffsetVal* reqOffset, STqOffsetVal* rspOffset, int64_t sver, int64_t ever, int64_t consumerId, bool hasData){
   if (!pVg->seekUpdated) {
     tscDebug("consumer:0x%" PRIx64" local offset is update, since seekupdate not set", consumerId);
-    pVg->offsetInfo.beginOffset = *reqOffset;
+    if(hasData) pVg->offsetInfo.beginOffset = *reqOffset;
     pVg->offsetInfo.endOffset = *rspOffset;
   } else {
     tscDebug("consumer:0x%" PRIx64" local offset is NOT update, since seekupdate is set", consumerId);
@@ -1929,7 +1929,7 @@ static void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) {
           pVg->epSet = *pollRspWrapper->pEpset;
         }
 
-        updateVgInfo(pVg, &pDataRsp->reqOffset, &pDataRsp->rspOffset, pDataRsp->head.walsver, pDataRsp->head.walever, tmq->consumerId);
+        updateVgInfo(pVg, &pDataRsp->reqOffset, &pDataRsp->rspOffset, pDataRsp->head.walsver, pDataRsp->head.walever, tmq->consumerId, pDataRsp->blockNum != 0);
 
         char buf[TSDB_OFFSET_LEN] = {0};
         tFormatOffset(buf, TSDB_OFFSET_LEN, &pDataRsp->rspOffset);
@@ -1979,7 +1979,7 @@ static void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) {
           return NULL;
         }
 
-        updateVgInfo(pVg, &pollRspWrapper->metaRsp.rspOffset, &pollRspWrapper->metaRsp.rspOffset, pollRspWrapper->metaRsp.head.walsver, pollRspWrapper->metaRsp.head.walever, tmq->consumerId);
+        updateVgInfo(pVg, &pollRspWrapper->metaRsp.rspOffset, &pollRspWrapper->metaRsp.rspOffset, pollRspWrapper->metaRsp.head.walsver, pollRspWrapper->metaRsp.head.walever, tmq->consumerId, true);
         // build rsp
         SMqMetaRspObj* pRsp = tmqBuildMetaRspFromWrapper(pollRspWrapper);
         taosFreeQitem(pollRspWrapper);
@@ -2007,7 +2007,7 @@ static void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) {
           return NULL;
         }
 
-        updateVgInfo(pVg, &pollRspWrapper->taosxRsp.reqOffset, &pollRspWrapper->taosxRsp.rspOffset, pollRspWrapper->taosxRsp.head.walsver, pollRspWrapper->taosxRsp.head.walever, tmq->consumerId);
+        updateVgInfo(pVg, &pollRspWrapper->taosxRsp.reqOffset, &pollRspWrapper->taosxRsp.rspOffset, pollRspWrapper->taosxRsp.head.walsver, pollRspWrapper->taosxRsp.head.walever, tmq->consumerId, pollRspWrapper->taosxRsp.blockNum != 0);
 
         if (pollRspWrapper->taosxRsp.blockNum == 0) {
           tscDebug("consumer:0x%" PRIx64 " taosx empty block received, vgId:%d, vg total:%" PRId64 ", reqId:0x%" PRIx64,

source/client/test/CMakeLists.txt

@@ -20,7 +20,7 @@ TARGET_LINK_LIBRARIES(
 ADD_EXECUTABLE(smlTest smlTest.cpp)
 TARGET_LINK_LIBRARIES(
         smlTest
-        PUBLIC os util common transport parser catalog scheduler function gtest taos_static qcom
+        PUBLIC os util common transport parser catalog scheduler function gtest taos_static qcom geometry
 )
 
 TARGET_INCLUDE_DIRECTORIES(

source/client/test/clientTests.cpp

@@ -1442,178 +1442,4 @@ TEST(clientCase, sub_tb_mt_test) {
   }
 }
 
-TEST(clientCase, ts_3756) {
-//  taos_options(TSDB_OPTION_CONFIGDIR, "~/first/cfg");
-
-  TAOS* pConn = taos_connect("localhost", "root", "taosdata", NULL, 0);
-  ASSERT_NE(pConn, nullptr);
-
-  tmq_conf_t* conf = tmq_conf_new();
-
-  tmq_conf_set(conf, "enable.auto.commit", "false");
-  tmq_conf_set(conf, "auto.commit.interval.ms", "2000");
-  tmq_conf_set(conf, "group.id", "group_id_2");
-  tmq_conf_set(conf, "td.connect.user", "root");
-  tmq_conf_set(conf, "td.connect.pass", "taosdata");
-  tmq_conf_set(conf, "auto.offset.reset", "latest");
-  tmq_conf_set(conf, "msg.with.table.name", "false");
-
-  tmq_t* tmq = tmq_consumer_new(conf, NULL, 0);
-  tmq_conf_destroy(conf);
-
-  // 创建订阅 topics 列表
-  tmq_list_t* topicList = tmq_list_new();
-  tmq_list_append(topicList, "tp");
-
-  // 启动订阅
-  tmq_subscribe(tmq, topicList);
-  tmq_list_destroy(topicList);
-
-  TAOS_FIELD* fields = NULL;
-  int32_t     numOfFields = 0;
-  int32_t     precision = 0;
-  int32_t     totalRows = 0;
-  int32_t     msgCnt = 0;
-  int32_t     timeout = 200;
-
-  int32_t count = 0;
-
-  tmq_topic_assignment* pAssign = NULL;
-  int32_t numOfAssign = 0;
-
-  int32_t code = tmq_get_topic_assignment(tmq, "tp", &pAssign, &numOfAssign);
-  if (code != 0) {
-    printf("error occurs:%s\n", tmq_err2str(code));
-    tmq_free_assignment(pAssign);
-    tmq_consumer_close(tmq);
-    taos_close(pConn);
-    fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-    return;
-  }
-
-  for(int i = 0; i < numOfAssign; i++){
-    printf("assign i:%d, vgId:%d, offset:%lld, start:%lld, end:%lld\n", i, pAssign[i].vgId, pAssign[i].currentOffset, pAssign[i].begin, pAssign[i].end);
-  }
-
-//  tmq_offset_seek(tmq, "tp", pAssign[0].vgId, 4);
-  tmq_free_assignment(pAssign);
-
-  code = tmq_get_topic_assignment(tmq, "tp", &pAssign, &numOfAssign);
-  if (code != 0) {
-    printf("error occurs:%s\n", tmq_err2str(code));
-    tmq_free_assignment(pAssign);
-    tmq_consumer_close(tmq);
-    taos_close(pConn);
-    fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-    return;
-  }
-
-  for(int i = 0; i < numOfAssign; i++){
-    printf("assign i:%d, vgId:%d, offset:%lld, start:%lld, end:%lld\n", i, pAssign[i].vgId, pAssign[i].currentOffset, pAssign[i].begin, pAssign[i].end);
-  }
-
-  tmq_free_assignment(pAssign);
-
-  code = tmq_get_topic_assignment(tmq, "tp", &pAssign, &numOfAssign);
-  if (code != 0) {
-    printf("error occurs:%s\n", tmq_err2str(code));
-    tmq_free_assignment(pAssign);
-    tmq_consumer_close(tmq);
-    taos_close(pConn);
-    fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-    return;
-  }
-
-  for(int i = 0; i < numOfAssign; i++){
-    printf("assign i:%d, vgId:%d, offset:%lld, start:%lld, end:%lld\n", i, pAssign[i].vgId, pAssign[i].currentOffset, pAssign[i].begin, pAssign[i].end);
-  }
-
-  while (1) {
-    printf("start to poll\n");
-    TAOS_RES* pRes = tmq_consumer_poll(tmq, timeout);
-    if (pRes) {
-      char buf[128];
-
-      const char* topicName = tmq_get_topic_name(pRes);
-//      const char* dbName = tmq_get_db_name(pRes);
-//      int32_t     vgroupId = tmq_get_vgroup_id(pRes);
-//
-//      printf("topic: %s\n", topicName);
-//      printf("db: %s\n", dbName);
-//      printf("vgroup id: %d\n", vgroupId);
-
-      printSubResults(pRes, &totalRows);
-
-      tmq_topic_assignment* pAssignTmp = NULL;
-      int32_t numOfAssignTmp = 0;
-
-      code = tmq_get_topic_assignment(tmq, "tp", &pAssignTmp, &numOfAssignTmp);
-      if (code != 0) {
-        printf("error occurs:%s\n", tmq_err2str(code));
-        tmq_free_assignment(pAssign);
-        tmq_consumer_close(tmq);
-        taos_close(pConn);
-        fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-        return;
-      }
-
-      for(int i = 0; i < numOfAssign; i++){
-        printf("assign i:%d, vgId:%d, offset:%lld, start:%lld, end:%lld\n", i, pAssignTmp[i].vgId, pAssignTmp[i].currentOffset, pAssignTmp[i].begin, pAssignTmp[i].end);
-      }
-      if(numOfAssign != 0){
-        int i = 0;
-        for(; i < numOfAssign; i++){
-          if(pAssign[i].currentOffset != pAssignTmp[i].currentOffset){
-            break;
-          }
-        }
-        if(i == numOfAssign){
-          printf("all position is same\n");
-          break;
-        }
-        tmq_free_assignment(pAssign);
-      }
-      numOfAssign = numOfAssignTmp;
-      pAssign = pAssignTmp;
-
-    } else {
-//      tmq_offset_seek(tmq, "tp", pAssign[0].vgId, pAssign[0].currentOffset);
-//      tmq_offset_seek(tmq, "tp", pAssign[1].vgId, pAssign[1].currentOffset);
-//      tmq_commit_sync(tmq, pRes);
-      continue;
-    }
-
-//    tmq_commit_sync(tmq, pRes);
-    if (pRes != NULL) {
-      taos_free_result(pRes);
-      //      if ((++count) > 1) {
-      //        break;
-      //      }
-    } else {
-//      break;
-    }
-
-//    tmq_offset_seek(tmq, "tp", pAssign[0].vgId, pAssign[0].begin);
-  }
-
-  tmq_free_assignment(pAssign);
-
-  code = tmq_get_topic_assignment(tmq, "tp", &pAssign, &numOfAssign);
-  if (code != 0) {
-    printf("error occurs:%s\n", tmq_err2str(code));
-    tmq_free_assignment(pAssign);
-    tmq_consumer_close(tmq);
-    taos_close(pConn);
-    fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-    return;
-  }
-
-  for(int i = 0; i < numOfAssign; i++){
-    printf("assign i:%d, vgId:%d, offset:%lld, start:%lld, end:%lld\n", i, pAssign[i].vgId, pAssign[i].currentOffset, pAssign[i].begin, pAssign[i].end);
-  }
-
-  tmq_consumer_close(tmq);
-  taos_close(pConn);
-  fprintf(stderr, "%d msg consumed, include %d rows\n", msgCnt, totalRows);
-}
 #pragma GCC diagnostic pop

source/common/CMakeLists.txt

@@ -16,6 +16,14 @@ ENDIF ()
 IF (TD_STORAGE)
   ADD_DEFINITIONS(-D_STORAGE)
   TARGET_LINK_LIBRARIES(common PRIVATE storage)
+
+  IF(${TD_LINUX})
+    IF(${BUILD_WITH_COS})
+      add_definitions(-DUSE_COS)
+    ENDIF(${BUILD_WITH_COS})
+
+  ENDIF(${TD_LINUX})
+
 ENDIF ()
 
 target_include_directories(

source/common/src/systable.c

@@ -314,7 +314,7 @@ static const SSysDbTableSchema userUserPrivilegesSchema[] = {
 static const SSysTableMeta infosMeta[] = {
     {TSDB_INS_TABLE_DNODES, dnodesSchema, tListLen(dnodesSchema), true},
     {TSDB_INS_TABLE_MNODES, mnodesSchema, tListLen(mnodesSchema), true},
-    {TSDB_INS_TABLE_MODULES, modulesSchema, tListLen(modulesSchema), true},
+    // {TSDB_INS_TABLE_MODULES, modulesSchema, tListLen(modulesSchema), true},
     {TSDB_INS_TABLE_QNODES, qnodesSchema, tListLen(qnodesSchema), true},
     {TSDB_INS_TABLE_SNODES, snodesSchema, tListLen(snodesSchema), true},
     {TSDB_INS_TABLE_CLUSTER, clusterSchema, tListLen(clusterSchema), true},

source/common/src/tdatablock.c

@@ -26,7 +26,7 @@ int32_t colDataGetLength(const SColumnInfoData* pColumnInfoData, int32_t numOfRo
     if (pColumnInfoData->reassigned) {
       int32_t totalSize = 0;
       for (int32_t row = 0; row < numOfRows; ++row) {
-        char* pColData = pColumnInfoData->pData + pColumnInfoData->varmeta.offset[row];
+        char*   pColData = pColumnInfoData->pData + pColumnInfoData->varmeta.offset[row];
         int32_t colSize = 0;
         if (pColumnInfoData->info.type == TSDB_DATA_TYPE_JSON) {
           colSize = getJsonValueLen(pColData);
@@ -142,7 +142,8 @@ int32_t colDataSetVal(SColumnInfoData* pColumnInfoData, uint32_t rowIndex, const
   return 0;
 }
 
-int32_t colDataReassignVal(SColumnInfoData* pColumnInfoData, uint32_t dstRowIdx, uint32_t srcRowIdx, const char* pData) {
+int32_t colDataReassignVal(SColumnInfoData* pColumnInfoData, uint32_t dstRowIdx, uint32_t srcRowIdx,
+                           const char* pData) {
   int32_t type = pColumnInfoData->info.type;
   if (IS_VAR_DATA_TYPE(type)) {
     int32_t dataLen = 0;
@@ -164,7 +165,6 @@ int32_t colDataReassignVal(SColumnInfoData* pColumnInfoData, uint32_t dstRowIdx,
   return 0;
 }
 
-
 static int32_t colDataReserve(SColumnInfoData* pColumnInfoData, size_t newSize) {
   if (!IS_VAR_DATA_TYPE(pColumnInfoData->info.type)) {
     return TSDB_CODE_SUCCESS;
@@ -188,16 +188,17 @@ static int32_t colDataReserve(SColumnInfoData* pColumnInfoData, size_t newSize)
 }
 
 static int32_t doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t currentRow, const char* pData,
-                         int32_t itemLen, int32_t numOfRows, bool trimValue) {
+                            int32_t itemLen, int32_t numOfRows, bool trimValue) {
   if (pColumnInfoData->info.bytes < itemLen) {
-    uWarn("column/tag actual data len %d is bigger than schema len %d, trim it:%d", itemLen, pColumnInfoData->info.bytes, trimValue);
+    uWarn("column/tag actual data len %d is bigger than schema len %d, trim it:%d", itemLen,
+          pColumnInfoData->info.bytes, trimValue);
     if (trimValue) {
       itemLen = pColumnInfoData->info.bytes;
     } else {
       return TSDB_CODE_TDB_INVALID_TABLE_SCHEMA_VER;
     }
   }
-  
+
   size_t start = 1;
 
   // the first item
@@ -230,8 +231,8 @@ static int32_t doCopyNItems(struct SColumnInfoData* pColumnInfoData, int32_t cur
   return TSDB_CODE_SUCCESS;
 }
 
-int32_t colDataSetNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData,
-                            uint32_t numOfRows, bool trimValue) {
+int32_t colDataSetNItems(SColumnInfoData* pColumnInfoData, uint32_t currentRow, const char* pData, uint32_t numOfRows,
+                         bool trimValue) {
   int32_t len = pColumnInfoData->info.bytes;
   if (IS_VAR_DATA_TYPE(pColumnInfoData->info.type)) {
     len = varDataTLen(pData);
@@ -262,7 +263,7 @@ static void doBitmapMerge(SColumnInfoData* pColumnInfoData, int32_t numOfRow1, c
 
   uint8_t* p = (uint8_t*)pSource->nullbitmap;
   pColumnInfoData->nullbitmap[BitmapLen(numOfRow1) - 1] &= (0B11111111 << shiftBits);  // clear remind bits
-  pColumnInfoData->nullbitmap[BitmapLen(numOfRow1) - 1] |= (p[0] >> remindBits);  // copy remind bits
+  pColumnInfoData->nullbitmap[BitmapLen(numOfRow1) - 1] |= (p[0] >> remindBits);       // copy remind bits
 
   if (BitmapLen(numOfRow1) == BitmapLen(total)) {
     return;
@@ -350,7 +351,7 @@ int32_t colDataMergeCol(SColumnInfoData* pColumnInfoData, int32_t numOfRow1, int
 
       pColumnInfoData->pData = tmp;
       if (BitmapLen(numOfRow1) < BitmapLen(finalNumOfRows)) {
-        char*    btmp = taosMemoryRealloc(pColumnInfoData->nullbitmap, BitmapLen(finalNumOfRows));
+        char* btmp = taosMemoryRealloc(pColumnInfoData->nullbitmap, BitmapLen(finalNumOfRows));
         if (btmp == NULL) {
           return TSDB_CODE_OUT_OF_MEMORY;
         }
@@ -622,7 +623,7 @@ int32_t blockDataToBuf(char* buf, const SSDataBlock* pBlock) {
 
     if (pCol->reassigned && IS_VAR_DATA_TYPE(pCol->info.type)) {
       for (int32_t row = 0; row < numOfRows; ++row) {
-        char* pColData = pCol->pData + pCol->varmeta.offset[row];
+        char*   pColData = pCol->pData + pCol->varmeta.offset[row];
         int32_t colSize = 0;
         if (pCol->info.type == TSDB_DATA_TYPE_JSON) {
           colSize = getJsonValueLen(pColData);
@@ -698,8 +699,7 @@ int32_t blockDataFromBuf(SSDataBlock* pBlock, const char* buf) {
   return TSDB_CODE_SUCCESS;
 }
 
-static bool colDataIsNNull(const SColumnInfoData* pColumnInfoData, int32_t startIndex,
-                                          uint32_t nRows) {
+static bool colDataIsNNull(const SColumnInfoData* pColumnInfoData, int32_t startIndex, uint32_t nRows) {
   if (!pColumnInfoData->hasNull) {
     return false;
   }
@@ -880,7 +880,6 @@ int32_t dataBlockCompar(const void* p1, const void* p2, const void* param) {
 }
 
 static int32_t blockDataAssign(SColumnInfoData* pCols, const SSDataBlock* pDataBlock, const int32_t* index) {
-
   size_t numOfCols = taosArrayGetSize(pDataBlock->pDataBlock);
   for (int32_t i = 0; i < numOfCols; ++i) {
     SColumnInfoData* pDst = &pCols[i];
@@ -1131,6 +1130,7 @@ static int32_t doEnsureCapacity(SColumnInfoData* pColumn, const SDataBlockInfo*
     if (tmp == NULL) {
       return TSDB_CODE_OUT_OF_MEMORY;
     }
+    // memset(tmp, 0, numOfRows * pColumn->info.bytes);
 
     // copy back the existed data
     if (pColumn->pData != NULL) {
@@ -1474,8 +1474,8 @@ size_t blockDataGetCapacityInRow(const SSDataBlock* pBlock, size_t pageSize, int
   int end = nRows;
   while (start <= end) {
     int mid = start + (end - start) / 2;
-    //data size + var data type columns offset + fixed data type columns bitmap len 
-    int midSize = rowSize * mid + numVarCols * sizeof(int32_t) * mid + numFixCols * BitmapLen(mid); 
+    // data size + var data type columns offset + fixed data type columns bitmap len
+    int midSize = rowSize * mid + numVarCols * sizeof(int32_t) * mid + numFixCols * BitmapLen(mid);
     if (midSize > payloadSize) {
       result = mid;
       end = mid - 1;
@@ -1669,7 +1669,7 @@ int32_t tEncodeDataBlock(void** buf, const SSDataBlock* pBlock) {
 
     if (pColData->reassigned && IS_VAR_DATA_TYPE(pColData->info.type)) {
       for (int32_t row = 0; row < rows; ++row) {
-        char* pData = pColData->pData + pColData->varmeta.offset[row];
+        char*   pData = pColData->pData + pColData->varmeta.offset[row];
         int32_t colSize = 0;
         if (pColData->info.type == TSDB_DATA_TYPE_JSON) {
           colSize = getJsonValueLen(pData);
@@ -1772,7 +1772,7 @@ static char* formatTimestamp(char* buf, int64_t val, int precision) {
 
 // for debug
 char* dumpBlockData(SSDataBlock* pDataBlock, const char* flag, char** pDataBuf) {
-  int32_t size = 2048*1024;
+  int32_t size = 2048 * 1024;
   *pDataBuf = taosMemoryCalloc(size, 1);
   char*   dumpBuf = *pDataBuf;
   char    pBuf[128] = {0};
@@ -1780,8 +1780,8 @@ char* dumpBlockData(SSDataBlock* pDataBlock, const char* flag, char** pDataBuf)
   int32_t rows = pDataBlock->info.rows;
   int32_t len = 0;
   len += snprintf(dumpBuf + len, size - len,
-                  "===stream===%s|block type %d|child id %d|group id:%" PRIu64 "|uid:%" PRId64
-                  "|rows:%" PRId64 "|version:%" PRIu64 "|cal start:%" PRIu64 "|cal end:%" PRIu64 "|tbl:%s\n",
+                  "===stream===%s|block type %d|child id %d|group id:%" PRIu64 "|uid:%" PRId64 "|rows:%" PRId64
+                  "|version:%" PRIu64 "|cal start:%" PRIu64 "|cal end:%" PRIu64 "|tbl:%s\n",
                   flag, (int32_t)pDataBlock->info.type, pDataBlock->info.childId, pDataBlock->info.id.groupId,
                   pDataBlock->info.id.uid, pDataBlock->info.rows, pDataBlock->info.version,
                   pDataBlock->info.calWin.skey, pDataBlock->info.calWin.ekey, pDataBlock->info.parTbName);
@@ -2156,21 +2156,21 @@ int32_t blockEncode(const SSDataBlock* pBlock, char* data, int32_t numOfCols) {
     data += metaSize;
     dataLen += metaSize;
 
-     if (pColRes->reassigned && IS_VAR_DATA_TYPE(pColRes->info.type)) {
-        colSizes[col] = 0;
-        for (int32_t row = 0; row < numOfRows; ++row) {
-          char* pColData = pColRes->pData + pColRes->varmeta.offset[row];
-          int32_t colSize = 0;
-          if (pColRes->info.type == TSDB_DATA_TYPE_JSON) {
-            colSize = getJsonValueLen(pColData);
-          } else {
-            colSize = varDataTLen(pColData);
-          }
-          colSizes[col] += colSize;
-          dataLen += colSize;
-          memmove(data, pColData, colSize);
-          data += colSize;
+    if (pColRes->reassigned && IS_VAR_DATA_TYPE(pColRes->info.type)) {
+      colSizes[col] = 0;
+      for (int32_t row = 0; row < numOfRows; ++row) {
+        char*   pColData = pColRes->pData + pColRes->varmeta.offset[row];
+        int32_t colSize = 0;
+        if (pColRes->info.type == TSDB_DATA_TYPE_JSON) {
+          colSize = getJsonValueLen(pColData);
+        } else {
+          colSize = varDataTLen(pColData);
         }
+        colSizes[col] += colSize;
+        dataLen += colSize;
+        memmove(data, pColData, colSize);
+        data += colSize;
+      }
     } else {
       colSizes[col] = colDataGetLength(pColRes, numOfRows);
       dataLen += colSizes[col];
@@ -2181,7 +2181,8 @@ int32_t blockEncode(const SSDataBlock* pBlock, char* data, int32_t numOfCols) {
     }
 
     colSizes[col] = htonl(colSizes[col]);
-//    uError("blockEncode col bytes:%d, type:%d, size:%d, htonl size:%d", pColRes->info.bytes, pColRes->info.type, htonl(colSizes[col]), colSizes[col]);
+    //    uError("blockEncode col bytes:%d, type:%d, size:%d, htonl size:%d", pColRes->info.bytes, pColRes->info.type,
+    //    htonl(colSizes[col]), colSizes[col]);
   }
 
   *actualLen = dataLen;
@@ -2283,7 +2284,7 @@ const char* blockDecode(SSDataBlock* pBlock, const char* pData) {
 }
 
 void trimDataBlock(SSDataBlock* pBlock, int32_t totalRows, const bool* pBoolList) {
-//  int32_t totalRows = pBlock->info.rows;
+  //  int32_t totalRows = pBlock->info.rows;
   int32_t bmLen = BitmapLen(totalRows);
   char*   pBitmap = NULL;
   int32_t maxRows = 0;
@@ -2310,8 +2311,9 @@ void trimDataBlock(SSDataBlock* pBlock, int32_t totalRows, const bool* pBoolList
         if (colDataIsNull_var(pDst, j)) {
           colDataSetNull_var(pDst, numOfRows);
         } else {
-          // fix address sanitizer error. p1 may point to memory that will change during realloc of colDataSetVal, first copy it to p2
-          char* p1 = colDataGetVarData(pDst, j);
+          // fix address sanitizer error. p1 may point to memory that will change during realloc of colDataSetVal, first
+          // copy it to p2
+          char*   p1 = colDataGetVarData(pDst, j);
           int32_t len = 0;
           if (pDst->info.type == TSDB_DATA_TYPE_JSON) {
             len = getJsonValueLen(p1);

source/common/src/tglobal.c

@@ -216,7 +216,11 @@ uint32_t tsCurRange = 100;                      // range
 char     tsCompressor[32] = "ZSTD_COMPRESSOR";  // ZSTD_COMPRESSOR or GZIP_COMPRESSOR
 
 // udf
-bool tsStartUdfd = true;
+#ifdef WINDOWS
+bool tsStartUdfd = false;
+#else
+bool    tsStartUdfd = true;
+#endif
 
 // wal
 int64_t tsWalFsyncDataSizeLimit = (100 * 1024 * 1024L);
@@ -236,6 +240,15 @@ int64_t tsStreamBufferSize = 128 * 1024 * 1024;
 int64_t tsCheckpointInterval = 3 * 60 * 60 * 1000;
 bool    tsFilterScalarMode = false;
 int32_t tsKeepTimeOffset = 0;  // latency of data migration
+int     tsResolveFQDNRetryTime = 100; //seconds
+
+char   tsS3Endpoint[TSDB_FQDN_LEN] = "<endpoint>";
+char   tsS3AccessKey[TSDB_FQDN_LEN] = "<accesskey>";
+char   tsS3AccessKeyId[TSDB_FQDN_LEN] = "<accesskeyid>";
+char   tsS3AccessKeySecret[TSDB_FQDN_LEN] = "<accesskeysecrect>";
+char   tsS3BucketName[TSDB_FQDN_LEN] = "<bucketname>";
+char   tsS3AppId[TSDB_FQDN_LEN] = "<appid>";
+int8_t tsS3Enabled = false;
 
 #ifndef _STORAGE
 int32_t taosSetTfsCfg(SConfig *pCfg) {
@@ -258,7 +271,43 @@ int32_t taosSetTfsCfg(SConfig *pCfg) {
 int32_t taosSetTfsCfg(SConfig *pCfg);
 #endif
 
-struct SConfig *taosGetCfg() { return tsCfg; }
+int32_t taosSetS3Cfg(SConfig *pCfg) {
+  tstrncpy(tsS3AccessKey, cfgGetItem(pCfg, "s3Accesskey")->str, TSDB_FQDN_LEN);
+  if (tsS3AccessKey[0] == '<') {
+    return 0;
+  }
+  char *colon = strchr(tsS3AccessKey, ':');
+  if (!colon) {
+    uError("invalid access key:%s", tsS3AccessKey);
+    return -1;
+  }
+  *colon = '\0';
+  tstrncpy(tsS3AccessKeyId, tsS3AccessKey, TSDB_FQDN_LEN);
+  tstrncpy(tsS3AccessKeySecret, colon + 1, TSDB_FQDN_LEN);
+  tstrncpy(tsS3Endpoint, cfgGetItem(pCfg, "s3Endpoint")->str, TSDB_FQDN_LEN);
+  tstrncpy(tsS3BucketName, cfgGetItem(pCfg, "s3BucketName")->str, TSDB_FQDN_LEN);
+  char *cos = strstr(tsS3Endpoint, "cos.");
+  if (cos) {
+    char *appid = strrchr(tsS3BucketName, '-');
+    if (!appid) {
+      uError("failed to locate appid in bucket:%s", tsS3BucketName);
+      return -1;
+    } else {
+      tstrncpy(tsS3AppId, appid + 1, TSDB_FQDN_LEN);
+    }
+  }
+  if (tsS3BucketName[0] != '<' && tsDiskCfgNum > 1) {
+#ifdef USE_COS
+    tsS3Enabled = true;
+#endif
+  }
+
+  return 0;
+}
+
+struct SConfig *taosGetCfg() {
+  return tsCfg;
+}
 
 static int32_t taosLoadCfg(SConfig *pCfg, const char **envCmd, const char *inputCfgDir, const char *envFile,
                            char *apolloUrl) {
@@ -580,6 +629,11 @@ static int32_t taosAddServerCfg(SConfig *pCfg) {
   if (cfgAddInt32(pCfg, "keepTimeOffset", tsKeepTimeOffset, 0, 23, CFG_SCOPE_SERVER) != 0) return -1;
   if (cfgAddInt32(pCfg, "maxStreamBackendCache", tsMaxStreamBackendCache, 16, 1024, CFG_SCOPE_SERVER) != 0) return -1;
   if (cfgAddInt32(pCfg, "pqSortMemThreshold", tsPQSortMemThreshold, 1, 10240, CFG_SCOPE_SERVER) != 0) return -1;
+  if (cfgAddInt32(pCfg, "resolveFQDNRetryTime", tsResolveFQDNRetryTime, 1, 10240, 0) != 0) return -1;
+
+  if (cfgAddString(pCfg, "s3Accesskey", tsS3AccessKey, CFG_SCOPE_SERVER) != 0) return -1;
+  if (cfgAddString(pCfg, "s3Endpoint", tsS3Endpoint, CFG_SCOPE_SERVER) != 0) return -1;
+  if (cfgAddString(pCfg, "s3BucketName", tsS3BucketName, CFG_SCOPE_SERVER) != 0) return -1;
 
   GRANT_CFG_ADD;
   return 0;
@@ -979,6 +1033,7 @@ static int32_t taosSetServerCfg(SConfig *pCfg) {
   tsKeepTimeOffset = cfgGetItem(pCfg, "keepTimeOffset")->i32;
   tsMaxStreamBackendCache = cfgGetItem(pCfg, "maxStreamBackendCache")->i32;
   tsPQSortMemThreshold = cfgGetItem(pCfg, "pqSortMemThreshold")->i32;
+  tsResolveFQDNRetryTime = cfgGetItem(pCfg, "resolveFQDNRetryTime")->i32;
 
   GRANT_CFG_GET;
   return 0;
@@ -1502,6 +1557,7 @@ int32_t taosInitCfg(const char *cfgDir, const char **envCmd, const char *envFile
     if (taosSetServerCfg(tsCfg)) return -1;
     if (taosSetReleaseCfg(tsCfg)) return -1;
     if (taosSetTfsCfg(tsCfg) != 0) return -1;
+    if (taosSetS3Cfg(tsCfg) != 0) return -1;
   }
   taosSetSystemCfg(tsCfg);
 

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

@@ -46,7 +46,7 @@ static int32_t mmDecodeOption(SJson *pJson, SMnodeOpt *pOption) {
     if (code < 0) return -1;
     tjsonGetInt32ValueFromDouble(replica, "role", pOption->nodeRoles[i], code);
     if (code < 0) return -1;
-    if(pOption->nodeRoles[i] == TAOS_SYNC_ROLE_VOTER){
+    if (pOption->nodeRoles[i] == TAOS_SYNC_ROLE_VOTER) {
       pOption->numOfReplicas++;
     }
   }
@@ -65,7 +65,7 @@ int32_t mmReadFile(const char *path, SMnodeOpt *pOption) {
   char      file[PATH_MAX] = {0};
   snprintf(file, sizeof(file), "%s%smnode.json", path, TD_DIRSEP);
 
-  if (taosStatFile(file, NULL, NULL) < 0) {
+  if (taosStatFile(file, NULL, NULL, NULL) < 0) {
     dInfo("mnode file:%s not exist", file);
     return 0;
   }